1 NO. 90-CI-6033 JEFFERSON CIRCUIT COURT DIVISION ONE (1) 2 3 JOYCE FENTRESS, ET AL. PLAINTIFFS 4 5 VS. DEPOSITION FOR PLAINTIFFS 6 7 SHEA COMMUNICATIONS, ET AL. DEFENDANTS 8 * * * * * * * * * * 9 10 DEPONENT: LORI SHELDON 11 DATE: SEPTEMBER 15, 1993 12 13 * * * * * * * * * * 14 15 16 REPORTER: KATHY NOLD 17 18 KENTUCKIANA REPORTERS SUITE 260 19 730 WEST MAIN STREET LOUISVILLE, KENTUCKY 40202 20 (502) 589-2273 Page 1 1 * * * * * * * * * * 2 3 UNITED STATES DISTRICT COURT SOUTHERN DISTRICT OF INDIANA 4 INDIANAPOLIS DIVISION 5 IN RE ELI LILLY AND COMPANY ) Prozac Products Liability ) MDL Docket No. 907 6 Litigation ) 7 * * * * * * * * * * 8 NO. 91-02496-A 9 JACKIE LYNN BIFFLE, ET AL ) IN THE DISTRICT ) COURT OF 10 V. ) DALLAS COUNTY, TEXAS ) 11 ELI LILLY & COMPANY AND ) 14TH JUDICIAL DISTA PRODUCTS COMPANY ) DISTRICT 12 * * * * * * * * * * 13 NO. 92-14775-E 14 RICHARD HAROLD CROSSETT, JR., ) IN THE 15 CHAD H. CROSSETT, AMY MICHELLE ) DISTRICT CROSSETT AND KRISTEN ANN CROSSETT, ) COURT OF 16 INDIVIDUALLY AND AS SURVIVORS OF ) AND ON BEHALF OF THE ESTATE OF ) 17 JOCQUETTA ANN CROSSETT, DECEASED ) ) 18 V. ) DALLAS COUNTY, ) TEXAS 19 ELI LILLY & COMPANY, DISTA ) PRODUCTS COMPANY, TEXAS ) 20 PSYCHIATRIC COMPANY, INC. ) D/B/A/ HCA WILLOW PARK ) 101ST JUDICIAL 21 HOSPITAL, JAMES K. WITSCHY, M.D., ) DISTRICT AND DOUG BELLAMY, ED.D. ) Page 2 1 * * * * * * * * * * 2 NO. A-921,405-C 3 MARIA GUADALUPE REVES ) IN THE 4 INDIVIDUALLY AND AS NEXT ) DISTRICT COURT FRIEND OF GRANT JULIAN REVES ) OF 5 A MINOR CHILD, AND ON BEHALF ) OF THE ESTATE OF CHRISTIAN ) 6 MARIE REVES, DECEASED ) ) ORANGE COUNTY, 7 V. ) TEXAS ) 8 ELI LILLY & COMPANY, DISTA ) PRODUCTS COMPANY, RAVIKUMAR ) 9 KANNEGANTI, M.D., HOSPITAL ) CORPORATION OF AMERICA, A ) 10 TENNESSEE CORPORATION, HEALTH ) SERVICES ACQUISITION CORP., ) 11 A DELAWARE CORPORATION, ) HCA PSYCHIATRIC COMPANY, A ) 12 DELAWARE CORPORATION, TEXAS ) PSYCHIATRIC CO., INC.. A/K/A ) 13 AND/OR D/B/A HCA BEAUMONT ) NEUROLOGICAL HOSPITAL, AND HCA ) 14 HEALTH SERVICES OF TEXAS, INC. ) 128TH JUDICIAL A/K/A AND/OR BEAUMONT ) DISTRICT 15 NEUROLOGICAL HOSPITAL ) 16 * * * * * * * * * * Page 3 1 IN THE UNITED STATES DISTRICT COURT 2 FOR THE WESTERN DISTRICT OF TEXAS SAN ANTONIO DIVISION 3 ELIZABETH T. SANCHEZ, ) 4 INDIVIDUALLY AND AS THE ) SURVIVING SPOUSE, MARGARET R. ) 5 SANCHEZ, INDIVIDUALLY AND NEXT ) OF FRIEND OF DEBRA JEAN ) 6 SANCHEZ, VERONICA MARIE ) SANCHEZ, EDWARDO ESTEBAN ) 7 SANCHEZ, AND MICHAEL ANTHONY ) SANCHEZ, CHILDREN; AND ALL ON ) 8 BEHALF OF THE ESTATE OF ) EDWARDO SANCHEZ ) 9 ) V. ) CIVIL ACTION NO. 10 ) SA93CA367 ELI LILLY AND COMPANY AND ) 11 DISTA PRODUCTS COMPANY ) 12 * * * * * * * * * * 13 IN THE UNITED STATES DISTRICT COURT FOR THE SOUTHERN DISTRICT OF TEXAS 14 HOUSTON DIVISION 15 MARIA SANCHEZ, INDIVIDUALLY ) AND AS NEXT FRIEND OF DEBORAH ) 16 SANCHEZ, VERONICA SANCHEZ, ) EDDIE SANCHEZ, AND MICHAEL ) 17 SANCHEZ, AND ON BEHALF OF THE ) ESTATE OF EDUARDO SANCHEZ ) 18 ) V. ) CIVIL ACTION NO. 19 ) H-93-1469 ELI LILLY AND COMPANY AND ) 20 DISTA PRODUCTS COMPANY, A ) DIVISION OF ELI LILLY AND ) 21 COMPANY ) Page 4 1 * * * * * * * * * * 2 STATE OF NEW YORK 3 SUPREME COURT COUNTY OF JEFFERSON 4 _____________________________________________ 5 STEPHANIE CAPONE, AS EXECUTOR OF THE ESTATE OF JOSEPH J. CAPONE, JR., AND 6 STEPHANIE CAPONE, INDIVIDUALL, NOTICE TO TAKE 7 PLAINTIFF, DEPOSITION UPON ORAL EXAMINATION 8 VS. INDEX NO. 93-251 9 ELI LILLY AND COMPANY, DISTA PRODUCTS 10 COMPANY, A DIVISION OF ELI LILLY AND COMPANY, FLOYD BAJJALY, M.D, 11 DEFENDANTS. 12 _____________________________________________ 13 * * * * * * * * * * 14 SUPREME COURT OF TEH STATE OF NEW YORK COUNTY OF ORANGE 15 --------------------------------------X BRUCE R. MALEN AS EXECUTOR OF THE : INDEX NO. 16 ESTATE OF BARBARA E. MALEN, AND OF : 4119/92 BRUCE R. MALEN, INDIVIDUALLY, : 17 : HON. PETER PLAINTIFF : PATSALOS, 18 : J.S.C. -against- : 19 : ELI LILLY & COMPANY, DISTA PRODUCTS : 20 COMPANY, A DIVISION OF ELI LILLY & : COMPANY, BARRY SINGER AND UNITED : 21 HOSPITAL, : : 22 DEFENDANTS. : --------------------------------------X 23 * * * * * * * * * * Page 5 1 ---------------------------------X 2 VALARIE J. FRIEDMAN AND DAVID : SUPERIOR COURT FRIEDMAN, HER HUSBAND, : OF NEW JERSEY 3 : LAW DIVISION: PLAINTIFF, : MIDDLESEX COUNTY 4 : DOCKET NO. : L-3191-91 5 VS. : : 6 ELI LILLY & COMPANY; DISTA : PRODUCTS INC, A DIVISION OF : 7 ELI LILLY & COMPANY; LISS : PHARMACY; MADISON PHARMACY AND : 8 JOHN DOES NOS. 1-25 (UNKNOWN : ENTITIES), : 9 : DEFENDANTS. : 10 ---------------------------------X 11 * * * * * * * * * * 12 SUPREME COURT OF THE STAET OF NEW YORK COUNTY OF SUFFOLK 13 -------------------------------------x 14 RHOMDA L. HALA and JOSEPH L. HALA, : 15 Plaintiffs, : Index No. 14869/90 16 - against - : 17 ELI LILLY & COMPANY and DISTA : PRODUCTS COMPANY, a DIVISION OF 18 ELI LILLY & COMPANY : 19 Defendants. : -------------------------------------x 20 21 * * * * * * * * * * Page 6 1 IN THE CIRCUIT COURT OF COOK COUNTY, ILLINOIS 2 COUNTY DEPARTMENT, LAW DIVISION 3 PATRICIA BRACH, ) ) 4 Plaintiff ) ) 5 v. )No. 92 L 13369 ) 6 ELI LILLY AND COMPANY, a foreign ) corporation; ALAN N. MILLER, M.D., ) 7 WILLIAM BRUINSMA, Psy.D., and ) CONDELL MEMORIAL HOSPITAL, ) 8 ) Defendants. ) 9 * * * * * * * * * * 10 IN THE CIRCUIT COURT OF COOK COUNTY, ILLINOIS 11 COUNTY DEPARTMENT - LAW DIVISION 12 RENATO DI SILVESTRO, Individually ) and as Special Administrator of ) 13 the Estate of JOHN DI SILVESTRO, ) Deceased, ) 14 ) Plaintiff, ) 15 ) v. ) No. 91 L 7881 16 ) ROBERT L. NELSON, et al., ) 17 ) Defendants, ) 18 ) GEORGE MELNICK, M.D. and PETER ) 19 FINK, M.D. ) ) 20 Respondents in Discovery.) 21 * * * * * * * * * * Page 7 1 IN THE CIRCUIT COURT OF COOK COUNTY, ILLINOIS 2 COUNTY DEPARTMENT, LAW DIVISION 3 JOAN M. GRYER, ) ) 4 Plaintiff, ) ) 5 v. ) No. 92 L 7387 ) 6 ELI LILLY AND COMPANY, et al., ) ) 7 Defendants. ) 8 * * * * * * * * * * 9 IN THE CIRCUIT COURT OF COOK COUNTY, ILLINOIS 10 COUNTY DEPARTMENT, LAW DIVISION 11 JENNIFER HAMMERLI, as Plenary ) Guardian of the Estate of RAY B. ) 12 HAMMERLI, a disabled person, ) ) 13 Plaintiff, ) ) 14 v. ) No. 92 L 2365 ) 15 ELI LILLY AND COMPANY, THE ) UPJOHN COMPANY, DICKIE KAY, M.D., ) 16 (former Respondent in Discovery), ) and RICHARD CZECHOWICZ (former ) 17 Respondent in Discovery), ) ) 18 Defendants. ) 19 * * * * * * * * * * Page 8 1 IN THE CIRCUIT COURT OF THE SIXTH JUDICIAL CIRCUIT 2 CHAMPAIGN COUNTY, ILLINOIS 3 LINDA GARDNER, Individually and ) as Special Administrator of ) 4 the Estate of SHANE GARDNER, ) deceased, ) 5 ) Plaintiff, ) 6 ) v. ) No. 91 L 1066 7 ) ELI LILLY AND COMPANY, a foreign ) 8 corporation, ) ) 9 Defendant. ) 10 * * * * * * * * * * 11 IN THE NINETEENTH JUDICIAL CIRCUIT COURT 12 LAKE COUNTY, ILLINOIS 13 JAMES E. SHEPPARD, Special ) Administrator of the Estate of ) 14 KENNETH K. SHEPPARD, Deceased, ) ) 15 Plaintiff ) ) 16 v. ) No. 93 L 124 ) 17 GOOD SHEPHERD HOSPITAL, a ) corporation, DR. STEWART SEGAL, ) 18 DR. SANFORD SHERMAN, DR. BRUCE ) CARLSON, DR. R. BERGLUND, and ELI ) 19 LILLY & COMPANY, a corporation, ) ) 20 Defendants. ) 21 * * * * * * * * * * Page 9 1 SUPERIOR COURT OF THE STATE OF CALIFORNIA 2 FOR THE COUNTY OF LOS ANGELES 3 DR. MARIUS SAINES, etc., et al., ) Case No: 4 ) SC 008331 Plaintiffs, ) 5 ) vs. ) 6 ) ELI LILLY & COMPANY, a corporation; ) 7 DISTA PRODUCTS COMPANY, a division ) of Eli Lilly & Company; and DOBS 1- ) 8 100, inclusive, ) ) 9 Defendants. ) ____________________________________) 10 11 * * * * * * * * * * Page 10 1 THE DEPOSITION OF LORI SHELDON, TAKEN AT THE 2 OFFICE OF BAKER & DANIELS, 300 NORTH MERIDIAN 3 STREET, SUITE 2700, INDIANAPOLIS, INDIANA 46204, 4 ON SEPTEMBER 15, 1993; SAID DEPOSITION TAKEN 5 PURSUANT TO NOTICE IN ACCORDANCE WITH THE RULES 6 OF CIVIL PROCEDURE. 7 * * * * * * * * * * 8 A P P E A R A N C E S 9 10 NANCY ZETTLER COUNSEL FOR GROUP A PLAINTIFFS 11 LEONARD M. RING AND ASSOCIATES, P.C. 111 WEST WASHINGTON AVENUE, SUITE 1333 12 CHICAGO, ILLINOIS 60602 13 LAWRENCE J. MYERS COUNSEL FOR ELI LILLY AND COMPANY 14 FREEMAN & HAWKINS 4000 ONE PEACHTREE CENTER 15 303 PEACHTREE STREET, N.E. ATLANTA, GEORGIA 30308-3243 16 MARGARET M. HUFF 17 ELI LILLY AND COMPANY LILLY CORPORATE CENTER 18 INDIANAPOLIS, INDIANA 46285 19 DENISE BRODSKY COUNSEL FOR GOOD SHEPHERD HOSPITAL 20 415 WASHINGTON STREET, SUITE 214 WAUKEGAN, ILLINOIS Page 11 1 MIGUEL A. RUIZ COUNSEL FOR DEFENDANTS CZECHOWICZ, FINK, BRUINSMA 2 CLAUSEN MILLER GORMAN CAFFREY & WITOUS 10 SOUTH LASALLE 3 CHICAGO, ILLINOIS 60603 4 PAUL J. CLEMENTI COUNSEL FOR DR. DICKIE KAY 5 HINSHAW & CULBERTSON 222 NORTH LA SALLE STREET, SUITE 300 6 CHICAGO, ILLINOIS 60601-1081 7 ROBIN HILL COUNSEL FOR DRS. WITSCHY AND KANNEGANTI 8 BAILEY AND WILLIAMS 3500 NCNB PLAZA 9 901 MAIN STREET DALLAS, TEXAS 75202-3714 10 PAUL SMITH 11 COUNSEL FOR PLAINTIFFS 745 CAMPBELL CENTER 2 12 8115 NORTH CENTRAL EXPRESSWAY DALLAS, TEXAS 75206 Page 12 1 I N D E X 2 3 DEPOSITION OF LORI SHELDON 4 5 DIRECT EXAMINATION BY MR. SMITH 15 6 CROSS EXAMINATION BY MS. ZETTLER 141 7 CROSS EXAMINATION BY MR. CLEMENTI 264 8 RECROSS EXAMINATION BY MR. ZETTLER 268 9 REDIRECT EXAMINATION BY MR. SMITH 270 10 CERTIFIED QUESTION 32 11 CERTIFICATE 275 12 ERRATA 276 13 14 EXHIBITS 15 PLAINTIFFS' EXHIBIT NO. 1 103 16 PLAINTIFFS' EXHIBIT NO. 2 109 PLAINTIFFS' EXHIBIT NO. 3 127 17 PLAINTIFFS' EXHIBIT NO. 4 198 PLAINTIFFS' EXHIBIT NO. 5 216 18 PLAINTIFFS' EXHIBIT NO. 6 222 Page 13 1 COMES LORI SHELDON, CALLED BY THE 2 PLAINTIFFS, AND AFTER FIRST BEING DULY SWORN, WAS 3 DEPOSED AND TESTIFIED AS FOLLOWS: 4 * * * * * * * * * * 5 MR. SMITH: What agreements have we 6 been doing? 7 MS. ZETTLER: We have a basic on-going 8 objection to the cross-noticing of these 9 depositions in cases other than Fentress. I 10 think the agreement, as far as objections, is all 11 objections except to form and -- what was the 12 other one? 13 MR. MYERS: Responsiveness has been 14 made pursuant to the request from somebody in one 15 of your Texas cases. So that's what we've been 16 doing. 17 MR. SMITH: Why don't we put the 18 agreement that we entered into in the last 19 deposition that was taken and stick it in this 20 one. 21 MR. MYERS: That's fine, but we've been 22 objecting to the form of the question and 23 responsiveness of the answer. The plaintiffs 24 have some objection about the cross-notices, Page 14 1 we've cross-noticed the depositions and I assume 2 somebody will take that up at some point at a 3 later date. 4 * * * * * * * * * * 5 DIRECT EXAMINATION 6 BY MR. SMITH: 7 Q. Would you state your name, 8 please, ma'am? 9 A. Lori Sheldon. 10 Q. How old are you, Ms. Sheldon? 11 A. I'm sorry? 12 Q. How old a lady are you? 13 A. Thirty-two years old. 14 Q. Where do you live? 15 A. XXXXXXXXXXXXXXXXXXXXXX 16 XXXXXXXXXXXX. 17 Q. Are you married, Ms. Sheldon? 18 A. Yes. 19 Q. What is your husband's name? 20 A. XXXXXXXXXXXXXXXXXXX. 21 Q. Does he work outside the home? 22 A. XXXX. 23 Q. What does he do? 24 A. XXXXXXXXXXXXXXXXXXXXXXXXXXX. Page 15 1 Q. For whom? 2 A. XXXXXXXXXXXXXXXXXXX. 3 Q. Has he ever worked for Eli 4 Lilly and Company in any capacity? 5 A. XXXX. 6 Q. Do you work outside of the 7 home, Ms. Sheldon? 8 A. Yes. 9 Q. What do you do for a living? 10 A. I'm a systems analyst for Eli 11 Lilly and Company. 12 Q. How long have you worked for 13 Eli Lilly and Company? 14 A. Nine years. 15 Q. So you would have started with 16 Lilly in approximately 1984? 17 A. 1984. 18 Q. Do you know your exact start 19 date? 20 A. Yes, June 18, 1984. 21 Q. What were you hired on to do at 22 Lilly initially? 23 A. I was hired as a systems 24 analyst. Page 16 1 Q. What is a systems analyst? 2 A. It's a job in which -- there 3 are a variety of job descriptions of a systems 4 analyst depending on the department. I guess I'm 5 not sure what question you're asking. What my 6 job description is now or what it was when I 7 hired? 8 Q. When you hired. 9 A. The offer letter was for a 10 systems analyst, and it was nondescript as to 11 which department. My initial job was in medical 12 systems. 13 Q. Does this have to do with 14 computers? 15 A. Yes. 16 Q. You're sort of a trained 17 computer nerd? 18 MR. MYERS: Object to the form. 19 A. Thank you. I work with 20 computers in varying capacities depending on the 21 assignment I'm on. 22 Q. What educational background did 23 you have to become a systems analyst? 24 A. A Bachelor's degree. Page 17 1 Q. From where? 2 A. From Iowa State University. 3 Q. When did you get that degree? 4 A. In May, 1984. 5 Q. Where did you go to high 6 school? 7 A. Denison High School. 8 Q. Denison? 9 A. Denison, D-E-N-I-S-O-N, in 10 Denison, Iowa. 11 Q. Did you know there's a Denison, 12 Texas? 13 A. Yes. 14 Q. Did you have any jobs while you 15 were at Iowa State as a student? 16 A. Yes. 17 Q. What did you do there? 18 A. Worked as a clerk in a Target 19 store. 20 Q. What was your major at Iowa 21 State? 22 A. It was a management degree with 23 emphasis in management information systems. 24 Q. And those management Page 18 1 information systems were related to computer-type 2 materials? 3 A. Yes. 4 Q. Would you now consider yourself 5 an expert in computer data entry and research? 6 MR. MYERS: Before she answers, let me 7 object to the form. What do you mean by computer 8 data entry and research? That's awfully broad, 9 Paul. 10 Q. Well, with respect to the 11 computers that are used at Eli Lilly and Company, 12 would you consider yourself as an expert in the 13 use of those computers? 14 A. Which type of computer? 15 Q. Any type of computer. 16 A. In that case, my answer would 17 be no. 18 Q. Why? 19 A. Because there are several 20 different types of systems and computers used at 21 Eli Lilly and Company, and I'm not an expert in 22 all of those systems. 23 Q. Which systems are you an expert 24 in? Page 19 1 A. I'm not sure I would classify 2 myself as an expert, but I'm familiar with the 3 use of Windows and in particular a software tool 4 to develop data entry systems in Windows. 5 Q. Does that software tool have a 6 particular name? 7 A. Power Builder is the name of 8 the tool. 9 Q. Does Lilly call that Power 10 Builder software tool any particular in-house -- 11 does it give it any particular in-house 12 designation? 13 A. No, just the the same name the 14 software company gives it. 15 Q. Power Builder? 16 A. Power Builder. 17 Q. Any other computer systems 18 there at Lilly that you consider yourself expert 19 in? 20 A. Again, I'm not sure I 21 understand exactly what expert -- what your 22 definition of expert is, I guess, so I'm not sure 23 how to answer the question. 24 Q. If you have an ability to use Page 20 1 these computers and answer questions and explain 2 how to use these computers of individuals you 3 work with? 4 A. Okay. I also have knowledge of 5 using DB2 on our corporate main frame. 6 Q. What is DB2? 7 A. DB2 is a data base management 8 system, and I have used the programming language 9 SAS to read data from that data base system. 10 Q. What is SAS? 11 A. SAS is a programming language. 12 Q. What does S-A-S stand for? 13 A. Statistical Analysis System. 14 Q. Any other particular 15 programming language, computer package, software 16 package, that you consider yourself expert in? 17 A. Not at this time. 18 Q. Before we get into these 19 systems in detail, why don't you take us, Ms. 20 Sheldon, through your experience at Lilly. Why 21 don't you start with the first job you did at 22 Lilly. 23 A. My first job at Lilly as a 24 systems analyst was in the medical department Page 21 1 working on Prozac. Initially in providing data 2 entry systems and edit of the data as it was 3 entered. And from that, I went to working with 4 reporting the data that had been entered. 5 Q. Your first job function would 6 have started in June of 1984? 7 A. Yes. 8 Q. How long did you do that? 9 A. As far as I can remember, for 10 approximately nine to twelve months. 11 Q. Was your work at that time 12 exclusively in the medical department and 13 exclusively working on Prozac clinical trials? 14 A. At that time it was, yes. 15 Q. And what exactly were you doing 16 during that period of time? 17 A. I would have been creating data 18 entry systems, writing programs that edited the 19 data that had been entered. 20 Q. What else? 21 A. Providing queries of data that 22 had been entered. 23 Q. Anything else during that first 24 nine to twelve months? Page 22 1 A. That's pretty much it, that's 2 all I can really remember of the job. 3 Q. And all of that you did 4 exclusively with Prozac? 5 A. No. I did some work with 6 another drug at the same time. 7 Q. What drug was that? 8 A. Tomoxetine, which was under 9 study at the same time. 10 Q. What is Tomoxetine? 11 A. An anti-depressant. 12 Q. Is it a specific serotonin 13 reuptake inhibitor? 14 A. I do not remember. 15 Q. Were there clinical trials on 16 Tomoxetine in progress at the time there were 17 clinical trials in progress on Prozac? 18 A. Yes. 19 Q. Can you give us an estimate 20 during that first nine to twelve months of the 21 amount of time you spent on Prozac versus 22 Tomoxetine? 23 A. It's purely an estimate, but it 24 was probably somewhere between sixty and Page 23 1 seventy-five percent on Prozac. 2 Q. Who was your immediate 3 supervisor during that period of time? 4 A. During the first nine to twelve 5 months? 6 Q. Yes, ma'am. 7 A. Maury Wolfred. 8 Q. Maury? 9 A. Maury, M-A-U-R-Y, Wolfred, 10 W-O-L-F-R-E-D. 11 Q. And what was his job title? 12 A. He was a department head. 13 Q. Department head of what? 14 A. Medical systems. 15 Q. And your official title during 16 that first nine to twelve months was? 17 A. I do not remember, I believe it 18 was information analyst. 19 Q. Did you receive any training, 20 Ms. Sheldon, there at Lilly concerning how to do 21 your job for a specific training period of time? 22 A. I would have received training 23 on the System 38, which was the computer used for 24 data entry. Page 24 1 Q. Did you go for classes on that? 2 A. Go for classes where? 3 Q. There at Lilly? 4 A. At Lilly, yes, yes, at Lilly. 5 Q. Did you have a period of time 6 initially where you spent most of your day in 7 class learning how to use these various systems? 8 A. Where I spent -- yes, where I 9 would have spent most of the day. 10 Q. How long did that last? 11 A. I don't remember, it would have 12 been less than a week, but I don't remember the 13 exact number of days. 14 Q. Did I understand it that the 15 System 38 was the name of the computer that data 16 from the clinical trials was being placed into? 17 A. That's where the data was being 18 entered. 19 Q. In a System 38? 20 A. In a System 38. 21 Q. Is that part of the main frame 22 computer there at Lilly? 23 A. No, it's not. 24 Q. It's a separate system? Page 25 1 A. Yes. 2 Q. On a separate computer? 3 A. Yes. 4 Q. And was that system at that 5 time being used for anything other than data 6 entry on Prozac-related clinical trials? 7 A. Yes, it was the machine used 8 for data entry of all clinical trials at Lilly. 9 Q. What was the capacity of that 10 system? 11 A. I do not know. 12 Q. Was the System 38 a computer 13 system designed specifically for Lilly? 14 A. No. 15 Q. Who designed that system? 16 A. IBM. 17 Q. Was it modified for use by 18 Lilly in their clinical trials? 19 A. I do not know that. 20 Q. Who at that time did you 21 consider most knowledgable there at Lilly 22 concerning the System 38 computer? 23 A. I don't remember who was in 24 charge of the operations at that time. Page 26 1 Q. Who was your teacher, who 2 taught you how to use the System 38? 3 A. I don't remember who taught the 4 training, who ran the training. 5 Q. Did you consider yourself, by 6 the time you were trained and by the time you 7 used the system for a while, as being proficient 8 in the use of the System 38? 9 A. After I used it for a while, 10 yes. 11 Q. By the end of that first nine 12 to twelve month period of time? 13 A. Yes. 14 Q. Any other specific training 15 that you got there initially at Lilly with 16 respect to your job duties? 17 A. I don't remember any other 18 particular training. There may have been, I just 19 don't recall. 20 Q. You say that first nine to 21 twelve months you were creating data entry 22 systems; is that right? 23 A. Uh-huh. 24 MR. MYERS: Yes? Page 27 1 A. Yes. 2 Q. You have to say yes and no. 3 A. Yes. 4 Q. And if I ask you any question 5 you don't understand and you need it to be 6 rephrased, would you let me know, please? 7 A. Sure will. 8 Q. Why was there a need for you to 9 create data entry systems for that first nine to 10 twelve months? 11 A. There were ongoing clinical 12 trials on the drugs and new trials starting 13 during that period. 14 Q. Would you need to create new 15 systems for ongoing clinical trials? 16 A. Not for ongoing clinical 17 trials, no. 18 Q. But if a new clinical trial was 19 begun during that period of time, you would have 20 to help design the system? 21 A. Yes, if it did not fit into the 22 current data entry systems. 23 Q. I notice in those documents 24 produced to us for your deposition, there were Page 28 1 several protocols in your file. Do you recall 2 that? 3 A. I don't know what was produced 4 for the deposition, so I don't know what was in 5 the files. 6 Q. You haven't had an opportunity, 7 Ms. Sheldon, to review documents that have been 8 produced by Lilly in connection with your 9 deposition here today? 10 A. No, I have not reviewed those. 11 Q. Have you reviewed any documents 12 in connection with giving your deposition here 13 today? 14 A. No, I have not. 15 Q. Have you asked to see some 16 documents prior to your giving your deposition 17 here today? 18 A. No, I have not. 19 Q. Have you discussed the subject 20 of your testimony here today with anyone? 21 A. Only with the lawyers 22 representing. 23 Q. And with what lawyers did you 24 have discussions? Page 29 1 A. With Larry and with Mary, 2 representing Lilly. 3 Q. Are you speaking of Ms. Huff 4 and Mr. Myers? 5 A. Yes. 6 Q. And when did you have those 7 discussions, Ms. Sheldon? 8 A. Approximately three weeks ago. 9 Q. Where? 10 A. At Lilly. 11 Q. In your office? 12 A. In a conference room. 13 Q. Were any documents shown to you 14 at that time? 15 A. No. 16 Q. And that discussion lasted 17 approximately three hours? 18 A. Yes. 19 Q. Was anybody else there present 20 at that meeting besides you, Mr. Myers and Ms. 21 Huff? 22 A. No. 23 Q. Since that time, have you done 24 any research or done anything to prepare yourself Page 30 1 for this deposition here this morning, Ms. 2 Sheldon? 3 A. No. 4 Q. Have you discussed the 5 substance of the conversation that you had with 6 Mr. Myers or Ms. Sheldon -- or Ms. Huff with 7 anyone since that meeting? 8 A. Yes. 9 Q. With whom? 10 A. With my husband. 11 Q. All right. And what did you 12 tell your husband concerning what was said at 13 that meeting? 14 MR. MYERS: Don't tell anything that 15 you discussed with us. I don't care who you told 16 it to, that's privileged. She's not going to 17 answer that, Paul. 18 MR. SMITH: Don't you think she 19 probably has violated the privilege once she 20 discussed it with somebody outside of Lilly? 21 MR. MYERS: I don't think she has, and 22 she's not going to answer that, so ask another 23 question. 24 MR. SMITH: Are you instructing her not Page 31 1 to answer that question? 2 MR. MYERS: Yes, I am. 3 MS. ZETTLER: Certify that. 4 MR. SMITH: Do you have to certify 5 these questions in Kentucky in order to take them 6 to the judge? 7 MR. MYERS: Yes, sir. 8 MR. SMITH: Then we would like to 9 certify that question. 10 (QUESTION CERTIFIED.) 11 Q. Have you discussed those 12 conversations that you had with Mr. Myers and Ms. 13 Huff with anybody else other than your husband? 14 A. No. 15 Q. Now, so we're clear, it's my 16 understanding you stated earlier that your 17 husband does not work for Eli Lilly and Company? 18 A. That's correct. 19 Q. And has never worked for Eli 20 Lilly and Company? 21 A. No, he's not. 22 Q. Is he a pharmacologist? 23 A. No. 24 Q. Is he a medical doctor? Page 32 1 A. No. 2 Q. Is he a systems analyst? 3 A. No. 4 Q. We were talking before we got 5 on that aside about the protocols that were in 6 your file. And the reason for my question was, 7 in creating these data entry systems, would you 8 have to refer to the protocols in order to 9 establish a data entry system for a new clinical 10 trial that was going to be done? 11 A. Yes. 12 Q. And what type of problems were 13 you attempting to solve as a computer analyst in 14 doing that? 15 MR. MYERS: In designing the system or 16 in looking at the protocols? 17 MR. SMITH: In designing the system in 18 order to comply with the protocols or to get the 19 information that the protocols required. 20 A. I'm not sure I understand how 21 to answer that question as it was stated. Could 22 you restate it? 23 Q. I'm not a systems analyst, and 24 I'm not a computer programmer. I'm a nerd, but Page 33 1 not a computer nerd. So I'm going to have some 2 difficulty with you, Ms. Sheldon, in explaining 3 myself because I'm certainly not sophisticated in 4 many areas, including computer programming, so 5 would you bear with me, please? 6 A. Yes. 7 Q. It's just that my initial 8 function is to ask you about how you would go 9 about creating a system designed to get 10 information that's going to be given to you with 11 respect to a particular clinical trial protocol. 12 A. The data entry systems are 13 designed based on clinical report forms. 14 Q. Right. 15 A. And the data entry system must 16 collect all the data that may be written on the 17 clinical report form. The clinical report form 18 is designed based on what is written in the 19 protocol. 20 Q. Does systems analysts have 21 anything to do with the design of the clinical 22 report form? 23 A. Yes. 24 Q. And in what respect do they Page 34 1 assist in the design of the clinical report form? 2 A. From a standpoint of 3 aesthetics, how it would be easiest to enter data 4 into the clinical report form, how would it be 5 easiest for a key punch operator to read the data 6 from the form and put it into the system. Some 7 checks and balances as to whether all of the 8 statements in the protocol have been represented 9 on the clinical report form. 10 Q. Were you confident during the 11 time that you were creating and using these 12 systems at Lilly that the clinical report form 13 and the particular data entry process was 14 consistent? In other words, was the clinical 15 report form easy to use or did it produce some 16 difficulties? 17 A. I'm not an expert on that. I 18 didn't use the clinical report forms, so I can't 19 answer that question from that standpoint. 20 Q. Who did use the clinical report 21 form with respect to the entry of data onto 22 computer systems at Lilly? 23 A. There was a department 24 specifically for data entry from the clinical Page 35 1 report forms, and it would be the people in that 2 department that actually entered the data from 3 the clinical report forms. 4 Q. What was the name of that 5 department? 6 A. It was -- I don't remember the 7 official name, it was a data entry department 8 within medical systems. 9 Q. All right. But in your job 10 function that first nine to twelve months, you 11 were working on creating data entry systems for 12 use by that department, were you not? 13 A. That is correct. 14 Q. You say additionally, that you 15 were writing programs to edit data? 16 A. Yes. 17 Q. Why would there be a necessity 18 to edit data? 19 A. To make sure that the data 20 entered matched what was on the clinical report 21 form, to provide checks of the data that was 22 actually on the clinical report form for 23 reasonableness. 24 Q. All right. So if you had a Page 36 1 blood pressure out of norm, within a certain 2 parameter, you would write a program that would 3 pick this up within the computer and 4 automatically flag it somehow? 5 A. Yes. 6 Q. Was that difficult to do? 7 MR. MYERS: To write the program? 8 MR. SMITH: Yes. So that it would pick 9 up abnormal values. 10 A. When I first started on the 11 job, it seemed difficult. By the time I had been 12 at it for a while, it was second nature, I 13 suppose. 14 Q. Additionally, you say that you 15 were involved in writing programs that would 16 query data that had been entered or that you 17 physically did query data that had been entered, 18 which one? 19 A. Yes -- I'm sorry, can you read 20 the question back? 21 (THE COURT REPORTER READ BACK THE 22 REQUESTED TESTIMONY.) 23 A. Both. 24 Q. Why would you want to query Page 37 1 data that had been entered? 2 A. That would have come as a 3 request from the medical community, just 4 fulfilling the request. 5 Q. So during that period of time 6 in early or in mid-1984 until mid-1985, Lilly had 7 the capacity to obtain information from their 8 data sources? 9 A. Yes. Information is a very 10 broad term. 11 Q. The information from the 12 clinical report forms that was put on the 13 computer? 14 A. Yes. 15 Q. That's the only information 16 that you were involved with at that time, wasn't 17 it? 18 A. No. 19 Q. What else did you do? 20 A. In addition, there was 21 information stored in files pertaining to the 22 date the clinical report form data was entered 23 into the system, the date that the data was 24 released for analysis on the main frame, Page 38 1 information about who entered the data, 2 information about who worked on cleaning the data 3 after it was entered, those types of data were 4 not entered physically by a person, but were 5 entered by the system any time activity was done 6 on the data. And I would have queried or written 7 queries for that data as well. 8 Q. When you say cleaning the data, 9 what do you mean by that? 10 A. Cleaning the data would be 11 reconciling discrepancies between what was in the 12 data base and what was on the physical form, or 13 making corrections that were what the physician 14 who was on the clinical trial would request. If 15 the physician sent in a correction, that would be 16 corrected in the data base. 17 Q. Why would there be a necessity 18 to correct information that was entered onto the 19 data base during this cleaning process? 20 A. That, I was not involved in 21 that. It was -- that would be speculation for me 22 to answer that question. 23 Q. Well, did you ever observe that 24 there were are errors made by individuals in the Page 39 1 data entry process in entering data onto the 2 computer system? 3 A. I found several cases where 4 errors like that would have occurred. 5 Q. Simply human errors? 6 A. Yes. 7 Q. What else did you do during 8 that first nine to twelve months? 9 A. That was my primary job 10 responsibility. It included attending status 11 meetings or staff meetings. There were other 12 parts of my job that were pretty much part of 13 every systems analyst's job at that point. 14 Q. Were you a systems analyst on 15 the Prozac team? 16 A. Yes. 17 Q. How many members were on the 18 Prozac team when you first joined the team? 19 A. I don't know the exact number. 20 Q. Well, you would have meetings, 21 would you not? 22 A. Yes. 23 Q. Where most of the members of 24 the team were present? Page 40 1 A. Yes. 2 Q. And back during that first nine 3 to twelve months, approximately how many members 4 would be there at those meetings? I know you 5 can't give me an exact figure. 6 A. There would be approximately 7 eight to ten people in any of those meetings. 8 Q. Initially during that first 9 nine to twelve months, who was heading up those 10 meetings? 11 A. The department head of CRAs, 12 Earleen Ashbrook. 13 Q. Were there medical personnels -- 14 was there physicians on these Prozac team project 15 meetings? 16 A. Yes. 17 Q. But they weren't in charge? 18 A. They were not in charge of the 19 status meetings. 20 Q. But would you meet with them? 21 A. Yes. 22 Q. But Ms. Ashbrook headed up the 23 team as far as the status meetings? 24 A. She led the meetings. Page 41 1 Q. And how long did she lead those 2 meetings? 3 A. During that nine to twelve 4 months or during the entire nine to twelve months 5 in question? 6 Q. Do you know how long she 7 continued in that capacity? 8 A. No, I don't. 9 Q. After that nine to twelve month 10 period, what was your next job assignment at 11 Lilly? 12 A. The next job assignment was 13 actually a continuation of this job assignment, 14 with additional responsibilities. 15 Q. Tell me about that. 16 A. During that period, I began 17 reporting data from the main frame, using the SAS 18 language I referred to before. 19 Q. When did you begin that? 20 A. I don't recall exactly, it 21 would have been sometime between December of '84 22 and February of '85. 23 Q. Did your title change? 24 A. No. Page 42 1 Q. Did you continue on the Prozac 2 team? 3 A. Yes. 4 Q. Did you report to someone else 5 after that period of time or at that period of 6 time? 7 A. At that period of time, no, I 8 did not change. 9 Q. You were still reporting to 10 Maury -- 11 A. Maury Wolfred. 12 Q. Maury Wolfred. 13 A. Yes. 14 Q. Who did he report to? 15 A. Charlie Sampson. 16 Q. And what was Charlie Sampson's 17 job title? 18 A. Manager. 19 Q. Manager of what? 20 A. Medical systems. 21 Q. Did your job title change when 22 you began reporting data directly from the main 23 frame using the SAS language? 24 A. No. Page 43 1 Q. You were still a systems 2 analyst? 3 A. An information analyst. 4 Q. What was different about what 5 you began doing in December to February than what 6 you had been doing the nine to twelve months 7 before? 8 A. During the first few months or 9 the first six to nine to twelve months, I worked 10 on the System 38 on data entry systems. During 11 that time, I also started picking up the 12 responsibility for writing reports in SAS against 13 data that was on the main frame. 14 Q. All right. Is that what -- 15 you're telling me that your job was enlarged to 16 where you began, after that first nine to twelve 17 months, writing reports in SAS against data on 18 the main frame? 19 A. Yes. 20 Q. The difference is that you 21 weren't using the main frame before this, but 22 then you began using the main frame as opposed to 23 the System 38? 24 A. I personally started working on Page 44 1 the main frame at that point. 2 Q. And the System 38 and the main 3 frame are two different? 4 A. They are two different 5 machines. 6 Q. What is the difference in those 7 two machines in lay language, please? 8 A. What type of difference are you 9 looking for? 10 Q. Something that I can 11 understand. 12 A. There's a difference in how the 13 data bases are structured, there's a difference 14 in the tools that are available for reporting and 15 analyzing, there are differences in the physical 16 structure of the machines, that's beyond what I'm 17 comfortable with, I'm not an expert in that 18 field. There are differences in sign-on 19 procedures, there are many differences in the 20 machines, the major difference being the tools 21 that are available. 22 Q. I would assume that the main 23 frame has much greater capacity than the System 24 38 with respect to both the storage capacity and Page 45 1 the ability to retrieve that information that's 2 stored on the computer? 3 A. Yes. The CPU or central 4 processing unit cycling would be faster on the 5 main frame, it's a faster machine. Some of the 6 differences I mentioned in how data is stored 7 would have made it more desirable to do analysis 8 on the main frame, it was a different data base 9 structure. 10 Q. Why would you have two 11 different computers and computer systems, such as 12 the System 38 and the main frame? 13 A. I was not involved in the 14 decision on where to put the data, I can't answer 15 how that decision was made. 16 Q. Are both systems two different 17 storage systems? 18 A. Essentially, yes. 19 Q. And both systems have the 20 ability to input data; correct? 21 A. Yes. 22 Q. And both systems have the 23 ability to retrieve data? 24 A. Yes. Page 46 1 Q. My question is: From a 2 practical standpoint, do you have any idea why 3 there would be use of two different computers in 4 this storage and retrieval process? 5 A. Again, it goes back to the 6 types of tools that are available on the 7 different machines. It's my understanding that 8 on the System 38 there are better data entry 9 tools, on the main frame there are better 10 reporting analysis tools. 11 Q. So by moving the information 12 from the System 38 to the main frame, you might 13 increase the ability to get information from that 14 pool of information? 15 A. Yes. 16 Q. But the System 38 might be 17 easier to actually program into? 18 A. Could you rephrase that 19 question, I'm not sure how you mean the term 20 program into. 21 Q. Well, it's easier to enter data 22 into, maybe. 23 A. Yes. 24 Q. All this data that's entered Page 47 1 into the System 38 is entered based on the 2 clinical report forms, as I understand it; is 3 that correct? 4 A. Yes. 5 Q. And the System 38 contains all 6 the data that is received from the case report 7 forms? 8 A. Yes. 9 Q. And, then, is that data then 10 transferred to the main frame? 11 A. What do you mean by 12 transferred? 13 Q. Physically put into one 14 physical computer into the main frame. 15 A. There is a copy of that data 16 made on the main frame. 17 Q. So it -- the information 18 remained on the 38, but it was put in addition 19 onto the main frame? 20 A. Yes. 21 Q. Was this done at one time or 22 was this done at a different time? 23 A. I'm not sure I understand. 24 Q. Originally, was the data just Page 48 1 put into the System 38 and then later it was 2 decided to also enter in that information into 3 the main frame? 4 A. I'm still not understanding the 5 question. 6 Q. It's probably because I don't 7 understand the process of how this was done. 8 Would one individual, as this data was coming in, 9 program -- or not program, but enter data in the 10 System 38 and the main frame at the same time -- 11 A. No. 12 Q. -- with the same data entry? 13 A. No. 14 Q. So how would that get from one 15 system to the other? 16 A. A copy of the data would have -- 17 there were procedures in place to take copies of 18 the data that was to be transferred to the main 19 frame on tape, and then procedures on the main 20 frame to load that data into the data bases on 21 the main frame. 22 Q. Could it be said that all data 23 that was ever entered into the System 38 would 24 have been at some time, some how, transferred Page 49 1 onto the main frame? 2 A. All data related to -- 3 MR. MYERS: Are you talking Fluoxetine? 4 A. Are you talking about 5 Fluoxetine or are we talking about medical, all 6 medical data across all drugs we study or -- 7 Q. Just Prozac. 8 A. Just Prozac. I believe all 9 data was transferred to the main frame. 10 Q. You say you believe. Are you 11 uncertain about that? 12 A. I'm uncertain about that. 13 Q. Why? 14 A. I do not know for sure whether 15 the data that was stored in the system regarding 16 the dates of data entry, some of that information 17 that I described earlier, I do not know whether 18 that was transferred to the main frame. 19 Q. Do you know whether or not all 20 data that was entered into the System 38 with 21 respect to information off the case report forms 22 got transferred from the System 38 into the main 23 frame? 24 A. I believe it all did. Page 50 1 Q. Do you know of anybody that 2 would be certain about that, who would be most 3 knowledgable concerning the data that was 4 transferred from the System 38 into the main 5 frame? 6 A. That might be an analyst who is 7 currently working in that area and has access to 8 that data. 9 Q. Well, who, at that time that 10 you were working with, would you think would be 11 the most knowledgable about that? 12 A. At that time, I would have 13 been. 14 Q. All right. So at that time, 15 would it be your testimony that you felt like 16 that you were of the opinion at that time that 17 all case report form data had been or was being 18 transferred from the System 38 to the main frame? 19 A. That is -- my memory is -- this 20 is going back almost nine years. To the best of 21 my recollection, it was. 22 Q. All right. You don't recall 23 any instances where there was some data -- that 24 you have a recollection that there was some data Page 51 1 that was not transferred from the System 38 into 2 the main frame? 3 A. I do not recall any cases where 4 that would have been. 5 Q. Was that part of your job 6 during that period of time to ensure that the 7 data was transferred? 8 A. Yes. 9 Q. Help me, if you would, with 10 what S-A-S language is. You say it stands for 11 Statistical Analysis Systems; is that right? 12 A. I believe so, yes. I may be 13 off on the acronym slightly, but it is a 14 statistical analysis software, it's both a 15 programming language and there is a data base 16 structure in the SAS products. This is a 17 language that we would program reports to 18 retrieve the data out of the data bases, and 19 produce paper reports of what data was in those 20 data bases. 21 Q. When you talk about computer 22 language, what do you mean? 23 A. It is a -- there is a mechanism 24 for writing lines of programming language code, Page 52 1 there is a compiler that interprets that code and 2 does -- basically turns it into machine level 3 instruction to the computer with a what to do. 4 In the case of SAS, it's a high level language. 5 Basically, it keeps a systems analyst from having 6 to know, to flip switches, ones and zeros like 7 when programs were first written on computers 8 back in the '50s. It's a means of interpreting 9 what to tell the machine to do is what a 10 programming language is. 11 Q. And you were instrumental in 12 writing the SAS language or in using it? 13 A. No, in using the SAS language. 14 Q. Who wrote it? 15 A. A company called SAS Institute. 16 Q. All right. They come or came 17 to Lilly and wrote a particular language? 18 A. No. This is a commercially 19 available software language used by many 20 different companies. 21 Q. Was this S-A-S language in 22 place when you switched over to using the main 23 frame? 24 A. Yes. Page 53 1 Q. Did you at any time have any 2 input into adding or deleting S-A-S type 3 language? 4 A. I don't understand the 5 question. 6 Q. Well, S-A-S is a specific type 7 of language; is that right? 8 A. Yes. 9 Q. The language of a computer 10 tells you -- tells the programer or the person 11 using the computer what specifically to type in 12 to get specific information? 13 A. Yes. 14 Q. Did you at any time have any 15 input in making additions, deletions or 16 corrections to the way you used the language? 17 A. I think that might be a 18 different question than the first one. I'm 19 trying to understand exactly what you're asking 20 here. Did I have any input into changing the 21 language, no. 22 Q. All right. What kind of 23 changes, then, did you make? 24 A. To what, in reference to what? Page 54 1 Q. To in any way effect the 2 language? 3 A. I did not have any input in 4 effecting the SAS language, that is written by 5 SAS Institute and available to many different 6 companies, I did not impact the language. 7 Q. Did you do anything that in any 8 way affected how the SAS language was used? 9 A. Yes. 10 Q. All right. Tell me about that, 11 please? 12 A. I worked with several other 13 people in the department to help write some 14 programming standards for how to use the 15 language. 16 Q. Anything else? 17 A. No. 18 Q. All right. What do you mean by 19 programming standards? 20 A. That would mean naming 21 conventions for variables that are used in the 22 programs. 23 Q. What does the word conventions 24 mean in connection with computer technology? Page 55 1 A. Conventions means a definition 2 or a written standard to try and get all 3 programers to write in a similar style so that 4 it's understood by other programers. 5 Q. Give me an example. 6 A. An example would be if 7 referring to the field that identifies a patient 8 by number, that all programers would refer to 9 that field as patient rather than pat numb or 10 patient numb or -- it's just -- we all use the 11 same word so that we understand that that's what 12 it means in all programs. 13 Q. What do you mean by variables 14 in computer terminology? 15 A. A variable is a way of 16 identifying to the system a holding place for a 17 piece of information. 18 Q. So these computers had specific 19 holding places for pieces of information? 20 A. Yes. 21 Q. And you would have to know the 22 right term to get into that holding place for a 23 piece of information? 24 A. No, it's more -- it's not -- Page 56 1 that's not exactly the way it is. That holding 2 place can be referred to however the analyst 3 wants to refer to it, it's just a bucket that 4 data goes into and can be brought back out. And 5 what we did with conventions was attempt to use 6 those buckets with all the same names so that we 7 understood from one program to another, from one 8 report to another, how to refer to those holding 9 places for data during the program. 10 Q. What else did you do in 11 connection with the S-A-S language? 12 A. I wrote programs using that 13 language to retrieve data from data bases and 14 report that data. 15 Q. Okay. Would somebody give you 16 an assignment and say Ms. Sheldon, we need to 17 know a specific, particular, group of 18 information. For instance we need to know how 19 many males were in a particular study, during a 20 particular month, in the Prozac studies, and you 21 would go write a program that would get that 22 information out of the buckets; is that what 23 you're saying? 24 A. No, I would write -- it's a Page 57 1 difference in terminology. I would write a 2 program that would get that data from the data 3 bases where that data was stored. 4 Q. So you know there is, as it 5 comes in, data is stored from the CRF concerning 6 the sex of a participant in a clinical trial; 7 correct? 8 A. Yes. 9 Q. And you know the data comes in 10 with respect to the date that particular 11 individual was within a clinical trial? 12 A. Yes. 13 Q. And, so, you need to write a 14 program in order to get those two pieces of 15 information out of the data base; is that what 16 you're saying? 17 A. Yes. 18 Q. Would you use the S-A-S 19 language or would you have to convert that into 20 the S-A-S language? 21 A. I would use the S-A-S, the SAS 22 language for that. 23 Q. Anything else you did in 24 connection with the use of the main frame and the Page 58 1 S-A-S language during that period of time that 2 you were doing that? 3 A. What period of time are we 4 taking about here? 5 Q. We're talking that next period 6 of time when you were still on the Prozac team, 7 but apparently moved from working primarily with 8 the System 38 into that period of time from 9 December of '84 to February of '85 where you used 10 the main frame and S-A-S language. 11 A. And the question was -- can you 12 read the question back? 13 Q. What else did you do during 14 that period of time? 15 A. I, as part of the team, would 16 work with reading and reviewing protocols, 17 helping with the design of case report forms or 18 clinical report forms. 19 Q. What else? 20 A. I think that's pretty much all. 21 Q. How long did you do that? 22 A. Until the Summer of 1987, 23 sometime in the Summer of 1987. 24 Q. Did you work during that period Page 59 1 from December '84 or February '85 to the Summer 2 of 1987, did that work exclusively deal with 3 Prozac? 4 A. No. 5 Q. What other drugs did you work 6 on? 7 A. I worked for a brief period on 8 another drug called Indecainide. 9 Q. I-N-D-E-C-A-N-I-N-E? 10 A. I-N-D-E-C-A-I-N-I-D-E, I 11 believe. 12 Q. What kind of drug is that? 13 A. It is a cardiovascular drug. 14 Q. No other anti-depressants? 15 A. No. 16 Q. Did you discontinue your work 17 on Tomoxetine -- 18 A. Yes. 19 Q. -- when you switched to using 20 the main frame and the S-A-S language? 21 A. I don't know the exact time 22 frame that I stopped working with Tomoxetine. It 23 was sometime during the first year that I worked 24 at the company, but I don't know the exact time Page 60 1 frame of that. 2 Q. Do you know what happened to 3 Tomoxetine? 4 A. No, I don't. 5 Q. You don't know whether it was 6 ever approved? 7 A. It was not approved, I don't 8 know what the status of that drug is. 9 Q. Okay. In the Summer of 1987, 10 what did you begin to do? 11 A. I took a new assignment within 12 medical as a data base administrator. 13 Q. Did you work on Prozac while 14 you were a data base administrator? 15 A. Yes. 16 Q. How long were you a data base 17 administrator? 18 A. Until January, 1989. 19 Q. Did you work on Prozac from the 20 Summer of 1987 until January of 1989 when you 21 were a data base administrator? 22 A. Not during that entire time 23 frame, no. 24 Q. Did you discontinue your work Page 61 1 on Prozac data at one specific time? 2 A. I discontinued working on 3 Prozac in the Summer of 1987, and in the Fall of 4 1987 was asked to return to working on it for a 5 brief period of time, and then went back to my 6 data base administrator role when that particular 7 project was completed. 8 Q. How long during that brief 9 period of time from the Fall of 1987 did you work 10 on Prozac? 11 A. I don't know the exact amount 12 of time, it was approximately two to three 13 months. 14 Q. Would it have run into 1988 or 15 do you think it was all done in 1987? 16 A. It was all done in 1987. 17 Q. What specific project was that? 18 A. Response to the approvable 19 letter from the FDA. 20 Q. Response to the approval 21 letter? 22 A. Approvable letter from the FDA. 23 Q. What was the approvable letter 24 from the FDA? Page 62 1 A. That was a letter from the FDA 2 with -- I don't know the exact language of the 3 letter. My interpretation of it was that they 4 intended to approve the drug once we had 5 responded to the remaining questions that were in 6 the letter. 7 Q. All right. But this was some 8 document that was sent to Lilly from the Food and 9 Drug Administration? 10 A. Yes. 11 Q. And there needed to be response 12 to specific questions raised by that letter? 13 A. Yes. 14 Q. And they asked you to help with 15 that? 16 A. Yes. 17 Q. I assume they asked for your 18 help in the statistical data processing area or 19 systems analyst analysis area? 20 A. Could you repeat the question? 21 Q. What did you do? 22 A. Thank you. I wrote programs to 23 report the data that was in the Fluoxetine data 24 bases. Page 63 1 Q. I guess you had a copy of that 2 letter from the Food and Drug Administration? 3 A. Yes, I did. 4 Q. And that became a part of your 5 file? 6 A. Yes. 7 Q. And did you make notes and 8 write memos in response to that letter? 9 A. I do not remember. 10 Q. But you know you had the 11 letter? 12 A. Yes. 13 Q. And you know it was in your 14 file? 15 A. Yes. 16 Q. And after that two or three 17 month period of time, what did you do with your 18 file? 19 A. It was in my file drawer in my 20 desk. 21 Q. All right. And then at some 22 point did you leave that particular office? 23 A. Yes. 24 Q. What did you do with that file? Page 64 1 A. It would have been moved to the 2 file drawer in my next office, whatever that was. 3 Q. Then you maintained that file 4 for a while? 5 A. What do you mean by maintained? 6 Q. That file that had that FDA 7 approval letter in it, that went with you? 8 A. Yes. 9 Q. How long -- do you still have 10 that letter in your file? 11 A. No. 12 Q. When did you give up that 13 letter and your file? 14 A. I'm not sure of the exact time 15 frame, approximately two years ago or eighteen 16 months ago. 17 Q. And what did you do with it? 18 A. I sent it to our legal 19 department at Lilly. 20 Q. When you sent it to your legal 21 department at Lilly, was that in response to some 22 particular memo? 23 A. Yes. 24 Q. And did that memo request that Page 65 1 you transfer to them all information you had 2 concerning your work on Fluoxetine or Prozac? 3 A. Yes. 4 Q. And did you do that? 5 A. Yes. 6 Q. Did you make some type of log 7 or record in any way what information was 8 contained in your file? 9 A. No. 10 Q. Was that done by the law 11 department? 12 A. I do not know. 13 Q. Did they ever return to you 14 some log on that information that was in your 15 file? 16 A. No. 17 Q. But the last time -- or you 18 know that at the time you turned over your file, 19 this FDA approvable letter was in your file? 20 A. Yes. 21 MR. SMITH: We didn't get it, Larry. 22 MR. MYERS: I haven't looked through 23 the file to know what's in it or what's not in 24 it, so I don't know that you got it or not, I Page 66 1 can't comment on that. 2 MR. SMITH: Well, as I understand it, 3 we have requested that this lady's file be 4 produced for us to question her, and four boxes 5 were delivered to Ms. Zettler last week, and the 6 approvable letter is not in the file. 7 MR. MYERS: I have an understanding 8 that in fact there were four boxes transferred, I 9 don't know whether the approvable letter is in 10 there or not, I just don't know that. 11 MR. SMITH: Who would know that? 12 MR. MYERS: I don't know, Paul, and I'm 13 not under examination. 14 MR. SMITH: I'm not trying to examine 15 you, Larry. The reason I want to know is because 16 I have looked at every page in those four boxes 17 of documents, and the approvable letter is not 18 within that set of documents. I need, in order 19 to question this lady intelligently, to have that 20 information. As I understand it from the 21 witness, the information was in her file and 22 transferred to legal. 23 MR. MYERS: I'm really not sure what 24 you're saying or asking me. Page 67 1 MR. SMITH: Can you give me any 2 assistance in getting me that document or 3 explaining why that document was not produced? 4 MR. MYERS: I can probably look into it 5 at the lunch break, but I doubt if I can give you 6 any assistance immediately in going through those 7 boxes. 8 MR. SMITH: Well, our problem is this 9 is not the first time this has occurred where a 10 witness has spoken of a document that was turned 11 over to legal that has not been produced to us. 12 MR. MYERS: I don't know that the 13 letter has not been produced to you at some other 14 time, I just don't know that. 15 MR. SMITH: Whether it was produced at 16 any other time, I'm telling you it has not been 17 produced for purposes of questioning this 18 witness. 19 MR. MYERS: Well, I can make an inquiry 20 is all I can tell you, Paul. I mean I really 21 don't know what else to tell you. 22 Q. Do you remember, Ms. Sheldon, 23 how many pages this document was? 24 A. No, I do not. Page 68 1 Q. Can you give me -- was it five 2 pages, thirty pages? 3 A. I do not remember, I'm sorry. 4 Q. Do you remember it as being a 5 lengthy document? 6 A. I do not remember the size of 7 the document. 8 Q. You worked with the document 9 for two or three months, did you not? 10 A. Yes, I did. 11 Q. And in answer or attempted to 12 write programs to answer specific questions 13 concerning that document? 14 A. Yes. 15 Q. And that document was a letter 16 from the Food and Drug Administration where they 17 were requesting information in order to approve 18 Prozac for human consumption, were they not? 19 A. Yes. 20 Q. What did you do, what specific 21 assignments were you given in connection with 22 that particular document? 23 A. I was assigned to writing 24 programs to respond to specific questions. I do Page 69 1 not recall the questions, the specific questions. 2 Q. Were those specific questions 3 listed in the letter? 4 A. I believe they were. 5 Q. And were there memos in 6 addition to those specific questions that posed 7 other questions to you, to answer, to help in 8 giving this response to the FDA letter? 9 A. I do not know if there were 10 other memos. 11 Q. Do you know if you addressed 12 related issues brought up by the FDA approvable 13 letter? 14 A. Yes. 15 Q. What related issues did you 16 address? 17 A. I do not remember the specific 18 issues. 19 Q. Do you remember that you made 20 notes or had memos concerning those related 21 issues? 22 A. I do not remember what memos or 23 notes there would have been. 24 Q. If there had been memos or Page 70 1 notes concerning that, would you have put them in 2 your file? 3 A. At the time they were written, 4 I would have, if there were any. 5 Q. And would you have maintained 6 them in your file? 7 A. I don't know that. 8 Q. Did you make a practice of 9 maintaining memos from your superiors and 10 subordinates during that time in your file? 11 A. During which time? 12 Q. During that time, that two to 13 three month period, at the end of 1987? 14 A. Yes, I would have. 15 Q. Would this be in that file that 16 you turned over to legal two years ago to 17 eighteen months ago? 18 A. I don't know the answer to 19 that. 20 Q. If it were in there, it would 21 have been turned over to legal, would it not? 22 A. Yes. 23 Q. You didn't make a -- you didn't 24 make a practice of taking material out of your Page 71 1 file once you put it in your file, did you? 2 A. Yes, when I left the job I 3 would have given any of my old memos to the 4 person replacing me. 5 Q. Who replaced you? 6 A. I don't remember. 7 Q. Well, take some time and see if 8 you can remember who replaced you. 9 A. I don't remember who was 10 assigned to the drug when I left it. There were 11 other analysts on the drug when I left it, I 12 don't know who specifically replaced me on any 13 given task. 14 Q. What other analysts were there 15 then? 16 A. Greg Cunningham was the other 17 analyst that was working on it. 18 Q. Is he the only other analyst 19 that was working on it? 20 A. I don't know that, that's the 21 only name I can recall. 22 Q. Would it be reasonable to 23 assume that Mr. Cunningham might have received 24 some of the information that you had generated Page 72 1 and maintained in your file? 2 A. That is possible. 3 Q. Or it's possible that it was 4 maintained in your file, also. And not 5 transferred to him? 6 A. I don't know, I don't know 7 that. 8 Q. I may have asked you this, and 9 I apologize if I have, did you go through your 10 file when you turned it over to legal, did you 11 review it? 12 A. I don't remember whether I did 13 or not. I presume I did to determine that what 14 was in the file was related to Prozac, but I do 15 not specifically remember a review. 16 Q. Did you have any role in 17 determining who would be given your assignments 18 when you left? 19 A. No. 20 Q. Do you know who would have made 21 that determination? 22 A. The management at that time. 23 Q. Who was your supervisor at that 24 time? Page 73 1 A. My supervisor at that time was 2 Tim Krupa. 3 Q. Tim Krupa, K-R-U-P-A? 4 A. K-R-U-P-A. 5 Q. Was there a particular 6 individual that was heading up the writing of 7 this response to the approval letter from the 8 Food and Drug Administration, approvable letter? 9 A. I do not recall which 10 individual was in charge. 11 Q. Who do you think it would have 12 been? 13 A. I would care not to speculate. 14 Q. Can you give me an either/or? 15 A. It would have been Max Talbott 16 from regulatory, Earleen Ashbrook from the 17 medical community or Joe Wernicke, the physician. 18 Q. Was there anybody within the 19 systems analyst group during that period of time 20 that was assisting you in this? 21 A. Yes. 22 Q. Who? 23 A. Greg Cunningham. 24 Q. Anybody else? Page 74 1 A. In the medical systems 2 community, no. 3 Q. In any systems community? 4 A. I would have received help from 5 operations in getting jobs run. 6 Q. Who would have been the person 7 in operations that was helping you? 8 A. I don't know, I did not know 9 the individuals. 10 Q. And after this two or three 11 month period of time, what did you do? 12 A. I returned to my job as a data 13 base administrator. 14 Q. Once you returned to that job 15 in, I guess it would have been January, 1988, or 16 early 1988 -- 17 A. It was either late 1987 or 18 early 1988. 19 Q. You didn't work on Prozac after 20 that? 21 A. No. 22 Q. Have you worked on Prozac since 23 then? 24 A. No. Page 75 1 Q. You say you continued as a data 2 base administrator until January, 1989. What did 3 you do after January, 1989? 4 A. I moved to another department 5 in marketing systems. 6 Q. What was that department? 7 A. Marketing information systems. 8 Q. Did that have to do at all with 9 the marketing of Prozac? 10 A. Did what have to do with the 11 marketing of Prozac? 12 Q. What you were doing in the 13 marketing information systems in January of 1989 14 have anything to do with Prozac? 15 A. Yes, it had to do with the 16 marketing of all of our drugs at Lilly. 17 Q. In what respect? 18 A. I was responsible for reporting 19 marketing activity on all drugs without regard to 20 which drug, basically any drug that was being 21 marketed. Prozac would have been included in 22 that information at some point. 23 Q. What type of information would 24 this have been? Page 76 1 A. Sampling activity. 2 Q. Sampling activity. You mean 3 samples that were given out to physicians? 4 A. Yes. 5 Q. For their use? 6 A. Yes. 7 Q. How long did you -- how long 8 were you in the marketing information systems? 9 A. I'm still in marketing 10 information systems. 11 Q. At any time while you were 12 working on Prozac, did you ever do any systems 13 analysis or data processing analysis or analysis 14 of any form concerning suicide in connection with 15 Prozac? 16 A. I don't remember any specific 17 reports or analyses relating to suicide. 18 Q. So is it correct that your 19 testimony is that at no time were you asked to 20 make a specific analysis or write a program to be 21 used in an analysis concerning the issue of 22 whether or not Prozac and suicide were in any way 23 related? 24 A. Could you read the question Page 77 1 back, please? 2 (THE COURT REPORTER READ BACK THE 3 REQUESTED TESTIMONY.) 4 A. No. 5 Q. That's not your testimony? 6 A. No. My testimony is that I do 7 not remember any specific reports. I may have 8 been asked to do that type of analysis, I do not 9 remember any specific reports. 10 MS. ZETTLER: Do you remember anything 11 generally? 12 A. There would have been reporting 13 at the end of clinical trials regarding efficacy 14 data which would have included suicidal thoughts 15 or tendencies. 16 Q. The efficacy data would have 17 included suicidal thoughts and tendencies? 18 A. Yes. 19 Q. What about safety data? 20 A. Yes. 21 Q. Did you have any work or did 22 you do any work on the safety data at the end of 23 the clinical trials? 24 A. Yes. Page 78 1 Q. What work did you do? 2 A. Reported instances of adverse 3 events and laboratory data that would have been 4 collected during the clinical trial. 5 Q. Now let's talk about adverse 6 events for a minute. Would those adverse events 7 have been adverse events that were reported by 8 the clinical investigators during clinical trials 9 as per clinical report forms? 10 A. Yes. 11 Q. Any other source of information 12 concerning adverse events of Fluoxetine? 13 A. No. 14 Q. The only information you've 15 ever analyzed or ever used was that information 16 reported during the clinical trials? 17 A. Yes. 18 Q. Are you familiar with the drug 19 epidemiology unit there at Lilly? 20 A. Yes. 21 Q. Did they ever report to you in 22 any fashion any adverse events or did you ever 23 secure their information? 24 MR. MYERS: With respect to Prozac? Page 79 1 A. With respect to Prozac? 2 Q. Yes. 3 A. I don't recall that we did 4 anything that would have been reported to the 5 DEU, would have been reported on clinical report 6 forms during that time period. 7 Q. What is the DEU? 8 A. Drug epidemiology unit, I'm 9 sorry. 10 Q. Was that unit collecting data 11 on adverse experiences that occurred during the 12 clinical trials? 13 A. I don't know the answer to 14 that, I didn't work with that group. 15 Q. Do you know what computer 16 system they used? 17 A. No, I do not. 18 Q. Would they have used the same 19 or a different computer system? 20 A. I do not know the answer to 21 that. 22 Q. Were there more than one 23 computer system there at Lilly that collected 24 information concerning adverse events? Page 80 1 A. Yes. 2 Q. All right. Identify for me the 3 various systems that collected information with 4 respect to adverse events in connection with 5 Prozac. 6 A. In connection with Prozac. 7 Early clinical trials were collected using an 8 Entrex data entry system. 9 Q. Give me that again? 10 A. Entrex, E-N-T-R-E-X. 11 Q. What did you call that, an 12 Entrex -- 13 A. Data entry system. 14 Q. All right. 15 A. Clinical trial data from the 16 clinical report forms was recorded on one of 17 several System 38 systems. There were -- there 18 was more than one data entry system on the System 19 38 for that. 20 Q. Would you give me those, 21 please, ma'am? 22 A. I do not not know what those 23 systems are, it's been several years. 24 Q. Give me any that you can Page 81 1 recall. 2 A. I do not remember the names of 3 them specifically. One was for collecting safety 4 data on patients that were not in efficacy 5 trials, one was -- 6 Q. Just a second. All right, go 7 ahead. 8 A. One was for a series of 9 clinical trials that was done during the time I 10 was there, and I believe there were several 11 systems written to collect data on new trials. I 12 was not involved in the writing of those systems. 13 Q. Were you involved in writing 14 any of them? 15 A. I was involved in writing 16 enhancements to those systems. 17 Q. All right. In what way? 18 A. In writing a new data entry 19 screen for additional data that was collected 20 would be an enhancement. 21 Q. What else? 22 A. That's the only thing I can 23 think of. 24 Q. All right. What other System Page 82 1 38 entry systems were used for the clinical trial 2 data, adverse experiences? 3 A. For Fluoxetine? 4 Q. Yes. 5 A. I believe there's a data entry 6 system for the spontaneous reports that come in 7 to the company, I do not know anything about that 8 system. That would be the system the drug 9 epidemiology unit would use. 10 MS. ZETTLER: Would that be the DEN, 11 system, D-E-N? 12 A. I think so, yes. 13 Q. But in all the information that 14 you collected and then reported, you never 15 obtained information within the DEN system or the 16 drug epidemiology unit? 17 A. No. 18 Q. Is that correct? 19 A. That's correct. 20 Q. That you didn't? 21 A. That I did not. 22 Q. What other systems were used to 23 collect adverse experiences while you were 24 working on Prozac? Page 83 1 A. Those are the only systems. 2 Q. The Entrex data entry system 3 was used in early clinical trials? 4 A. Yes. 5 Q. The System 38 was used to 6 obtain clinical trial data system, and there was 7 a lot of different entry systems? 8 A. Yes. 9 Q. Under the System 38? 10 A. Yes. 11 Q. The spontaneous reports were 12 collected in the DEN unit somehow, but you never 13 used that information? 14 A. I did not work with that 15 system. 16 Q. So you never obtained 17 information from that system since you didn't 18 work with it? 19 A. Right. 20 Q. Now what about the S-A-S main 21 frame information, was that ever used to collect 22 adverse experience data? 23 A. It was not used to collect 24 data. Page 84 1 Q. Was it used to store data? 2 A. Yes. 3 Q. Were there any other systems 4 used to store data? 5 A. Not while I was working on 6 Prozac. 7 Q. How about after you were 8 working on Prozac, are you familiar with any 9 other systems used to store data after you left 10 Prozac? 11 A. It's my understanding that DB2 12 has been used to store data. 13 Q. DB2? 14 A. Yes. 15 Q. What is DB2? 16 A. That is a data base management 17 system. 18 Q. Is it on the main frame? 19 A. Yes. 20 Q. When did that system come into 21 existence with respect to Prozac? 22 A. I do not know, I was not 23 working on Prozac. 24 Q. So it would have been after Page 85 1 1988? 2 A. Yes. 3 Q. How did you learn that that 4 system was being used for Prozac? 5 A. I have ongoing friendships with 6 people who have worked in that department. 7 Q. Who? 8 A. With Greg Cunningham, with 9 Janet Bosomworth, the statistician, with Nick 10 Miller, who is, I believe, working in that area 11 now. 12 Q. Any other computer systems or 13 entry systems of any other sources of entry or 14 storage of Fluoxetine adverse event information 15 within the Lilly organization that you're aware 16 of? 17 A. I don't know of any. 18 Q. And you've named all of those 19 that you're familiar with when you were working 20 on Prozac? 21 A. Yes. 22 Q. Now, give me the names of all 23 the systems used to analyze this data that's 24 stored in the computers? Page 86 1 A. That are currently used or were 2 used at that time? 3 Q. That were used at that time. 4 A. To my knowledge, SAS was the 5 only language that we used for reporting the 6 data, for the reporting -- SAS was the only 7 language that we used for reporting the data. 8 Q. How about for entering the 9 data, any other language for entering the data? 10 A. None other than what I've 11 already described on the System 38. 12 Q. Do you remember whether or not 13 there was any question or issue raised in this 14 response -- or was there ever any question in the 15 approvable letter from the Food and Drug 16 Administration in latter 1987 concerning the 17 issue of suicidality or homicidality? 18 A. I do not remember any specific 19 questions. 20 Q. Do you remember addressing at 21 all the issue of suicide or homicide or violent 22 activity in connection with that response that 23 you all made to that letter? 24 A. I do not remember that there Page 87 1 was any. 2 Q. There could be or couldn't be, 3 you just don't recall? 4 A. Right, I do not remember. 5 MS. ZETTLER: Would looking at that 6 letter refresh your recollection as to whether or 7 not you worked specifically on any of those 8 issues? 9 THE WITNESS: I couldn't say without 10 knowing what's in the letter. 11 Q. You would have to see the 12 letter, wouldn't you? 13 A. I don't know the answer to 14 that. 15 Q. Without seeing the letter? 16 A. Right. 17 Q. If you saw the letter, you 18 would know the answer, wouldn't you, probably? 19 A. I don't know that, I did not -- 20 I did not decide how the responses would be done, 21 so I don't recall exactly what responses would 22 have been written to that letter. 23 Q. I understand that, but you 24 would have to see the original approvable letter Page 88 1 and the responses in order to know that, wouldn't 2 you? 3 A. Or all of the information that 4 was sent in the response, I would know that. 5 Q. Correct. Did you put the 6 response letter in your file also? 7 A. No. 8 Q. Who would have that? 9 A. I do not know. 10 Q. How do you know you didn't put 11 the response, because you never got a copy of it? 12 A. I never received a copy of it 13 and it was several boxes of paper. 14 Q. Were you ever given a specific 15 task of seeking or reporting information 16 contained within Lilly's computer system, 17 information pertaining to adverse events in 18 connection with Prozac or Fluoxetine? 19 A. Yes. 20 Q. What specific work did you do 21 in connection with that? 22 A. I was requested to report all 23 adverse events that had occurred on clinical 24 trials using Fluoxetine. Page 89 1 Q. Who made this request? 2 A. All requests would have come 3 from the medical community, the physician or CRA 4 in the medical community. 5 Q. Who within the medical 6 community? 7 A. That would have been Earleen 8 Ashbrook or Joe Wernicke. 9 Q. When did you get this request? 10 A. I received numerous requests 11 during the time I was working on reporting the 12 data in supporting Prozac. 13 Q. When was the first time you got 14 that request? 15 A. In either late 1984 or early 16 1985, when I first started working with the data. 17 Q. Would this have come to you in 18 the form of a memo or in some written form or 19 would this have been an oral request? 20 A. Either case would have been 21 used. 22 Q. How did you go about getting 23 that information, what did you have to do? 24 A. Could you rephrase the Page 90 1 question? 2 Q. As I understand it, you were 3 requested by the medical community to report all 4 adverse affects -- adverse events that had 5 occurred during the clinical trials using 6 Fluoxetine. My question is how did you go about 7 getting that information? 8 A. By writing SAS programs to 9 retrieve the data from the data bases. 10 Q. Wasn't there something in place 11 at that time to retrieve adverse event 12 information? 13 A. Yes. 14 Q. What was in place at that time? 15 A. There were several SAS programs 16 in place. 17 Q. Why did you have to write 18 another one then? 19 A. The requests were not always 20 exactly the same. There were different requests 21 over the course of several years. 22 Q. Give me an example of how there 23 would be two different requests. 24 A. There would be a request for Page 91 1 reporting adverse events at the end of a study, 2 what adverse events occurred during that study, 3 in total. There would be requests to report on 4 an investigator by investigator basis so the 5 information could be sent back to the 6 investigator regarding their own patients. Those 7 would be different forms of the same report, but 8 would require slightly different programming to 9 get the data out in two different reports. 10 Q. Is the program and a query the 11 same thing? 12 A. A query, as I mentioned it 13 before, was usually a request on the System 38. 14 It took the form of a written program using the 15 programming language. 16 Q. And you would have to make this 17 request in S-A-S language? 18 A. On the main frame, it would be 19 in SAS. 20 Q. What language would you use for 21 the System 38? 22 A. There was a query language on 23 the System 38. 24 Q. What was the name of that query Page 92 1 language on the System 38? 2 A. I do not recall the name of 3 that, language name. 4 Q. Was it some commercial program? 5 A. Yes, provided by IBM. 6 Q. Have you ever heard of the 7 ELECT terminology? 8 A. Yes. 9 Q. What is that? 10 A. ELECT was a dictionary for 11 reporting adverse events. 12 Q. Was that used on the System 38 13 or the S-A-S? 14 A. It was used on both. 15 Q. For the entire time that you 16 were there? 17 A. No. 18 Q. Why? 19 A. Initial clinical trials on 20 Prozac were done using another event dictionary. 21 Q. What event dictionary was that? 22 A. It was -- the acronym for it 23 was S-S-A-I. 24 Q. S-S -- Page 93 1 A. A-I. 2 Q. So Lilly went from SSAI to 3 ELECT? 4 A. Yes. 5 Q. And these are two different 6 dictionaries? 7 A. Yes. 8 Q. And dictionaries of what? 9 A. Of adverse event terms. 10 Q. When did Lilly move from SSAI 11 to ELECT? 12 A. It was sometime -- is this a 13 question regarding Prozac or regarding the 14 medical community? 15 Q. Prozac. 16 A. Prozac. It was sometime during 17 1985. 18 Q. How do you know that, because 19 you worked on it? 20 A. Because I worked on it. 21 Q. What did you do? 22 A. I wrote a program that would 23 report any data that was collected using SSAI 24 would report those as ELECT terms. Page 94 1 Q. All right. What does SSAI 2 stand for? 3 A. I think it stands for Signs, 4 Symptoms and Illnesses. 5 Q. Okay. 6 A. I believe. 7 Q. Sounds good to me. So 8 originally the Prozac clinical trials used terms 9 of adverse events as per a dictionary called 10 SSAI? 11 A. Yes. 12 Q. And that was to ensure 13 consistency with respect to particular 14 terminology with respect to a particular adverse 15 event; is that right? 16 A. Yes. 17 Q. Then in 1985, the ELECT system 18 was put in place and it used a different 19 dictionary for adverse event terms? 20 A. That's two different questions. 21 Q. Yes. 22 A. The ELECT dictionary was put in 23 place before 1985, I do not know the date. 24 Q. But it was put in with Prozac Page 95 1 in 1985? 2 A. It was used on Prozac starting 3 in 1985, it was used on other drugs within the 4 medical community before 1985. 5 Q. Why did they put Prozac on 6 ELECT in 1985? 7 A. I believe it was to standardize 8 the terminology across all of medical. 9 Q. Well, when did the ELECT 10 dictionary originally come into existence at 11 Lilly? 12 A. I do not know. 13 Q. As far as you know? 14 A. I do not know, I do not know 15 the answer to that question. 16 Q. Was ELECT being used for other 17 compounds, used by Lilly that you know of, before 18 you came to work for Lilly? 19 A. I do not know the answer to 20 that. 21 Q. But you know when you came to 22 work Fluoxetine was used in the SSAI dictionary? 23 A. Yes. 24 Q. Was Tomoxetine used in the SSAI Page 96 1 dictionary when you came? 2 A. I do not remember that, I do 3 not remember the answer to that. 4 Q. When you wrote a program to 5 change SSAI terminology over to ELECT 6 terminology, did that have to do with Prozac or 7 did that have to do with Tomoxetine? 8 A. Neither, it had to do with the 9 entire medical community. It was a program that 10 was used on Prozac as well other drugs that had 11 been reporting in SSAI. 12 Q. Help me again. As I understand 13 it, you wrote a program that converted SSAI 14 terminology to ELECT terminology? 15 A. Yes. 16 Q. Give me an example of a term 17 that would be different with SSAI and ELECT, if 18 you can remember a specific term. 19 A. I do not remember any specific 20 terms. 21 Q. Well, would it be a system like -- 22 I mean, and I know this probably wasn't it, but 23 would you be changing an adverse event, say, like 24 sweating to maybe perspiration? Page 97 1 A. That would be a possibility. 2 Q. Bad breath to halitosis? 3 A. That would be a possibility. 4 Q. Different words for the same 5 condition? 6 A. Yes. 7 Q. Why would you want to change 8 that as a systems analyst? 9 A. It was at the request of the 10 medical community. 11 Q. Why would the medical community 12 want to change that? 13 A. To report all adverse events 14 using the same terminology. 15 Q. Was SSAI a dictionary that was 16 used throughout the medical community? 17 A. It was used on more than 18 Prozac. 19 Q. When I say throughout the 20 medical community, wax SSAI something that was 21 recognized as an adverse event term dictionary 22 with, say, another pharmaceutical company? 23 A. I don't know the answer to 24 that. Page 98 1 Q. ELECT is a particular Lilly 2 dictionary, is it not? 3 A. It was my understanding that 4 ELECT was based on the FDA dictionary, adverse 5 event dictionary. 6 Q. The first two letters in ELECT 7 stand for Eli Lilly, don't they? 8 A. I believe so. 9 Q. So it's something particularly 10 that was designed for use within the Lilly 11 organization, wasn't it? 12 A. I do not know, I did not design 13 the dictionary. 14 Q. Your understanding is that the 15 ELECT dictionary is based on the FDA dictionary? 16 A. Yes. 17 Q. Why didn't you just use the FDA 18 dictionary? 19 A. I did not make that decision, I 20 do not know what went into that decision. 21 Q. Do you have any idea to whom I 22 should be directing those questions? 23 A. No, I do not. 24 Q. Do you know the name of the FDA Page 99 1 dictionary, does it have a particular term? 2 A. As far as I can remember, it 3 was COSTART. 4 Q. COSTART? 5 A. Yes. 6 Q. And that's another dictionary? 7 A. That's the name for the FDA 8 dictionary. 9 Q. Is it your testimony that -- 10 it's your understanding that every word in the 11 COSTART dictionary is going to be the same as in 12 the ELECT dictionary? 13 A. I do not know the answer to 14 that. 15 Q. Do you know what differences 16 there were in the SSAI dictionary as opposed to 17 the ELECT dictionary? 18 A. I do not know the answer to 19 that. 20 Q. There were differences, though, 21 weren't there? 22 A. Yes. 23 Q. Because you had to write a 24 program to convert over to the ELECT terminology. Page 100 1 A. Yes. 2 Q. When you were writing the 3 program to convert over to the ELECT terminology, 4 did you ever consult the COSTART dictionary to 5 determine what term was being used by COSTART? 6 A. I did not work with the COSTART 7 dictionary. 8 Q. Any other dictionaries of 9 adverse events that you're familiar with? 10 A. No. 11 Q. Do you know if Lilly ever used 12 any other dictionary of adverse events other than 13 SSAI, ELECT -- or ELECT? 14 A. I do not know. 15 Q. Could an individual such as me 16 send and get a SSAI dictionary and an ELECT 17 dictionary and compare the two dictionaries? 18 MR. MYERS: Is that question limited to 19 a person such as you? 20 A. I guess I need to understand 21 what the frame of reference is. Is the frame of 22 reference a person who has no computer access? 23 Q. Yes. 24 A. I don't know the answer to Page 101 1 that, I don't know where those dictionaries are 2 stored at this point. 3 Q. Are these computer-type 4 dictionaries? 5 A. Yes. 6 Q. They're not like Webster's 7 dictionary sitting on the shelf? 8 A. In what way? 9 Q. Could you print out your 10 computer to say print ELECT dictionary of terms? 11 A. Yes. 12 Q. And you could run a printout of 13 all the adverse experience terminology used by 14 ELECT? 15 A. Yes. 16 Q. Did you write a program that 17 would automatically convert over SSAI terminology 18 that had been used in the past into ELECT 19 terminology to get that information stored 20 together? 21 A. No, it was not stored together. 22 Q. How did that work? 23 A. Any data that was in the data 24 base, the data bases that were out there at the Page 102 1 time, that were in SSAI would -- a report for 2 those adverse events would convert to the ELECT 3 term, and then report the terms out in ELECT, or 4 report the ELECT terms. 5 Q. Right. So if sweating had been 6 the SSAI term, and perspiration was the ELECT 7 term, if you entered sweating, it would 8 automatically come out as perspiration as far as 9 picking out adverse events having to do with 10 sweating or perspiration? 11 A. If those were the mappings that 12 were in the dictionary, yes. 13 Q. They would automatically map 14 from one to the other? 15 A. Yes. 16 (PLAINTIFFS' EXHIBIT NO. 1 WAS 17 MARKED FOR IDENTIFICATION AND 18 RECEIVED IN EVIDENCE.) 19 Q. Ms. Sheldon, we've handed you a 20 document that's been marked as Exhibit 1, and you 21 were listed on the CC portion of the top page. 22 MR. MYERS: Could she have a chance to 23 review it before you ask her questions about it? 24 Q. Yes. Would you please review Page 103 1 what's been marked Exhibit 1. 2 A. (WITNESS COMPLIES.) 3 Q. This appears to be a memo from 4 a Ms. Ashbrook to a group of individuals. Those 5 individuals are listed on the top of the first 6 page of this document. Generally, who are these 7 individuals? 8 A. In general, it appears that the 9 primary addressee list is people from the medical 10 community. 11 Q. When you say community, you're 12 not talking about community -- 13 MR. MYERS: At large. 14 A. I'm talking about the Lilly -- 15 Q. The Lilly -- 16 A. The Lilly medical community. 17 Q. All right. 18 A. The CC list is a combination, 19 there are systems analysts, a statistician, a 20 physician and I believe one of them is a 21 secretary. 22 Q. And you are the systems 23 analyst? 24 A. I'm one of the systems Page 104 1 analysts. 2 Q. The title is time cards and 3 project timesheets. Did you keep a record of 4 your time that you spent on particular projects? 5 A. Yes. 6 Q. Item one there says project 7 codes will be expanded up to ten characters long 8 will allow for more detailed tracking of how we 9 spend our time. See attached sheet for new 10 project codes; correct? 11 A. Yes. 12 Q. Does that mean that during your 13 day you would record how long you spent doing a 14 particular task? 15 A. Yes. 16 Q. You know lawyers -- or some 17 lawyers, that is, make records of their billable 18 time. You understand that? 19 A. Yes. 20 Q. And that's recorded and then 21 later transmitted to the client for billing 22 purposes. Why were you all keeping specific time 23 logs on the time you were spending on particular 24 projects? Page 105 1 A. The best I can remember, it was 2 for accounting for financial purposes, and for 3 future project planning, how much time was 4 necessary for each project. 5 Q. But could someone go back and 6 pull up your project code time card information 7 and find out what you were doing on a particular 8 project at a particular time? 9 A. I don't know the answer to 10 that. 11 Q. Is that, in theory, what the 12 purpose of time cards is? 13 A. In theory. 14 Q. Did you do that religiously, 15 keep your time cards and time sheets? 16 A. Religiously, what do you mean 17 by religiously? 18 Q. Where you accounted for most of 19 your day. 20 A. Yes. 21 Q. And is that entered into the 22 computer by you or did you just turn in your time 23 cards at the end of each day? 24 A. I don't remember for sure, I Page 106 1 think I just turned in a sheet on a periodic 2 basis. 3 Q. Then if you'll look on page two 4 of this document, it gives particular project 5 codes for particular projects; correct? 6 A. Yes. 7 Q. And it lists protocols, 8 Fluoxetine protocols, and then lists several 9 different studies that were in progress at that 10 time; correct? 11 A. I assume that's what this list 12 was. 13 Q. Do you recall getting documents 14 such as this and coding your time sheets as per 15 these particular -- for instance, Fluoxetine 16 protocols? 17 A. I don't remember the specific 18 activity. I'm not sure I know what you're 19 looking for in the question. 20 Q. Well, I'm just wondering if you 21 did in fact key your time to these particular 22 project codes that are mentioned here on page two 23 of this document. 24 A. That's something I don't Page 107 1 remember doing, I can't -- I assume I did it, I 2 don't recall doing it. 3 Q. For example, look down at the 4 last Fluoxetine protocol there on that page, it's 5 HCCW, psychomotor study, paren, seventy-two, 6 close paren. Do you remember if you were working 7 on a particular systems analyst analysis for that 8 study coding in that particular project code? 9 A. I don't remember doing that 10 activity. 11 Q. I know that, I'm just using 12 that as an example. But if you were doing a 13 particular activity, would you do that? 14 A. If that were the standard 15 procedure for the department, yes, I would have 16 done it. 17 Q. Page three, Ms. Ashbrook says 18 that she has a complete dictionary of project 19 codes in her office. Have you ever seen a 20 dictionary of project codes? 21 A. No. 22 Q. Were you assigned projects by 23 codes? 24 MR. MYERS: When you say was she, you Page 108 1 mean in addition to the project having a code? I 2 don't follow what you're saying. 3 Q. Yes. 4 A. No, my assignment -- I guess 5 I'm not sure how to answer this. The specific 6 work I did would have related to, at times, these 7 specific project codes, and at times to the 8 aggregation of all Fluoxetine projects. 9 Q. All right. 10 (PLAINTIFFS' EXHIBIT NO. 2 WAS 11 MARKED FOR IDENTIFICATION AND 12 RECEIVED IN EVIDENCE.) 13 Q. Would you review Exhibit 2. 14 A. (THE WITNESS COMPLIES.) 15 Q. Why don't you just review it. 16 If you feel comfortable just reviewing it 17 generally, and you don't have to read it word for 18 word because I'm not going to read it word for 19 word, Ms. Sheldon. 20 A. Okay. 21 Q. This document is apparently an 22 audit done by medical quality assurance; is that 23 correct? 24 A. Yes. Page 109 1 Q. And it's an audit of the 2 Fluoxetine safety update that was to be submitted 3 to the Food and Drug Administration, is it not? 4 A. Yes. 5 Q. And Lilly had to, with respect 6 to Prozac, submit from time to time safety 7 updates to the Food and Drug Administration? 8 A. Yes. 9 Q. Medical quality assurance, is 10 that a separate division within the Lilly 11 organization? 12 A. Yes. 13 Q. And did you have much 14 experience with the medical quality assurance 15 department? 16 A. In what time frame? 17 Q. During the time frame that you 18 were working on the Prozac clinical trials. 19 A. Some experience working with 20 them, not extensive. 21 Q. Beg your pardon? 22 A. It was not extensive. 23 Q. Is medical quality assurance 24 sort of a watchdog group within the medical Page 110 1 department? 2 MR. MYERS: I object to the form. What 3 do you mean by watchdog group? 4 MR. SMITH: Somebody that -- or some 5 group of individuals that oversee or -- 6 Q. Did it help to ensure accuracy 7 and quality in the work done within the medical 8 department? 9 A. Yes. 10 Q. How often did you have dealings 11 with the medical quality assurance group while 12 you were working with Fluoxetine? 13 A. I do not remember exactly. I 14 know specifically that we worked together during 15 the safety update that this document refers to. 16 Q. Did you work on any other 17 safety updates other than this June, 1986 safety 18 update? 19 A. No. 20 Q. That's the only one you've 21 worked on? 22 A. Yes. 23 Q. This table forty-one was a 24 listing of patients who discontinued from the Page 111 1 Fluoxetine study after submission of the NDA, 2 either because of death or adverse events. 3 MR. MYERS: You're reading from that 4 first paragraph? 5 MR. SMITH: Yes. 6 Q. Did you have anything to do 7 with the preparation of table forty-one? 8 A. I don't remember specifically 9 which reports I worked on and which I didn't. 10 Q. Do you remember working on a 11 table that was a listing of patients who 12 discontinued the Fluoxetine study because of 13 death or adverse event? 14 A. I don't remember what specific 15 reports I worked on for which safety update. 16 Q. Did some of those reports have 17 to do with discontinuation of patients within the 18 clinical studies? 19 A. I don't know the answer to that 20 because I don't know the specific reports that I 21 worked on for the safety update. 22 Q. Did you work on some reports 23 that became part of the safety update? 24 A. Yes. Page 112 1 Q. Did any of those reports have 2 to do with safety? 3 A. Yes. 4 Q. Did any of those reports have 5 to do with early discontinuation? 6 A. I don't know the answer to 7 that. 8 Q. Why? 9 A. Because I do not remember the 10 specific reports that I was involved in that were 11 included in the safety update. 12 Q. Why do you not remember that? 13 A. Because it was nine years ago, 14 eight years ago. 15 Q. Did you make a request or have 16 you made a request to see some of those reports? 17 A. No, I have not. 18 Q. Were you told that you were 19 going to be shown some documents here today? 20 MR. MYERS: Well, to the extent your 21 question has to do with any discussion she may 22 have had with me or Ms. Huff, she's not going to 23 answer that, Paul, about anything she's discussed 24 with us. Page 113 1 Q. Did you know that the file that 2 you had delivered to the legal department at 3 Lilly would or had been produced to the 4 plaintiffs in this case? 5 A. I did not know. 6 Q. Turn to page two of that audit. 7 See there where it says error rate in the middle 8 of the page? 9 A. Yes. 10 Q. It says there were three 11 hundred seventeen patients included in table 12 forty-one, there were forty-eight patients who 13 had discrepancies between table forty-one and the 14 termination summary. The error rate was fifteen 15 point one percent. Do you see that? 16 A. Yes. 17 Q. Now, I believe it's your 18 testimony here today that you don't know whether 19 you were involved in table forty-one or not? 20 A. That is correct. 21 Q. What would you have to do in 22 order to know whether or not you were involved in 23 table forty-one? 24 A. I would have to see the actual Page 114 1 program and output referred to. 2 Q. Then do you think that would 3 for sure give you the ability to know whether you 4 worked on table forty-one or anything specific in 5 connection with table forty-one? 6 A. Only if it were documented in 7 the program that I had worked on. 8 MS. ZETTLER: Do you keep copies of the 9 output in your files? 10 THE WITNESS: No, I do not. 11 MS. ZETTLER: What do you do with it? 12 THE WITNESS: The output is turned over 13 to the requester, which would be the physician or 14 the CRA. 15 MS. ZETTLER: How many other people in 16 systems would have worked on this project back in 17 1986 when you worked on it? 18 MR. MYERS: The audit or the safety 19 update? 20 MS. ZETTLER: Start with the audit. 21 THE WITNESS: The audit of the safety 22 update, I do not recall the number of systems 23 people working on it. 24 MS. ZETTLER: Would it have been more Page 115 1 than three? 2 THE WITNESS: I do not know. 3 Q. How about the safety update 4 itself? 5 A. Safety update itself -- 6 Q. I'm a ventriloquist, that was 7 actually me talking all that time. You didn't 8 know I could do that, did you? 9 MR. CLEMENTI: Your voice changed. 10 MR. MYERS: What is the question, how 11 many on the safety update? 12 A. Yes. During the course of the 13 safety update, there were probably three or four 14 different systems people who work on various 15 parts of it. 16 Q. Do you have any reason to 17 disagree with the medical quality assurance 18 conclusion that there was an error rate of 19 fifteen point one percent in the table forty-one? 20 A. In what copy of the table 21 forty-one? 22 Q. Well, the paragraph there says 23 there were three hundred and seventeen patients 24 included in table forty-one, there were Page 116 1 forty-eight patients who had a discrepancy 2 between table forty-one and the termination 3 summary. The error rate was fifteen point one 4 percent. 5 A. Can I confer? 6 (WITNESS CONFERS WITH COUNSEL.) 7 A. I agree that there was a 8 fifteen point one percent error rate on the test 9 copy that was reviewed by medical quality 10 assurance. 11 Q. What do you mean by test copy? 12 A. An initial run of the program 13 that was being reviewed during writing and 14 testing of the programs that were creating the 15 report. 16 Q. Is it your testimony that you 17 now remember that that program was changed and 18 corrected after this review? 19 A. It is my testimony that this 20 specific quality assurance report states that all 21 problems identified during the audit were 22 corrected before the report was sent to the FDA. 23 Q. Do you know that for a fact? 24 A. I'm testifying to what I'm Page 117 1 reading on this report. 2 Q. No, do you know that for a 3 fact? 4 A. I do not recall the specifics 5 of this program because I do not recall whether I 6 was specifically involved. 7 Q. So you don't know for a fact 8 that it was corrected, do you? 9 MR. MYERS: She's already answered the 10 basis on which she bases that answer that she 11 read it in the report, in this report that's in 12 front of her. 13 Q. You also read that there was an 14 error rate of fifteen point one percent, didn't 15 you? 16 A. That I did. 17 Q. Do you recall any other tables 18 that had an error rate of fifteen point one 19 percent in that safety update? 20 A. I do not recall. 21 Q. Do you disagree that there were 22 other tables submitted that had an error rate 23 greater than fifteen point one percent? 24 MR. MYERS: Submitted to who? Page 118 1 MR. SMITH: Submitted to medical 2 quality assurance. 3 MR. MYERS: As opposed to the FDA? 4 MR. SMITH: Yes. 5 A. I don't know the answer to 6 that. 7 Q. In fact there were, weren't 8 there? 9 MR. MYERS: Before you answer, Paul you 10 asked her to briefly look at this document, and 11 now you're starting to ask her to specifically 12 look at it. And if you are, that's fine, but I 13 think you ought to give her a chance to read it 14 in full if there's a specific portion of this 15 audit you would like her to look at. 16 Q. My question is not concerning 17 table forty-one, but other tables that were 18 submitted to become a part of the safety update 19 to quality assurance. There were other tables 20 that had an error rate of greater than fifteen 21 point one percent, weren't there, Ms. Sheldon? 22 MR. MYERS: If you know, tell him, and 23 if you don't, tell him that. 24 A. I don't know. Page 119 1 Q. Why don't you take a look at 2 this document and see if there's any basis to 3 improve your knowledge on that. 4 MR. MYERS: And by saying that, your 5 question being whether there were other tables 6 with an error rate identified by medical quality 7 assurance greater than the one for table 8 forty-one that you just talked about. 9 MR. SMITH: Yes. 10 A. Would you repeat your question, 11 please? 12 (THE COURT REPORTER READ BACK THE 13 REQUESTED TESTIMONY.) 14 A. I don't find anything in this 15 report about error rates on other tables other 16 than the table forty-one that we've already 17 discussed. 18 Q. So there's nothing there to 19 dispute the fact that there were other tables 20 that were reviewed where there was an error rate 21 greater than fifteen point one percent? 22 MR. MYERS: I object to the form, Paul, 23 you mischaracterized what she said, she said the 24 exact opposite, I think, if you'll listen to the Page 120 1 answer. 2 MR. SMITH: I did listen to the answer. 3 Q. My question is different. Do 4 you recall -- let me just ask it directly. Don't 5 you recall, Ms. Sheldon, that there were other 6 tables submitted to medical quality assurance or 7 that medical quality assurance reviewed other 8 tables, and some of those tables had error rates 9 greater than fifteen point one percent? 10 A. I do not remember anything like 11 that. 12 Q. Do you deny that that occurred? 13 A. I can't deny or confirm it 14 because I don't remember. 15 Q. Do you remember even that other 16 tables were submitted and reviewed by medical 17 quality assurance? 18 A. Other than the two tables that 19 were listed in this report, I do not know what 20 tables medical quality assurance reviewed and 21 what they did not review. 22 Q. Don't you know that there were 23 other tables that were reviewed that aren't 24 mentioned in this letter that had an error rate Page 121 1 of greater than fifteen point one percent? 2 MR. MYERS: She's answered that 3 already. 4 A. I don't know that. 5 Q. You just don't remember that? 6 A. I do not remember, yes. 7 Q. Do you consider an error rate 8 of fifteen point one percent as being acceptable? 9 A. I do not consider it acceptable 10 on the final form of a report. 11 Q. Do you consider it acceptable 12 in any form of a report? 13 A. I would -- my personal opinion 14 is that an initial printout that comes from a new 15 program may include an error rate, and that is 16 why we test and review the output from programs 17 to find those error rates and correct them before 18 they're submitted to the FDA. 19 Q. Well, is your answer to the 20 question, if it's a new program, you'll accept a 21 fifteen point one percent error rate? 22 A. No. 23 Q. Is the answer to a question 24 then that you will accept -- you will not accept Page 122 1 a fifteen point one error rate in a new program? 2 A. I would not accept fifteen 3 point one percent error rate in a new program. 4 Q. Would you accept it in any 5 program? 6 A. No. 7 Q. Do you have any explanation as 8 to why this table contained a fifteen point one 9 percent error rate? 10 A. Based on the writeup that is in 11 this report, I'm of the understanding that this 12 review was done on an initial copy of that report 13 prior to all testing and corrections being made. 14 Q. Was it the purpose of medical 15 quality assurance to review initial copies? 16 A. I do not know the answer to 17 that, I don't know what medical quality 18 assurance's purpose was. 19 Q. The audit itself doesn't say 20 that this was done on a draft copy of the safety 21 update to be submitted to the FDA, does it? 22 A. No. 23 Q. It was intended to be a final 24 copy if it passed medical audit, wasn't it? Page 123 1 A. I don't know that. 2 Q. As far as you know? 3 A. I don't know that. 4 Q. Okay. Turn to page four. Let 5 me come over, if you don't mind, Ms. Sheldon, and 6 point to you where I'm going to be reading 7 because there's so much stuff blacked out that 8 it's hard to give you a specific pointing spot. 9 It's right here that I'm going to start, all 10 right. It says table forty-on lists depression 11 as self-inflicted injury as adverse event causing 12 termination. Do you see that? 13 A. Yes. 14 Q. It says CRF case report form 15 lists both suicidal thoughts and self-inflicted 16 injury. Do you have any explanation for why 17 table forty-one that would be submitted as a 18 safety update to the Food and Drug Administration 19 would list depression and self-inflicted injury 20 when the case report form lists suicidal thoughts 21 and self-inflicted injury? 22 A. I would take it that that is 23 because of the next sentence in this document, 24 which says suicidal thoughts mapped to depression Page 124 1 in the SSAI dictionary. 2 Q. Do you know why suicidal 3 thoughts mapped to depression in the SSAI 4 dictionary? 5 A. No, I do not. 6 Q. Does that maybe suggest to you 7 that suicidal thoughts would map to depression in 8 the SSAI dictionary? 9 A. I'm not a physician, I can't 10 comment on that. 11 Q. Do you know whether or not 12 suicidal thoughts mapped to depression in the 13 ELECT dictionary? 14 A. No, I do not know the answer. 15 Q. Do you remember while you were 16 a systems analyst working on Prozac it ever 17 coming up whether or not to categorize suicidal 18 thoughts as depression or as suicidal thoughts? 19 A. I do not remember any 20 discussions related to that. 21 Q. You didn't have anything to do 22 with the summary itself, did you? 23 MR. MYERS: What summary? 24 Q. Well, it says summary states Page 125 1 patient was discontinued due to investigator's 2 decision that the patient had become unreliable 3 and non-compliant. You didn't do the summary 4 yourself, did you? 5 A. I don't know what the summary 6 refers to. 7 Q. Do you know M.K. Mattlar? 8 A. Yes, I do. 9 Q. Who is M.K. Mattlar? 10 A. Currently her name has changed, 11 it's Margaret Kunkel, K-U-N-K-E-L. 12 Q. So the M stands for Margaret? 13 A. Yes. 14 Q. What is her position with Lilly 15 at this time? 16 A. I don't know. 17 Q. Do you know if she's still with 18 Lilly? 19 A. Yes, I do. 20 Q. That same page, the last page 21 says, would Dr. Wernicke, Dr. Sampson and Ms. 22 Ashbrook, who was your supervisor at the time, 23 wasn't she? 24 A. No, she was not. Page 126 1 Q. Please explain in writing 2 within thirty days what procedures would be 3 implemented to prevent recurrence of the problems 4 identified during this audit. Do you see that? 5 A. Yes. 6 Q. Did you ever do anything to put 7 in place a procedure that would prevent 8 recurrence of the problems identified during that 9 audit? 10 A. I do not remember any 11 procedures that would have come out of this 12 audit. 13 Q. You don't remember anything you 14 did to prevent recurrence of the problems 15 identified by the audit? 16 A. That's correct. 17 MR. MYERS: It's 12:15, I would like to 18 break for lunch if that's all right. 19 MR. SMITH: Sure. 20 (A SHORT RECESS WAS TAKEN.) 21 (PLAINTIFFS' EXHIBIT NO. 3 WAS 22 MARKED FOR IDENTIFICATION AND 23 RECEIVED IN EVIDENCE.) 24 Q. Ms. Sheldon, would you look Page 127 1 over the document that's been marked as document 2 Number 3, and review that for a minute so we can 3 talk about that. 4 A. Okay. 5 Q. Exhibit Number 3 appears to be 6 some type of computer generated exchange within 7 Lilly; is that correct? 8 A. It's an electronic mail 9 message. 10 Q. It's E-Mail? 11 A. Yes. 12 Q. The typeset on this E-Mail is 13 different from other Lilly E-Mail that we've 14 seen. Is it different from other E-Mail that 15 you've seen? 16 A. It depends on the printer that 17 the E-Mail is printed on. 18 Q. This particular printer was 19 different than some others? 20 A. Yes. 21 Q. The first page is dated May 13, 22 1987, the second page is not dated, the third 23 page is May 14th, the fourth page is May 15th. 24 There are numbered -- the pages are numbered with Page 128 1 consecutive Pz numbers, but I'm not sure that 2 this is all of the E-Mail on this subject. Can 3 you tell from looking at it whether or not this 4 is all of the E-Mail on this subject? 5 A. I can't tell from looking at 6 it. 7 Q. The reason that I ask is the 8 first page there, starting at on May 13th, 1987, 9 is twelve nineteen. Then under that is May 13th, 10 1987, eight forty-nine. Would that mean that the 11 bottom chronologically should be on the top? 12 A. That is a convention in the way 13 our electronic mail system attaches forwarding on 14 a message. 15 Q. All right. 16 A. The top -- what you see dated 17 May 13th, twelve nineteen is a message that was 18 attached in forwarding the original message, 19 that's what that means. 20 Q. Is the message, May 13th, 1987, 21 timed at eight forty-nine, the original message, 22 as far as you can tell? 23 A. As far as this series of 24 messages is concerned, yes. Page 129 1 Q. It's identifying a problem that 2 has apparently occurred in connection with 3 almedica randomization; is that right? 4 A. Yes. 5 Q. What is almedica? 6 A. Almedica was a software package 7 Lilly purchased to manage randomization of 8 patients to clinical trial drugs. 9 Q. All right. In other words, the 10 patients were going to be randomized pursuant to 11 a software package? 12 A. I'm not sure I understand the 13 question. 14 Q. Well, was this software package 15 assigning particular patients to get particular 16 medications? 17 A. It was assigning a kit number 18 to a particular medication, and when a patient 19 started on the study, the physician would give a 20 kit number to a patient, and this was our way of 21 determining what drug the patient was on. 22 Q. So Almedica is just the name of 23 a software program? 24 A. It's a complete software Page 130 1 package that Lilly purchased. 2 Q. What is randomization in 3 connection with these messages? 4 A. A randomization in a clinical 5 trial study means that the patient will get one 6 of the study drugs, and the randomization is how 7 it's determined which patient gets which drug. 8 Q. Well, is the theory that the 9 patient is supposed to be picked at random? 10 A. Yes. Or the drug is supposed 11 to be picked at random for the patient. 12 Q. The drug. Well, don't you want 13 one patient to receive a consistent drug 14 throughout the clinical trial, be it Fluoxetine, 15 placebo or a comparitor? 16 A. That depends on the design of 17 the protocol for that study. 18 Q. Well, if the protocol requires 19 a double-blind study, in which the Fluoxetine is 20 being compared with placebo in a comparitor drug, 21 some other anti-depressant, isn't the purpose to 22 select patients at random to receive these drugs? 23 A. If that's the way the protocol 24 was written, yes. Page 131 1 Q. So what part does the computer 2 software program have to do with it? 3 A. The computer software program 4 creates -- generates a list of numbers that are 5 kit numbers, and the drug that goes into that kit 6 number, those kits are sent to the physician. 7 When a physician starts a patient on drug, they 8 give a kit to the patient, and the drug that is 9 in that kit is then the drug that the patient is 10 on during the course of the study. 11 Q. And does the kit contain enough 12 drugs for the entire study? 13 A. That, again, depends on the 14 protocol, how the protocol is written. 15 Q. What generally happens in that 16 regard? 17 A. In some protocols, that is the 18 case, in some protocols additional kits are sent. 19 Q. Depending on the length of the 20 study perhaps? 21 A. The length and the design of 22 the study. 23 Q. But it's important, is it not, 24 to have patients properly randomized? Page 132 1 A. Yes. 2 (DISCUSSION OFF THE RECORD.) 3 Q. Okay, what was the problem, Ms. 4 Sheldon? 5 A. As far as I can remember, a 6 particular kit number was used on one study. For 7 example, in the example here, it says one, two, 8 three, four. In another study, the same number 9 was reused, and when those studies were combined 10 for reporting, possibly for safety reporting, 11 there was confusion over whether the one, two, 12 three, four went with the patient on the first 13 study or the patient on the second study, because 14 the Almedica table did not contain any indication 15 which study -- internally, the table did not 16 contain any indication which study the 17 randomization was for. 18 Q. How long had this problem 19 existed? 20 A. I don't know, I have no idea 21 based on the information I have here. 22 Q. Do you recall this as being a 23 problem, I mean do you have an independent 24 recollection of this problem? Page 133 1 A. I do not. Other than what's in 2 the messages, I don't remember exactly what the 3 problem was or how it surfaced. 4 Q. But now that you see the 5 message, you remember that there was at one time 6 a problem with the Almedica kit? 7 A. Yes, I remember that the 8 problem occurred. 9 Q. Do you remember this as being a 10 serious problem? 11 A. I don't recall. 12 Q. The reason I ask is because on 13 page four, the last page, it appears that you say 14 thanks for soliciting input on this very, and 15 very is in caps, critical matter. Is that a 16 message transmitted by you? 17 A. That was a message I wrote. 18 Q. All right. Then right before 19 that, it states this problem could never happen 20 again, and, let's see, there's one, two, three, 21 four, five explanation marks behind that; 22 correct? 23 A. Yes. 24 Q. Do you have any other Page 134 1 recollection of this problem? 2 A. No. 3 Q. But you can't tell me as we sit 4 here how many patients this might have involved? 5 A. No. 6 Q. How many studies this might 7 have involved? 8 A. No. 9 Q. How this might have affected 10 the results of any particular studies? 11 A. I don't know the answer to 12 that, I don't know any specifics about the 13 problem. 14 Q. Obviously you were comparing 15 and trying to merge several studies, and you had 16 patients that had the same Almedica kit numbers. 17 You couldn't tell what patients were actually 18 getting what medications, could you? 19 A. That's true. 20 Q. So you couldn't make any 21 determination with respect to whether or not the 22 drug helped, injured or was totally inactive with 23 respect to a particular patient? 24 A. That is true. Page 135 1 Q. And if you found an adverse 2 reaction, reported by that patient with that 3 particular Almedica kit, you couldn't know 4 whether you could attribute that to Fluoxetine, 5 the comparitor drug or even the placebo? 6 A. No. If that's how it was 7 reported to the FDA, it would be a problem. 8 Q. Do you think that it wasn't 9 reported to the FDA in that manner? 10 A. That would be something we 11 would look for in testing the output from any 12 program that we wrote. 13 Q. Well, do you know whether or 14 not medical quality assurance was ever called in 15 to review this problem? 16 A. I do not know. 17 Q. Do you talk with Comrade 18 Mattlar at the medical quality assurance group? 19 I'm sorry for laughing, that wasn't professional. 20 Did you talk to Ms. Mattlar concerning this? 21 A. Concerning this, I do not 22 recall that I did. 23 Q. Do you know whether or not 24 there were any additional memos or corrective Page 136 1 measures taken other than what is demonstrated 2 here to solve this problem? 3 A. I don't remember. 4 Q. Do you have a recollection that 5 the problem was solved? 6 A. Oh, I don't have any 7 recollection of what solution would have come out 8 of this. 9 Q. Well, you say this problem 10 could never happen again. I would assume that it 11 was your feeling at that time that this did solve 12 the problem that had arisen? 13 A. The wording of the message that 14 I typed here, is that this is a proposal. Based 15 on the messages that I have here, I have no 16 indication whether this proposal was accepted or 17 not, and I do not remember what the solution of 18 this situation was, I don't have any information 19 in front of me to indicate that. 20 Q. Would that -- should that be in 21 your file since this was in your file? 22 A. I don't know that. 23 Q. Would it have been your 24 practice to put everything in connection with Page 137 1 this problem into your file? 2 A. At the time I was working on 3 the product, yes, it would have been. 4 Q. When you say product, you mean -- 5 A. On Fluoxetine. 6 Q. Did you make any, that you 7 recall now -- or could any inquiry have been made 8 to determine how many patients were affected by 9 this Almedica randomization problem? 10 A. I don't know that, I don't know 11 the answer. 12 Q. Based on the -- 13 A. There could have been an 14 investigation, I don't know if there was or what 15 the outcome of it was. 16 Q. Who would have made that 17 investigation? 18 A. Based on the date this was 19 sent, that would have been either myself or any 20 other analyst who was working on Fluoxetine at 21 the time. 22 Q. On the first page of this 23 document, there's a listing of a lot of 24 individuals. Why don't you identify those Page 138 1 individuals for me, starting with Kathy Andrews. 2 A. All of these on the list, on 3 the primary addressee list, are systems analysts 4 within medical systems department. The CC is a 5 department head in that same department. 6 Q. Just so I'm clear, the problem 7 was that there was an error that had occurred in 8 Fluoxetine where the patient or kit numbers were 9 reused in a different protocol and patients were 10 incorrectly randomized when the protocols were 11 combined; is that your -- is that an accurate 12 statement of what the problem was? 13 A. No. 14 Q. All right. 15 A. What this is describing is that 16 the patients for a given protocol received the 17 drug they should have or would have been 18 randomized to. They got the drug they should 19 have gotten during that protocol. When the study 20 was unblinded or we began reporting the data 21 based on what the randomizations were and 22 actually let people know what drug the patient 23 was on, there was an error in determining which 24 patient was on which drug, and that would have Page 139 1 been discovered by a report on a patient on a 2 study receiving drug that should not have been 3 included in that study. 4 Q. But it would still affect 5 whatever you had reported up to that point, would 6 it not? 7 A. It would reflect what was 8 reported using the Almedica randomization. 9 Almedica was not used during all Prozac clinical 10 trials. 11 Q. How many clinical trials was it 12 used on? 13 A. I do not know, it was purchased 14 after I started with the company. 15 Q. How was -- how were the 16 patients randomized prior to use of the Almedica 17 randomization? 18 A. There was another system in 19 place at that time, I'm not familiar with the 20 mechanics of that system. 21 Q. What was the name of that 22 system? 23 A. I don't know the name of the 24 system. Page 140 1 MR. SMITH: That's all I have at this 2 time. 3 * * * * * * * * * * 4 CROSS EXAMINATION 5 BY MS. ZETTLER: 6 Q. Ms. Sheldon, my name is Nancy 7 Zettler and I represent a group of plaintiffs in 8 a case called Fentress versus Shea Communications 9 that results from a number of deaths and injuries 10 resulting from a man named Joseph Wesbecker going 11 to his former place of employment and shooting a 12 bunch of people with an AK forty-seven. I guess 13 I'm a little confused what happened here. You 14 said that -- we were talking about Exhibit 3. 15 You said that the patients were randomized 16 through the Almedica properly, but it was a 17 problem with the reporting after the unblinding. 18 Is that what's being talked about here? 19 A. When data was combined from two 20 studies on a report, in determining which patient 21 was on which drug there was a possibility of -- 22 for example if there were two records that showed 23 kit number one, two, three, four, there was a 24 possibility that that patient had kit number one, Page 141 1 two, three, four, picking up the wrong 2 randomization or wrong drug because there was not 3 an indication on the record which study that 4 record belonged to. 5 Q. When you say on the record, 6 what do you mean, the Almedica record? 7 A. On the Almedica record. 8 Q. So is it possible that on one 9 study kit number one, two, three four could have 10 been Fluoxetine, and on another study it could 11 have been a placebo? 12 A. Yes. 13 Q. And unless you can match up the 14 study that -- the specific study that each one, 15 two, three, four was related to, you don't know 16 if that patient was on a placebo or on 17 Fluoxetine; is that correct? 18 A. If that were the only data that 19 we had. 20 Q. Did you have other data? 21 A. The name of the file. In this 22 case, at the bottom of the first page, is says 23 panvalet member name. The name of the actual 24 file contains what study the patient was on. The Page 142 1 problem being described here is that that 2 particular study code was not within the record 3 itself, this is a problem in combining data. If 4 the data is just -- if you try to put all the 5 data together in one place and combine it, 6 there's a risk that numbers that have been reused 7 would cause this problem, unless you have some 8 other indication. For example, the code, that it 9 says down, panvalet name, unless that information 10 is available on the record, there is that risk. 11 Q. What is a panvalet name? 12 A. Panvalet is just a type of file 13 that is used on the main frame. 14 Q. Is that exclusive to 15 Fluoxetine? 16 A. No. 17 Q. What does panvalet stand for? 18 A. I have no idea, I honestly 19 don't know. 20 Q. So the panvalet member name in 21 this situation would be B1Y-M-C-H-C-C-A? 22 A. Yes. 23 Q. And is it your understanding 24 that B1Y is an indication for Fluoxetine? Page 143 1 A. Yes. 2 Q. What does the MC stand for? 3 A. I'm not positive, I believe it 4 just stands for McCarty, that this is where the 5 study was run from. 6 Q. That would be the affiliate, so 7 to speak, the Lilly affiliate? 8 A. Yes, yes. 9 Q. And HCCA is the protocol 10 initials? 11 A. Yes. 12 Q. The study number initials? 13 A. Yes. 14 Q. So are you saying that we could -- 15 that you could -- were able to tell by looking at 16 this panvalet name, which drug the patient in 17 panvalet name B1Y-M-C-H-C-C-A was taking which 18 drug? 19 A. I'm not sure I understood that 20 question. It might just be that I need to hear 21 the whole question. 22 Q. It might be because I'm not 23 completely clear. So you're saying that the 24 Almedica record did not have the panvalet name in Page 144 1 it? 2 A. Right, the record itself did 3 not have the file name in the record. 4 Q. Okay. So do you know for sure 5 that there was a switch of numbers or a problem 6 with the numbers or is this just a possibility 7 that was brought up? 8 A. I don't know for sure that 9 there was a switch, I don't know. 10 Q. Did anybody come in and say we 11 ran a test program on this or a test analysis on 12 this information and it looks like there's a mix 13 up on the kit numbers? 14 A. I don't remember. 15 Q. How is the proposed 16 renumbering, so to speak, how does that differ 17 from the panvalet number that was already used? 18 A. The panvalet name -- this was 19 not indicated in my response, the panvalet file 20 name would have been the same, that would not 21 have changed. 22 Q. Okay. 23 A. If you look at the example in 24 the proposal that I stated, the last four letters -- Page 145 1 Q. Which page? 2 A. I'm sorry, I apologize. On the 3 third page. 4 Q. Pz 1235? 5 A. One seven five, right. At the 6 end of the line, it says one, two, three, four A 7 Fluoxetine. At the end of that line, it says A, 8 B, C, D. 9 Q. Okay. 10 A. That would have been -- that 11 would have matched -- and this example was 12 probably a poor example because it didn't match 13 the way this problem was originally stated. But 14 that would have been the last four characters of 15 the file name or it would have been the project 16 code. 17 Q. Okay. So it would have been 18 HCCA, in this case? 19 A. Right. 20 Q. I'm sorry, go ahead. 21 A. That was pretty much it, just 22 that that would -- the problem that I had 23 described, that would solve the problem because 24 it would put that code on the record so that it Page 146 1 would be more easily accessible. 2 Q. So on the third page here, your 3 example, this first four digits and that one, 4 two, three, four, and that A, Fluoxetine, is what 5 already existed in the Almedica records? 6 A. Yes. 7 Q. And you were adding the project 8 code? 9 A. Yes. 10 Q. So theoretically you could have 11 had a patient or a kit with one, two, three, four 12 for every patient in the study as long as you put 13 the project code in there to differentiate? 14 A. No. You could reuse the number 15 one, two, three, four on different projects. 16 Q. And when you say different 17 project, you mean? 18 A. Different protocols. 19 Q. Different protocols within 20 Fluoxetine? 21 A. Within Fluoxetine. 22 Q. So if there was more than one 23 study done under any given protocol, you wouldn't 24 use the one, two, three, four in the different Page 147 1 protocols? 2 A. I think there's a mis -- I'm 3 not sure I understand the difference between a 4 study and a protocol the way you're describing. 5 Q. Let me ask you this: Is it 6 your understanding that there could be numerous 7 sites, investigation sites conducting studies 8 under one protocol? 9 A. Yes. 10 Q. So if say you have investigator 11 number one who is in Chicago and he's conducting 12 a study, and investigator number two is here in 13 Indianapolis, and he's conducting a study, 14 although they're working under the same protocol, 15 you wouldn't be able to use one, two, three, four 16 with investigator number one, with one of his 17 patients, and one, two, three, four with 18 investigator number two's patients? 19 A. No, you would not. 20 Q. Have you ever heard of AS 400? 21 A. Yes. 22 Q. What is AS 400? 23 A. AS 400 is a computer or type of 24 hardware. Page 148 1 Q. Where does it fit into the 2 System 38 main frame, DB2, all that other stuff, 3 where does it fit in? 4 A. It is a System 38 replacement 5 machine, it replaced the System 38. 6 Q. Okay. Do you know when it 7 replaced System 38? 8 A. No, I do not, it was at some 9 point while I was a data base administrator. 10 Q. Before you came back to work on 11 the approval letter? 12 A. I do not know that, I don't 13 know for sure. 14 Q. Okay. So essentially it was -- 15 its function was the same as System 38, it was to 16 collect and store data from clinical trials? 17 A. Yes. 18 Q. It wasn't used to analyze data 19 like the S-A-S? 20 A. No. 21 Q. Do you know why they changed 22 from System 38 to AS 400? 23 A. IBM no longer supports a System 24 38 as a strategic platform. Page 149 1 Q. So it's basically being a lack 2 of servicing it if something happened to it? 3 A. Correct. 4 Q. Was there any other changes as 5 far as AS 400, were there any substantive 6 differences between the two? 7 A. I was not involved in any of 8 the conversion, I don't know the answer. 9 Q. Have you used the AS 400? 10 A. No, I have not. 11 Q. The S-A-S, and I'm about as 12 knowledgable in computers as Paul, so you're 13 going to have to bear with me, okay, but the 14 S-A-S -- my understanding is, and 15 correct me if I'm wrong, is that it's statistical 16 software, so to speak, it teaches your computer 17 how to analyze information from a statistical 18 point of view? 19 A. It can be used for statistical 20 analysis or any type of reporting. 21 Q. Was it used with regards to 22 Fluoxetine information for statistical analysis? 23 A. Yes. 24 Q. Was it used for reporting? Page 150 1 A. Yes. 2 Q. And was it used for reporting 3 without statistical analysis, I mean is there a 4 difference between using it for statistical 5 analysis and using it for reporting? 6 A. I'm not a statistician, so I 7 have a hard time understanding the definition of 8 statistical, maybe the way you intend it. So I'm 9 not sure I know how to answer that question. 10 There were -- the reports that I produced 11 included occurrences and percents of patients. 12 Whether that's considered statistics or not is 13 something that I'm not sure I have an opinion or 14 I'm knowledgable enough to give you an opinion on 15 that. 16 Q. Could you with this software 17 run or pull information out that simply just 18 lists adverse events, every single adverse event 19 suffered by a patient, per se? In other words, 20 patient number one suffered from a nosebleed, 21 patient number two suffered from a broken ankle, 22 you know, things of that nature? 23 A. Yes. 24 Q. So it wouldn't necessarily have Page 151 1 to be broken down into percentage or things like 2 that, you could just print out straight 3 information? 4 A. Yes. 5 Q. In other words, you could ask 6 for information such as what were the percentage 7 of people across the board who suffered from 8 depression? 9 A. Yes. 10 Q. And it would do that for you? 11 A. Yes. 12 Q. Were there specific types of 13 statistical analyses that were already included 14 in the software, or was that something that had 15 to be put into the computer using the software? 16 In other words, if you wanted to run a specific 17 type of statistical analysis using a specific P 18 value or variable, things of that nature, would 19 you have to enter all that information in to get 20 the computer to be able to do that analysis or 21 was that something already built into the 22 software? 23 A. What information are you 24 referring to when you ask would I have to enter Page 152 1 the information? 2 Q. I'm not talking about like the 3 raw data from clinical trials, but in other 4 words, if you wanted to -- and I don't know 5 anything about statistics either, but if you 6 wanted to use a particular analysis, okay, we 7 want to use this P value and -- or determine this 8 P value or figure out the percentage of this and 9 this, is that something you have to put into the 10 computer or is that something that's already in 11 the software? In other words, is there a 12 specific type of statistical analysis that's 13 already included in the software or does that 14 give you the ability to put your statistical 15 analysis into the software and use it? 16 A. The software language gives the 17 programer the ability to specify which fields are 18 being compared to get the analysis. 19 Q. Let me make sure I understand. 20 A. Does that answer your question? 21 Q. I think so, but let me make 22 sure I understand. A field would be -- for 23 instance, adverse events would be considered a 24 field; correct, like a COSTART or ELECT term Page 153 1 would be considered a field? 2 A. Yes, a term would be considered 3 a field. 4 Q. And like for the patients -- 5 the demographic information on a patient, such as 6 like age, sex, things like that, would each 7 individually be considered a field? 8 A. Yes. 9 Q. The age would be a field, the 10 sex would be a field? 11 A. Yes. 12 Q. Patient number would be a 13 field? 14 A. Yes. 15 Q. What I'm trying to find out, 16 rather ineloquently, is are those fields, are 17 they something that are already in the software 18 or does that software allow you to run a program, 19 number one, setting up those fields, and number 20 two, analyzing the data by various fields? 21 A. There would have to be a data 22 base designed to determine what those fields are, 23 and data would have to be -- there would have to 24 be a program to put the data into those fields. Page 154 1 Q. Okay. 2 A. In the data bases. 3 Q. So you would create -- let me 4 use adverse -- let me use ELECT term as a field, 5 for an example. You would create that field 6 within the software -- using the software, you 7 could create the field, pull that information 8 together -- in other words, if you just wanted a 9 list of all adverse events that were ever 10 reported with regards to Fluoxetine, you could 11 pull all that information from your System 38 12 data using the program? 13 A. Not from the System 38. The 14 SAS tool is not available on the System 38 or the 15 AS 400. 16 Q. What is it available on? 17 A. It's available on the main 18 frame, it's available on PCs. There are several -- 19 I'm not sure what they're advertising these days, 20 I don't know. 21 Q. So you would have to transfer 22 the information collected on the System 38 to the 23 main frame, but then using that example of 24 creating a field and pulling the information Page 155 1 together from the field of, say, adverse event, 2 you could then do that once it's on the main 3 frame, correct, or would you have -- 4 A. To do that across studies, it 5 may have to come from several different data 6 bases. If those data bases all are designed the 7 same way, yes, that's possible. 8 Q. Okay. Say you had SSAI 9 information, okay, coming from the CRF, and the 10 CRFs across the board were the same, and the SSAI 11 information on all those CRFs entered into the 12 System 38 or AS 400 by data processors, you know, 13 as it comes in it's entered in or whatever. If 14 it's then transferred to the main frame, can you 15 pull all that information if it's similar across 16 the board for all the CRFs, could you pull all 17 that information together to use with the S-A-S? 18 MR. MYERS: When you say all that 19 information -- 20 MS. ZETTLER: The SSAI. 21 MR. MYERS: Just the SSAI? 22 MS. ZETTLER: Yes. 23 A. Again, if it's all in the same 24 format, it can be all pulled together and Page 156 1 reported. 2 Q. Okay. Was there anything about 3 the S-A-S that would enable you to converge 4 information from all the different data bases? 5 A. Yes. Again, if they were all 6 designed the same. 7 Q. So it would have to be 8 identical formats of information? 9 A. Right. 10 Q. But somebody would have to go 11 into the S-A-S or use S-A-S to create that field 12 and pull together that information; correct? In 13 other words, S-A-S isn't a software program that 14 already sets out fields for clinical trials 15 saying these are the types of information you may 16 want to look at, like adverse events or efficacy 17 ratings, things of that nature? 18 A. Right. It's a programming 19 language that can be used in a variety of 20 business programming situations. Clinical 21 trials, in this case, being one of those. 22 Q. Okay. But to be able to use 23 that, you would have to have the information 24 already in the CRF that are transferred to the Page 157 1 System 38 and then taken and put into the main 2 frame; correct? 3 A. Yes, data would have to 4 originate from the paper source in that case. 5 Q. Okay. When -- did you use the 6 S-A-S to make the changes from the SSAI terms to 7 the ELECT terms? 8 A. Yes. 9 Q. Could you tell us how you did 10 that? 11 A. The dictionaries -- those two 12 dictionaries are stored electronically. The 13 program that I wrote would read in the SSAI data 14 and then use the ELECT dictionary to look up what 15 the ELECT term for that SSAI term would be. 16 Q. So the computer would do that, 17 you wouldn't have to sit there and actually look 18 up every single term that the SSAI would map to 19 in the ELECT? 20 A. Right. 21 Q. Did you program the computer to 22 make the changes? 23 A. I programmed the computer to 24 read those electronic files and report out those Page 158 1 changes. 2 Q. So you actually told the 3 computer -- again, for example, say SSAI said 4 suicide attempt and ELECT maps suicide attempt to 5 depression, like we talked about earlier in that 6 example? 7 A. Yes. 8 Q. Would you have, like, put in 9 the SSAI term and said find out what it maps to 10 in the ELECT dictionary and change all of those 11 to the term that it was mapped to, so in other 12 words change all suicide attempts to depression? 13 A. Change them in -- change them 14 where? 15 Q. Let me ask you this: My 16 understanding of what happened was the 17 information that was entered onto the CRFs as 18 SSAI information was transferred to the main 19 frame, and was the S-A-S program changed from, 20 again the example, suicide attempt to the ELECT 21 term, depression? In other words, any SSAI term 22 that mapped to a different term using ELECT was 23 changed to that ELECT term? 24 A. The best I remember, we did not Page 159 1 actually change data, but stored both versions, 2 what would have been reported from the clinical 3 report form and what it would have mapped to. 4 Q. Okay. Why was that done? 5 MR. MYERS: The change you made or the 6 data stored that way? 7 Q. Why was it being stored that 8 way? 9 A. Basically to serve as an audit 10 trail for where the data came from. If the data 11 were actually changed, there would be no way to 12 compare it back to the clinical report form with 13 any certainty that it was still the same data. 14 Q. So if you looked at the 15 original clinical report form and saw suicide 16 attempt, but after the change to ELECT, not 17 knowing about the change to ELECT, you found a 18 report saying that it said depression and you 19 wanted to find out what happened in between, you 20 could go back and see that it was originally 21 mapped to suicide attempt in SSAI and it was 22 remapped to depression under ELECT when the 23 change was made? 24 A. Yes. Page 160 1 Q. Was this a group of information 2 and was it all existing clinical report form data 3 that was transferred and changed at once, or was 4 this something that had to be done clinical 5 report form by clinical report form? 6 A. It was done -- again, it was 7 done programmatically. Any data that had already 8 been entered in SSAI, the program actually did 9 the mapping and reporting. 10 Q. Was SSAI in its own separate 11 data base? 12 A. It was a separate file or data 13 base, I believe. 14 Q. When you say -- when I say a 15 separate data base, like the DEN, drug event 16 network, was it something like that? 17 A. I don't know about the drug 18 network data base, I never worked with it so I 19 don't know. 20 Q. Do you have any understanding 21 whatsoever what the drug network data base was? 22 A. No, and I know in concept what 23 the purpose of it was, but I really don't know 24 anything about the data base itself. Page 161 1 Q. What is your understanding of 2 what the purpose was? 3 A. This was a means for collecting 4 spontaneous adverse event reports. 5 Q. Okay. 6 A. And reporting to the FDA. 7 Q. Would you consider that a 8 separate file or data base? 9 A. Yes. 10 Q. That's what I'm trying to 11 figure out, was SSAI, was that similar separate 12 file data base? 13 A. Yes. 14 Q. I'm not saying it has to be the 15 same type of information, but I mean just from a 16 computer industry standpoint, is that considered 17 a separate file? 18 A. Yes. 19 Q. To your knowledge, was there a 20 separate data file base for ELECT? 21 A. Yes. 22 Q. To your knowledge, was there 23 ever a separate file data base for COSTART? 24 A. I don't know, I don't know Page 162 1 whether there was or not. 2 Q. Are you aware that at some 3 point Lilly changed from ELECT to COSTART? 4 A. I don't know anything about 5 that, I haven't heard anything. 6 Q. Do you have any knowledge of a 7 separate file data base in which adverse events 8 where information was stored other than the SSAI 9 or the ELECT? 10 A. The events themselves or 11 dictionary for those events? 12 Q. The events themselves. 13 A. The events are not stored in 14 the dictionary. I'm not sure what you're asking. 15 Q. Okay. So when you say there's 16 a separate file data base for SSAI or ELECT, you 17 mean for the actual dictionary? 18 A. For the actual dictionary, yes. 19 Q. In other words, you could go in 20 and say, all right, I have an actual term of 21 nosebleed and I want to see where it maps to, 22 that's how you use that data base? 23 A. Yes, yes. 24 Q. But that doesn't talk about any Page 163 1 particular actual adverse event that actually 2 occurred, it talks about terminology, period. 3 A. Right. 4 Q. Do you have any -- I forgot 5 what your answer was to my original question. Do 6 you have any knowledge of a separate data base 7 that was used just to store adverse event 8 information from clinical trials? 9 A. There were several -- when I 10 was working on the drug, there were several 11 Fluoxetine data bases, and what data went into 12 which data base depended on which protocol or 13 clinical trial it was. 14 Q. Were there different data bases 15 set up for each protocol? 16 A. There were some data bases that 17 had more than one protocol. I don't recall 18 whether there were any that had only one protocol 19 in them. 20 Q. I think earlier you talked 21 about there was a data base for collecting safety 22 data; correct? 23 A. That was termed -- if I 24 remember correctly, that was in talking about the Page 164 1 data bases on the System 38 or the data entry 2 systems on the System 38, is that possible? 3 Q. Clinical report form 4 information, there were several Systems 38 5 systems? 6 A. Yes. It was referred to as the 7 safety system or the safety data entry system 8 because the only data that was collected on that 9 system was regarding the safety of the drug. 10 There were no efficacy data being collected in 11 that data base. 12 Q. Okay. So obviously adverse 13 event information would be collected in that data 14 base? 15 A. Yes. 16 Q. Any other types -- what other 17 types of safety information would be collected in 18 that data base? 19 A. Laboratory values for blood 20 tests that would have been done on a patient. 21 Q. Vital signs? 22 A. Vital signs, I think basic 23 demographic information on the patient. Anything 24 that might affect how the drug reacts on a Page 165 1 patient. 2 Q. And I think another one you 3 talked about was there was a data base for a 4 series of clinical trials that was done while you 5 were there, while you were working in that area? 6 A. Yes. 7 Q. Was that a specific series of 8 clinical trials or are you talking about all the 9 clinical trials that were run just while you were 10 working? 11 A. Just general clinical trials 12 that were run while I was working in the system. 13 Q. So it could have been 14 Fluoxetine compared to placebos or dose ranging, 15 and could have been Fluoxetine compared to 16 another anti-depressant, it doesn't matter, it 17 wasn't specific to one type of study? In other 18 words, all studies comparing Fluoxetine against 19 other anti-depressants or something of that 20 nature? 21 A. No, it was not. 22 Q. Okay. You said there were 23 several systems that were written to collect data 24 in new trials? Page 166 1 A. Yes. 2 Q. What do you mean by that? 3 A. The manner in which the case 4 report forms were written was changed, and there 5 were new systems written to reflect that style of 6 clinical report form. 7 Q. Okay. What was different about 8 the new clinical report forms opposed to the old 9 clinical report form? 10 A. I don't remember the 11 differences. A lot of it was aesthetic, I 12 believe, just layouts and so forth. 13 Q. You also said that you wrote 14 enhancements to these systems, in other words 15 additional -- I believe you said additional 16 screens for collection of additional data. Do 17 you remember saying that? 18 A. Yes. 19 Q. Fill me in on what you mean by 20 that? 21 A. I don't know if this is a 22 specific example I worked on, but it's an example 23 of what might have happened. We may have been 24 collecting blood pressure with the patient Page 167 1 sitting, and if we determined in clinical trials 2 to begin collecting blood pressure sitting, 3 standing and supine, we would have had to have 4 made enhancements to the system to capture all of 5 those different types of blood pressure 6 measurements. 7 Q. Were those done for a 8 retrospective analysis or retrospective 9 collection of data? In other words, did you 10 write additional -- did you write enhancements so 11 that the clinical investigator could go back and 12 look at his records and collect that information 13 if it was something that was done, or is that 14 something that you did for new trials that were 15 going to be done? 16 A. The requests were always for 17 new trials that were going to be run. 18 Q. Have you ever heard the term 19 rechallenge study? 20 A. I don't remember ever hearing 21 it. 22 Q. Are you aware of a study that 23 was done by Lilly where people who became 24 suicidal on Fluoxetine were taken off the drug Page 168 1 and then put back on the drug to see if they 2 would become suicidal again? 3 A. No, I was not aware of that. 4 Q. How about movement disorders, 5 were you involved in any studies that dealt with 6 movement disorders as related to Fluoxetine, like 7 tardive dyskinesia or tardive dystonia? 8 A. Those terms sound familiar, but 9 I don't remember any particular studies. I know 10 I've heard the terms, but I don't know if I was 11 involved in the studies, I can't remember that. 12 Q. Are there any other systems or 13 programs used at Lilly now, to your knowledge, to 14 analyze Prozac data or Fluoxetine data? 15 A. I don't know, I'm not involved 16 any longer. 17 Q. Earlier Paul was asking, I 18 can't remember the series of questions he was 19 asking you, but you were asked if he meant back 20 then, and I believe that was when you were first 21 working on Prozac. Was there a change -- other 22 than the AS 400, was there ever a change in the 23 systems or software or programs that you guys 24 used while you were working on Prozac? Page 169 1 A. I don't know the specific 2 answer to that. Computer hardware and software 3 always changes, and there are new releases, but I 4 don't know if there were any specific changes 5 while I was working on it. 6 Q. Other than keeping up with the 7 state of the art or whatever is out there or 8 upgrading systems that you already have or things 9 like that, do you recall any new programs or any 10 new systems that were substituted with regards to 11 Prozac while you were working there? 12 A. No. 13 Q. You said that you stopped 14 working with Prozac it the beginning of 1987? 15 A. I think it was about the Summer 16 of 1987. 17 Q. And you were off it for a short 18 period, and then in the Fall of 1987 they asked 19 you to come back and work on the response to the 20 approvable letter? 21 A. Yes. 22 Q. Were there any programs written 23 specifically for analyzing data to respond to 24 questions asked in the approvable letter to your Page 170 1 knowledge? 2 A. To the best of my recollection, 3 yes. 4 Q. Okay. Can you give me an 5 example of those programs? 6 A. I can't tell you exactly. 7 Q. Can you give me generally? 8 A. I really don't remember enough 9 specifics to tell you anything, I'm sorry. 10 Q. That's okay. Let me ask you 11 this: Is it your recollection that information 12 requested by FDA in the approvable letter related 13 to the safety of the drug in some respects? 14 A. I don't remember the specific 15 questions -- 16 Q. I'm not asking specifically. 17 A. -- on the approvable letter. 18 Q. I'm asking generally. Do you 19 remember -- let me ask it again. Is it your 20 understanding that through the approvable letter, 21 the FDA was saying we are ready to approve this 22 but we need clarification on certain aspects of 23 the drug whether efficacy or safety? 24 A. I think it's speculation. My Page 171 1 answer might be speculation, I'm not real sure 2 how to answer. 3 Q. Obviously they needed 4 additional information because you provided them 5 with additional information. 6 A. Yes, and I don't remember the 7 specific of what the programs were, so I can't -- 8 I'm not sure I can answer. 9 Q. Do you have a general 10 recollection -- and I'm not talking about 11 specific safety aspects or specific efficacy 12 aspects, but do you have a general recollection 13 of the letter asking for more information related 14 to various aspects of the drug including safety 15 and efficacy? 16 A. It may have been safety and 17 efficacy, but I don't recall. 18 Q. To your knowledge, they 19 wouldn't be asking -- would they be asking 20 questions on marketing of the drug? 21 A. I don't know that, I wouldn't 22 have been involved in answering marketing 23 questions at that time. 24 Q. Would you have been involved in Page 172 1 answering any questions other than safety or 2 efficacy questions? 3 A. I can't think of -- I don't 4 know what other data there would have been other 5 than safety and efficacy, and is depends on how 6 you would categorize the data we collected. But 7 in general there would be safety data and 8 efficacy data, that would be about all I would 9 have been responsible for. 10 Q. There's no other area of 11 possible FDA interest that you recall working on 12 other than safety and efficacy data? 13 A. I don't recall any. 14 Q. You said you were working in 15 the marketing information systems on sampling 16 activity? 17 A. I'm not currently working on 18 sampling, that's what Paul asked me what I was 19 working on when I left medical. 20 Q. What are you working on now? 21 A. I work on infrastructure of our 22 current systems, basically keeping our current 23 systems running. 24 Q. Trouble shooting type of stuff? Page 173 1 A. Yes. 2 Q. How long did you work on 3 sampling activity in the marketing information 4 systems department? 5 A. Approximately two years. 6 Q. And generally, that type of 7 work, the sampling activity, was related to 8 Prozac, as well as other drugs that were 9 manufactured by Lilly? 10 A. It was related to any drug that 11 was sampled by a sales rep. 12 Q. To your knowledge was Prozac 13 sampled by sales reps? 14 A. Yes. 15 Q. To your knowledge is it still 16 sampled by sales reps? 17 A. Yes. 18 Q. What type of work would your 19 department do related to the sampling? 20 A. Report balances of inventory to 21 the sales representives. 22 Q. I'm a little confused about 23 when you say for the salesmen. Did you report to 24 the salesmen what the balance of the inventory Page 174 1 was or were you reporting on the balance of the 2 inventory used by or to be used by the salesmen? 3 A. It was -- basically it was like 4 getting a bank -- we produced a bank statement 5 for the sales rep. It would be similar to you 6 balancing your checkbook. We provided the bank 7 statement, these are the records that we show or 8 the checks that -- for example the checks that we 9 see have cleared, and sales reps then would have 10 to sit down and balance, this is what I think I 11 have on hand, and this is what the computer 12 thinks I have. So it was basically like 13 balancing a checkbook. 14 Q. So, for example, you would say 15 Salesman A has ten thousand pills that he could 16 use for a sample in his account, and he takes 17 five thousand and he has to account for those 18 five thousand? 19 A. Yes. 20 Q. Is this done as a requirement 21 for reporting to federal regulatory agencies like 22 the FDA or DEA? 23 A. I'm not sure what you're 24 asking, I guess if you could rephrase the Page 175 1 question. 2 Q. Sure. Does Lilly have a 3 responsibility to keep track of the number of 4 samples that are given out to doctors because of 5 regulatory requirements? 6 A. I'm not sure of what the 7 specific law is on that. 8 Q. Okay, but generally -- 9 A. I couldn't answer that. 10 Q. Do you have any idea generally 11 if that's done as part of a requirement by 12 regulatory agencies? 13 A. I don't have any knowledge of 14 that. 15 Q. What happens with that 16 information after it's balanced, so to speak, 17 once the account is balanced, where does that 18 information go? 19 A. We maintain the data, we keep 20 the data in the data base. 21 Q. And what type of internal 22 personnel would want to see that data once it's 23 been collected and stored in the data base? 24 A. There's a department that's Page 176 1 responsible for helping the sales rep balance 2 their accounts, so to speak. 3 Q. Other than that department -- 4 what I'm trying to find out is if that data is 5 used for any other purpose within Lilly other 6 than just keeping track of the samples that are 7 given out? 8 A. Yes, I believe that it is. 9 Q. Can you give me an example of 10 what types of reasons somebody would want to look 11 at that data? 12 A. If a physician is under 13 investigation for misuse of samples, we may get a 14 request from whatever agency is investigating 15 that to report back what samples that the doctor 16 received from us. 17 Q. Does your department have 18 responsibility for keeping track of numbers of 19 prescriptions for Prozac that are written? 20 A. I'm not sure exactly what data 21 we keep on results of that nature, I'm not 22 familiar with that data base. 23 Q. But -- 24 A. So I couldn't tell you. Page 177 1 Q. Is there a data base in which 2 information is at least attempted to be 3 accumulated regarding prescriptions of Prozac? 4 A. Regarding results, but I don't 5 know if it's prescriptions, I honestly don't know 6 what data is kept. 7 Q. When you say results, what do 8 you mean? 9 A. How much in dollar value was 10 sold in a particular area of the country or 11 because of the confidential nature of 12 prescriptions, sometimes it's hard to pinpoint 13 how many prescriptions were written and where 14 they were written. I'm familiar with those 15 issues, but I don't know how much we've attempted 16 to collect, that's an area that's completely 17 outside of what I work with. 18 Q. To your knowledge, do 19 pharmacists have a responsibility to report to 20 the FDA or the DEA all prescriptions written by 21 them? 22 A. I don't know the answer to 23 that. 24 Q. But at least an attempt is made Page 178 1 at Lilly to predict the number of prescriptions 2 written through the number of sales or the amount 3 of sales that are produced in each region? 4 A. I don't know that, I'm not -- I 5 don't work with that area so I don't know. 6 Q. Could you tell me a little more 7 about the Entrex data system, the data entry 8 system? 9 A. What are you asking? 10 Q. Is it purely a data entry 11 system or is it used to store data also? 12 A. I'm not very well versed on 13 Entrex. When I joined the company, that was the 14 machine that was there and some of the data that 15 was on the main frame had originally been entered 16 in that system. But how that system was used and 17 any of the mechanics of that system, I'm not 18 familiar with. 19 Q. Were they firmly entrenched in 20 the System 38 when you started? 21 A. Yes. 22 Q. But it's your understanding 23 that the information from Entrex had been at 24 least copied into the main frame? Page 179 1 A. Yes. 2 Q. When I say that the adverse 3 event terms were changed from the SSAI terms to 4 the ELECT terms, was that done through what you 5 call a batch edit? 6 A. I'm not sure I know what you're 7 asking. 8 Q. Are you familiar with the term 9 batch edit? 10 A. Yes. The only editing that I 11 know that was going on or that I was ever 12 involved with was on the System 38, and I didn't 13 do anything there that would remap the SSAI terms 14 to the ELECT terms. So as far as that remapping 15 in terms of edit, I don't know of anything, I 16 don't know anything about that. 17 Q. Did you run any batch edits on 18 the System 38 data? 19 A. Yes, that was in my initial job 20 assignment the first several months I was with 21 the company. 22 Q. Can you give me an idea what 23 type of batch editing you ran? 24 A. The edits that were on the Page 180 1 System 38 were, for the most part, reasonableness 2 edits and consistency. If an adverse event was 3 reported in the third visit a patient made to the 4 physician, and it was not -- it didn't have an 5 end date, it had to continue to show up in the 6 data base until the physician put an end date. 7 There would have been an edit to make sure that 8 that had continued to show up, that it wasn't 9 just neglected. There would have been range 10 edits to flag anything -- for example if the 11 physician reported ninety-eight point six degree 12 temperature, and marked celsius, that might be a 13 problem that we would want to at least go back to 14 the paper and probably go back to the physician 15 and clarify. 16 MR. SMITH: That would be some hot 17 data. 18 A. Those were the types of things 19 that were done. 20 Q. So basically what you could do 21 is, with the batch edit, take all the information 22 collected during a study from CRFs and run that 23 on the entire batch of information altogether? 24 A. No. An edit would only be run Page 181 1 on data that was in that data entry system. 2 There was an edit for each data entry system, and 3 you couldn't combine data from the different data 4 entry systems. So when I talk about the safety 5 data entry system and then the one that primarily 6 contained most of the information for the studies 7 that were ongoing when I was there, I couldn't 8 combine the data from those two systems into one 9 edit, I don't think. To the best of my 10 recollection, that was not done, there were 11 actually two separate edits that were being done. 12 Q. When you say systems, in that 13 context are you talking about the data base 14 files, is it the same thing? 15 A. There were different data entry 16 screens, there were different files. The data 17 was -- there were different data entry systems. 18 The data that was entered in this data entry 19 system was stored in a different file from data 20 entered into the second data entry system. 21 Q. You're not talking about a 22 System 38 as opposed to another physical -- 23 A. Right, I'm talking about 24 different files on the System 38. Page 182 1 Q. Okay. So if you have what we 2 talked about earlier, the safety data base, and 3 the data base for the series of clinical trials 4 that were done while you were there, you could 5 run batch edits on each of those separately but 6 you couldn't merge everything together and run it 7 all together? 8 A. Right. 9 Q. But as far as each individual 10 system, you could run batch edits on all the 11 information that was included in that data base. 12 In other words, you could run a batch edit on all 13 the information that's included in the safety 14 data base? 15 A. I think so. 16 Q. Okay. 17 A. It's been a very long time 18 since I worked with those systems, so -- the best 19 of my recollection, that's true, that's a true 20 statement. 21 Q. Why wouldn't a changing of the 22 terminology from SSAI to ELECT be considered a 23 batch edit then in that situation? 24 A. The files on the System 38 were Page 183 1 used to store the data as it was entered from the 2 clinical report form, and the remapping of the 3 SSAI data would not have been stored in that data 4 base because it was not entered from the case 5 report form that way. 6 Q. Where was it from? 7 A. The remapping was entered into 8 the dictionary, there was one dictionary for all 9 of medical. And then our data would go through 10 that program that I wrote, and if it were an SSAI 11 term, my program would look up in the dictionary 12 and say here's what the ELECT term is, and that 13 was all done on the main frame or the reporting 14 of that was done on the main frame. 15 Q. Okay. When you say one 16 dictionary for all of medical, what do you mean? 17 A. What I mean is that if a bee 18 sting were mapped to a rash on a Fluoxetine 19 study, it was a bee sting was mapped to a rash on 20 any study that was done on any drug in medical, 21 that was a medical-wide data base, there were not 22 separate dictionaries for the different products 23 or different studies. 24 Q. Is there one dictionary -- this Page 184 1 one dictionary different than the ELECT 2 dictionary or the the SSAI dictionary? 3 A. I'm not sure I understood your 4 question. 5 Q. It seems to me that now you're 6 talking about another entire adverse event 7 dictionary. 8 A. No. 9 Q. So you're saying once they made 10 the decision to go from SSAI to ELECT, ELECT was 11 the medical community-wide dictionary that was 12 used? 13 A. Yes. 14 Q. Okay. 15 A. Right. And there was one file 16 that those terms would be kept in, and then all 17 systems would record terms and they could refer 18 to that file to see what those terms mapped to. 19 Q. And that's going back to the 20 actual, let's use, the ELECT dictionary file 21 where all you would do with that file was put in 22 an actual event term and see where it maps to 23 within the dictionary? 24 A. I'm not sure I'm following Page 185 1 which files you're talking about, you may want to 2 start over. 3 Q. I want to make sure that we're 4 not talking about the same files, okay, I want to 5 make sure that you're not talking about more 6 files than, at least I think, I understand 7 existed. But when you talk about -- you talk 8 about the one dictionary for all of medical being 9 the ELECT dictionary, okay? 10 A. Yes. 11 Q. We talked about earlier the 12 files that -- the data bases that were started or 13 maintained for the SSAI and for the ELECT, there 14 were two separate files, right? 15 A. Okay. There are separate files 16 of -- maybe the difference is there are clinical 17 trial data files and there were dictionary files 18 that were used as a reference for clinical trial 19 data. Does that help? 20 Q. I think so, yes, but let me 21 make sure I understand, though. So you had a 22 dictionary file for ELECT that was nothing more 23 than containing the ELECT dictionary, and to use 24 that or the way you would use that or want to use Page 186 1 that is that you would look to see where a term 2 mapped to in the ELECT, period? 3 A. Yes. 4 Q. Okay. No other data from any 5 particular clinical trial was stored in that 6 file? 7 A. In that dictionary, no. 8 Q. But then you had other files 9 related to the clinical trials where information 10 collected and termed as an adverse event was 11 stored? 12 A. Were stored in those files. 13 Q. Those files, separate and apart 14 from other information that was kept or collected 15 by the CRF? 16 A. No. The data would have been 17 collected from the clinical report form. 18 Q. Okay. But was it stored with 19 all the rest of the data? In other words, was 20 there one data base where all you had in the data 21 base was the clinical report form and computer 22 form, more or less, or was that a separate file, 23 in other words was there a separate file for 24 adverse event data that was collected through the Page 187 1 clinical report form? 2 A. There were several adverse 3 event files, but they related to how the data was 4 collected in the first place. Is that what 5 you're asking? 6 Q. Sort of. 7 A. The same as there were several 8 data entry systems, there were several data bases 9 on the main frame where the data was stored for 10 analysis. 11 Q. But it would depend on how the 12 original CRF was structured? 13 A. Yes. 14 Q. Let me ask you this: When a 15 CRF was created for a particular study or 16 protocol, I'll use those as one and the same in 17 this instance, okay, would a data base -- did a 18 data entry data base start for that particular 19 CRF? 20 A. In some cases, yes. 21 Q. And those would be the cases 22 where the CRF differed from previous CRFs or the 23 norm? 24 A. Yes. Page 188 1 Q. Once that information was 2 collected, was a separate data base started for 3 just adverse event information from those CRFs? 4 A. There was no separate data base 5 specifically for adverse events. There was a 6 portion of each data base that was dedicated to 7 adverse event. 8 Q. So almost like a separate file 9 within the data base? 10 A. Yes. 11 Q. So you could have a larger data 12 base for a CRF from study number one, and within 13 that data base have a bunch of separate smaller 14 files that contained, like I said, blood work 15 information, adverse event information, 16 demographic information, things of that nature? 17 A. Yes. 18 Q. So when you did the change from 19 SSAI to ELECT, did you have to do it per file 20 then? In other words, did you have to go through 21 and say for the CRF for study A, we're going to 22 go into this subfile, for lack of a better 23 phrase, and change -- you know, change the 24 mapping or not change it, but record the mapping Page 189 1 differences for that information? 2 A. It depended on whether the file 3 structure was the same. It really gets down to 4 the adverse event file here look like the adverse 5 event file here. If they looked the same, they 6 could run through the same program all together. 7 Q. All together. 8 Q. So for instance if studies A, 9 B, C and D, the portion of the CRF that reported 10 adverse events were identical, you could run the 11 change on all of that at once as opposed to 12 having to do each individual one separately? 13 A. Yes. 14 Q. Did you actually do -- other 15 than writing the program, did you actually do the 16 change of the information on all that data that 17 was in existence at the time you worked on it? 18 A. My job was to write the program 19 that would report the change. 20 Q. And somebody else would 21 actually do the manual going into each file and 22 making the change? 23 A. Going into which files, we're 24 back to the files. Page 190 1 Q. Sure. We just talked about the 2 different -- how you would do it. If the portion 3 of the CRFs were the same, you could do it for up 4 to twenty studies at a time, right, or however 5 many studies want to see information that was 6 recorded was identical, right? 7 A. Yes. 8 Q. What I'm just trying to find 9 out is were you the person who actually, after 10 you wrote the program, you know, changing, doing 11 the remapping, so to speak? 12 MR. MYERS: Is that what you mean when 13 you say change, the mapping of the SSAI term to 14 the ELECT term? 15 MS. ZETTLER: Right. 16 Q. Because I understand that it 17 was done, you maintained the original term and -- 18 but you also started utilizing the ELECT term; 19 correct? 20 A. Yes. The program -- any 21 storage of -- here's the adverse event this 22 patient had as it was recorded in SSAI, and 23 here's what it mapped to in ELECT, would have 24 been done though running the program and the Page 191 1 computer writing out that mapping, if it were 2 stored in the file. There would not have been a 3 person going in and manually doing data entry. 4 Q. Oh, no, I'm sorry, that's not 5 what I mean. What I mean was were you the person 6 who actually ran the program once you wrote it or 7 did somebody else run the program on the various 8 files? 9 A. I don't know, I don't know. 10 Q. Do you have an idea how long it 11 took to actually complete that project, changing 12 from the SSAI to the ELECT? 13 A. I don't remember, I don't 14 remember how long it took. 15 Q. If I wanted to go back and find 16 out what the original SSAI term was for an event 17 that occurred during the clinical trial, could I 18 find that out by referring to patient numbers? 19 A. It would be a patient number 20 within a study, I believe. 21 Q. Okay. 22 A. There was a series of -- it 23 wasn't just the patient number, but a combination 24 of which study they were on and what patient Page 192 1 number to uniquely read the patient. 2 Q. Investigator number also? 3 A. Yes, I believe that's right. 4 Q. So to do that I would have to 5 have the patient number, the investigator number 6 and study initials or numbers? 7 A. Yes. 8 Q. But I would be able to track 9 that backwards, then? 10 A. Yes. 11 Q. Any other way I could do it 12 without those numbers? 13 A. To find what event a particular 14 patient had? 15 Q. To find -- yes, how they 16 originally were reported through the SSAI 17 terminology. 18 A. That would have been all stored 19 by project and investigator number and patient 20 number. If you had that information, I'm 21 presuming you would be able to find the paper 22 copy as well, that's -- I never, I was never 23 involved in going back to the paper for it. 24 Q. But as far as -- do you know of Page 193 1 any other way I could track, let's say, from the 2 way it was remapped to ELECT, starting from there 3 and working backwards to the original paper, is 4 there any way that I could track that information 5 without having the patient number, the 6 investigator number and the study? 7 A. To get all the way back to the 8 paper, you would have to have all that 9 information. 10 Q. Okay. Is there a way that I 11 could track going backwards what the original 12 SSAI terminology was for a particular event for a 13 particular patient without that information? 14 A. For a particular patient, I 15 don't believe so. 16 Q. I would still need the patient 17 number, study number and investigator number? 18 A. That's the key to all the 19 records. 20 Q. Can you get printouts of 21 information from the System 38 that include 22 things like patient numbers, investigator 23 numbers, things like that or is that something 24 you have to do through the SSAI? Page 194 1 A. Simple listings, I think, would 2 be available from the System 38. 3 Q. But if you -- if somebody asked 4 you to list every patient who had suffered from 5 depression as an adverse event, including their 6 patient number, study number, investigator 7 number, you would have to go through, say, S-A-S 8 for something like that? 9 A. I think so, yes. 10 Q. Would it be a matter of -- with 11 the S-A-S would it be a matter of being able to 12 set up any number of fields that you wanted as 13 long as the information that was -- information 14 that was entered into the original system, like 15 the System 38, and then transferred to the data 16 base? 17 MR. MYERS: Are you talking about now 18 or then because -- 19 A. I don't know now. 20 MR. MYERS: -- the System 38 and AS 400 -- 21 MS. ZETTLER: She said she hadn't 22 worked with AS 400. 23 A. I don't know what the data base 24 structure is now. I can only answer what the Page 195 1 data base structure looked like at the time I 2 last worked on it. If they have restructured 3 that data base, I couldn't answer that. 4 Q. My understanding of what was 5 gathered and stored and eventually transferred or 6 copied onto the main frame was all the 7 information that was contained on the CRFs, 8 including the demographic information, the 9 testing information results, adverse events and 10 things of that nature, almost all of that stuff 11 was all put on the System 38 and then eventually 12 transferred or copied onto the main frame; 13 correct? 14 MR. MYERS: At a certain point in time? 15 MS. ZETTLER: When she was working. 16 A. Yes. 17 Q. So theoretically, with S-A-S, 18 you could be able to set up, say, a chart listing 19 everybody by study name, patient number, 20 investigator number, and then ask for any other 21 information you want to regarding that patient as 22 long as it is something that had been entered 23 into System 38 and then copied onto the main 24 frame; correct? Page 196 1 A. Yes. 2 Q. Did you get a list of all the 3 studies that have been done off S-A-S? 4 MR. MYERS: With or without writing a 5 program? 6 MS. ZETTLER: Any way. 7 Q. Is it possible to get 8 information from -- 9 A. As I mentioned before, every -- 10 the key to every record is the study and the 11 investigator and the patient, so every record of 12 clinical trial data has those three pieces on it. 13 The description, the description of the study 14 would not be there. 15 Q. Where would that be? 16 A. I think there was a file on the 17 System 38 that had that information. 18 Q. To your knowledge, was somebody 19 or some department charged with keeping track of 20 what studies were run and what the status of each 21 study was at any given time? 22 A. That, I do not know, that would 23 have been in the Lilly medical component, not in 24 the systems component. Page 197 1 Q. But at least you could get a 2 list from S-A-S listing the study names and 3 numbers and investigators that had been entered, 4 at least, right? I mean as long as you had 5 information entered about a study, you could 6 retrieve a listing from S-A-S? 7 A. If that data were available on 8 the main frame. I don't know whether that data 9 is available on the main frame, I don't know if 10 that data is either entered or transferred to the 11 main frame, I don't know. 12 Q. As long as it exists. So, say, 13 with regards to any study that was done, as long 14 as that information about that study is in the 15 main frame, you can retrieve it back out? 16 A. Yes. 17 Q. Okay. 18 MR. SMITH: Can we take a short break? 19 (A SHORT RECESS WAS TAKEN.) 20 (PLAINTIFFS' EXHIBIT NO. 4 WAS 21 MARKED FOR IDENTIFICATION AND 22 RECEIVED IN EVIDENCE.) 23 Q. Have you had a chance to look 24 at Exhibit 4? Page 198 1 A. Yes. 2 Q. I realize it's just one page, 3 but do you have any idea what this is? 4 A. It's a listing of 5 discontinuations in the adverse events at that 6 discontinuation based on the title. Other than 7 that, I couldn't tell you anymore about it. 8 Q. Do you remember what this was -- 9 this was produced as documents that were in your 10 file, okay. So do you have any recollection just 11 from looking at this one paper what this was 12 related to? 13 A. No, I do not. 14 Q. Okay. Does the date up in the 15 upper right-hand corner help you at all, April 16 17, 1986? 17 A. At that time I was probably 18 working on the safety update. I don't know 19 whether this was for the safety update or not, 20 though, there's no indication on the paper. 21 Q. If you look at this in 22 connection with Exhibit 2, I believe, would that 23 help refresh your recollection as to whether or 24 not you worked on the audit, and if you did what Page 199 1 you did with regards to the audit? 2 A. There's no table number on 3 this, I don't know whether this is the table 4 forty-one that's referred to in the audit or not. 5 Q. Okay, fair enough. Under the 6 titles of Fluoxetine study drug discontinuations, 7 adverse experiences by patient, it says all 8 Fluoxetine in blinded patients, all 9 investigators. Would that have been the result 10 of a query? 11 A. That would have been the result 12 of -- based on the printing of this, this came 13 from the main frame printer. This would have 14 been a report that was run from the main frame 15 through programming. 16 Q. Would that have been using the 17 S-A-S? 18 A. Yes. 19 Q. Underneath that it says drug 20 name, blinded, PIP, VIS, SSI term. The PIP, is 21 that patient investigator project, is that what 22 that stands for? 23 A. Project investigator patient, 24 in that order. Page 200 1 Q. And that's what we were talking 2 about before the break, how information per 3 patient is tracked? 4 A. Yes. 5 Q. And each patient is tracked 6 through the project initials, the investigator 7 number and patient number, right? 8 A. Yes. 9 Q. And then over to the right it 10 says SSI term? 11 A. Yes. 12 Q. Would that be signs, symptoms, 13 illnesses term? 14 A. I believe so. 15 Q. So these terms that are listed, 16 drug overdose, nausea, vomiting, insomnia, 17 nervousness, those all would have been terms that 18 were found in the SSI dictionary as opposed to 19 the ELECT dictionary? 20 A. They would have been found in 21 the SSAI dictionary, I don't know whether they 22 were in the ELECT dictionary or not. 23 Q. What I'm trying to find out, 24 though, these are SSAI terms, not ELECT terms? Page 201 1 A. Yes. 2 Q. Per this column? 3 A. Per this. 4 Q. Do you have any recollection as 5 to what terms changed when you went from the SSAI 6 to the ELECT dictionary? 7 A. No, I don't. 8 Q. And again, for this information 9 to be able to be retrieved in this form, it would 10 have had to have been somehow entered into the 11 main frame, either directly or by transferring it 12 from System 38? 13 A. It would have been -- the data 14 itself would have been transferred from the 15 System 38 and this would have been produced as a 16 result of the program. 17 Q. Okay. Would this information 18 all have been information that was found on a CRF 19 at some point in time? 20 A. The drug name would not have 21 been on the CRF. 22 Q. Where would the drug name have 23 come from? 24 A. That would have come from the Page 202 1 panvalet files that were referred to. 2 Q. Almedica? 3 A. I don't know whether it would 4 have been Almedica or not, Almedica was not used 5 on all studies. 6 Q. Was there a separate file data 7 base for the the panvalet information? 8 A. Panvalet itself is a separate 9 file system, a system of files, yes. 10 Q. And in that system there would 11 be a file for each study, for each panvalet 12 member name? 13 A. Yes, each panvalet member name 14 would have contained a randomization. It depends 15 on the protocol, there may have been one. I 16 don't know if there were any cases where there 17 were ever just one. 18 Q. Earlier you testified that once 19 the information was accumulated in the System 38, 20 it was copied into the main frame. In other 21 words, if you look back at System 38, the 22 information would still be there, but a copy of 23 it would be transferred to the main frame for 24 analysis or use in whatever way you wanted to use Page 203 1 it? 2 A. Yes. 3 Q. To your knowledge does that 4 information exist -- that information that was 5 originally in the System 38, does that still 6 exist in either the System 38 or AS 400? 7 A. I don't know, I don't know. 8 Q. Do you know if it was Lilly's 9 policy to keep some sort of a copy computerized 10 or hard copy of the information that was once 11 retained, contained in the System 38? 12 A. To the best of my knowledge, 13 it's policy to keep at least the hard copy. 14 Q. Okay. To your knowledge would 15 it be possible for somebody to go back and ask 16 for a listing of what the original SSAI terms 17 were and what they mapped to in ELECT? 18 MR. MYERS: You mean the program she 19 did, as she did it on the program? 20 MS. ZETTLER: No, as a result of the 21 program. 22 Q. I mean obviously some terms 23 were remapped, right, some terms were remapped to 24 terms that were in the ELECT that were not in Page 204 1 SSAI. In other words, if there was a term in 2 SSAI that wasn't identical to an ELECT term and 3 that term mapped to an ELECT term, you could 4 theoretically get information as to what the 5 original SSAI term was and what event it mapped 6 to; correct? 7 A. You could at the time I was 8 working on the drug, report that. I don't know 9 about the format of that now and how accessible 10 it is. 11 Q. Could the information from a 12 System 38 be transferred to the AS 400? 13 A. I don't know anything about the 14 AS 400 other than what it is and that it was a 15 replacement. 16 Q. Obviously at some point before 17 you ran your program doing the remap to the ELECT 18 terms, you transferred the SSAI terms to the main 19 frame to work with it; correct? 20 A. The clinical data? 21 Q. Right. 22 A. Yes. 23 Q. What happened to that clinical 24 data after you completed your program and did the Page 205 1 remapping to the ELECT terms? 2 A. At the time, that Summer, the 3 data remained on the data base. 4 Q. Did you personally ever take it 5 off of the data base? 6 A. Delete it out of the data 7 bases? 8 Q. Right. 9 A. Not to my knowledge. 10 Q. When you say data bases, you 11 mean the main frame? 12 A. Main frame data bases. 13 Q. Okay. When you ran the program 14 switching from the SSAI to the ELECT terms, did 15 you do body systems as well as adverse event 16 terms? In other words -- well, did you do body 17 systems as well as event terms? 18 A. I don't recall -- I'm not sure 19 I understand the question. 20 Q. Let me rephrase it. My 21 understanding of how the COSTART dictionary 22 works, the FDA dictionary, is that you can 23 categorize or they categorize event terms by body 24 systems involved, for instance, central nervous Page 206 1 system, special senses, meaning hearing, sight, 2 things of that nature, muscle systems, different 3 systems of the body. 4 A. Okay. 5 Q. And there's also a category for 6 general, in which certain event terms are listed 7 that don't necessarily relate to one specific 8 body system. 9 A. Okay. 10 Q. Is that, to your knowledge, the 11 way the SSAI dictionary was set up also? 12 A. I do not remember, I don't 13 remember whether. 14 Q. Do you have a recollection of 15 changing, when you changed from the SSAI to the 16 ELECT dictionary, do you remember there being a 17 change in body systems? 18 A. I don't remember because I 19 don't remember what was in SSAI. 20 Q. Have you ever had -- have you 21 ever seen a hard copy of the SSAI dictionary? 22 A. I don't remember having one. 23 Q. The SSAI dictionary, are you 24 familiar with how the SSAI dictionary worked Page 207 1 before you changed over to the ELECT? 2 A. Aspects of it. 3 Q. Did they have different 4 subfiles within the SSAI data base? In other 5 words different ways you could use the dictionary 6 other than just mapping terms? 7 A. The dictionary itself? 8 Q. Right. 9 A. I don't remember that there was 10 anything like that. 11 Q. Have you ever seen a hard copy 12 of the ELECT dictionary? 13 A. I don't believe I have. 14 Q. Have you ever seen a copy of 15 the COSTART dictionary? 16 A. I don't think so. 17 Q. Have you heard of the DEN 18 dictionary? 19 A. No. 20 Q. How about the Lilly drug 21 dictionary? 22 A. Oh, boy, you're stretching my 23 memory. I believe there may have been a drug 24 dictionary, I haven't -- that's stretching, I'm Page 208 1 not real sure. It rings a bell, but it's very 2 faint. 3 Q. To your knowledge was that 4 dictionary, if it existed, would that have been a 5 computerized dictionary like ELECT or SSAI or was 6 that a hard copy form? 7 A. I don't remember ever seeing a 8 hard copy of a drug dictionary, so if there was 9 something we used, I don't know. It would have 10 been computerized -- I would assume it would have 11 been computerized. 12 Q. Look back at Exhibit 4. We're 13 talking about field. Would that be the things 14 listed at the top, drug name, PIP. I assume VIS 15 is visit? 16 A. Yes, that would be a field. 17 Q. These would each be a field? 18 A. Yes. 19 Q. And you could, theoretically, 20 if this type of information was available in the 21 data base, you could construct these fields in 22 any order you wanted to, right, if you wanted to, 23 in this example, you could just take PIP and 24 blinded, PIP and drug name, as opposed to all Page 209 1 four fields in this information? 2 A. That's the way I wrote the 3 program, yes. 4 Q. What's the difference between a 5 query and a program in this situation? 6 A. Query -- to use the term query 7 as I have used it and as I worked with it when I 8 was in medical, referred to very simple questions 9 that were written in a query language against the 10 System 38 data bases. And that was really the 11 only time that I used that terminology, query 12 terminology. 13 Q. So if it was a rather simple 14 request, then it would be considered a query as 15 it applied to a System 38, but if you got into 16 something more complicated that could be used 17 with S-A-S and the main frame, that you would 18 call it a program? 19 A. I'm not sure I agree with your 20 definition of the simplicity of the programs, 21 there were very complex queries done as well. 22 Q. When you say -- I guess I just 23 want to make sure that I'm not confusing the two, 24 okay. From what you've explained to me, and Page 210 1 again, it's probably because of my lack of 2 knowledge of computers, but from what you've told 3 me, it at least seems to me that there's not that 4 much difference between a query and a program 5 other than terminology and how it applies to the 6 main frame data base as opposed to the System 38? 7 A. A query -- the way -- like I 8 said, the way I used the term query, that was 9 based on using a query programming language on 10 the System 38 against System 38 data. In some 11 cases, the format of the data on the System 38 12 could not be listed out easily using the query, 13 and it made more sense to use a SAS program on 14 the main frame. 15 Q. Okay. When you set up a query 16 with the System 38, would you have to establish 17 fields in everything else like you would with the 18 program? 19 A. Yes. 20 Q. And in establishing fields, you 21 would have to, in effect, like, say, for 22 instance, and I don't know how many characters 23 are in this or not, but in Exhibit 4 start the 24 drug name, give it a certain number of characters Page 211 1 in which it can use up -- like in other words, 2 blinded looks like seven characters? 3 A. Uh -- 4 Q. In other words, could you start 5 like from one drug name would be in characters 6 one to seven? 7 A. I don't remember the query 8 language, I really can't tell you very much about 9 how that was done. 10 Q. Okay. But would you have to 11 set up fields, I mean columns, set up the colums 12 of fields and things like that also or was that 13 something already built into the system? 14 A. No, the language itself was 15 something that again was provided by IBM and was 16 not specific to clinical trial work, it was 17 simply a programming language. 18 Q. Okay. 19 A. It would have been the 20 analyst's responsibility to use that programming 21 language to extract information from the data 22 base and report it out. 23 Q. Okay. What I'm trying to do is 24 figure out what's involved in the query. Let me Page 212 1 give you an example. We have something as 2 lawyers call West Law, and that's a data base 3 that contains all kinds of reported cases, you 4 know, case opinions from judges in appellate 5 courts and things like that. And when we write a 6 query, we don't have to set up anything extremely 7 complicated, we can type in literally three or 8 four words if we think we're getting it narrowed 9 down to the information that we're looking for, 10 and it would go search the data base for any case 11 that has that information in it and report back 12 to us. 13 A. Okay. 14 Q. I'm getting an impression that 15 query, the way you use it, is more complicated 16 that that. 17 A. Yes. 18 Q. That what I'm trying to find 19 out, what a query is and how you would construct 20 a query. 21 A. It's more complicated. 22 Q. That's what I figured, but I 23 wanted to make sure. Why don't we try it this 24 way: Why don't you give me an example of using a Page 213 1 simple query, of how you would construct a query. 2 A. Again, if we're using query in 3 the sense that I've used the term query, I'm not 4 sure I can do that because I don't know the 5 language. As far as reporting something in SAS, 6 that's different. But I haven't used query 7 language on the System 38 or AS 400 for probably 8 eight years, so I'm not sure I can even give you 9 a description or an example. 10 Q. If I were to come to you and 11 say Lori, I want to know how many forty-five 12 years old white males on Fluoxetine attempted to 13 commit suicide. I would think of that as a 14 query, maybe, but you could construct a query 15 from that information to actually extract the 16 information from the data base? 17 A. That would probably be done on 18 the main frame to extract and report that 19 information. 20 Q. Okay. 21 A. If given -- again, I don't know 22 what the data base structure is now. If all the 23 data were in one place, the program would have to 24 be written to pull data from the demographic Page 214 1 file, adverse event file, and merge that data 2 together and report it. And I guess I don't 3 remember exactly what all you had in your 4 description, but it would take pulling data 5 together from several different files to get that 6 report. 7 Q. Okay. So you would have to 8 literally tell the computer go to the demographic 9 file and pull all the forty-five year old white 10 males -- obviously it's not that simple, but -- 11 and then go to the adverse event or safety file 12 and pull all of the corresponding patient numbers 13 for people who committed suicide. I mean, it 14 would be a matter of telling the computer to go 15 to different files to pull information to bring 16 it all together in one place? 17 A. Yes. 18 Q. And depending on the query -- 19 or depending on my request, as the person asking 20 for the information, would relate to how 21 complicated the query would be? 22 A. Yes. 23 Q. It could be something as simple 24 as pulling information from one data base, and Page 215 1 that would be relatively easy. It could be 2 pulling information from ten different data 3 bases, which would be relatively difficult or 4 very difficult? 5 A. Right. 6 (PLAINTIFFS' EXHIBIT NO. 5 WAS 7 MARKED FOR IDENTIFICATION AND 8 RECEIVED IN EVIDENCE.) 9 Q. Would this be an example, 10 Exhibit 5, would that be an example of a more 11 difficult query? 12 A. I think this would be a very 13 difficult query. 14 Q. Even I can recognize that. Let 15 me ask, do you recognize this? 16 A. I think I have seen this 17 report, but it's not something I'm familiar with. 18 Q. Okay. It says at the top 19 adverse experiences, first reportings and 20 discontinuations. If you were going to start 21 constructing something like this, would you start 22 with the adverse experiences as it's listed in 23 that far left-hand corner or would you first ask 24 for the first reports and the discontinuations? Page 216 1 MR. MYERS: Are you asking her if she 2 was going to do something like this? 3 Q. I'm trying to get an idea of 4 how something like this is constructed in a real 5 general sense, I don't need to know about 6 characters and which file you went to, but would 7 you rely basically -- in this case would you go 8 to the file that contained the adverse 9 experiences first or would that matter? 10 A. I don't know whether that would 11 matter. 12 Q. Because it seems to me in this 13 instance you're going to want to tell the 14 computer to search for adverse experiences first 15 reported and causing discontinuations, and then 16 ask for the rest of this information as far as 17 what visit it was and things of that nature. 18 A. Every -- at the time I was 19 working with these data bases, every adverse 20 event record was identified by the visit, that 21 was not necessarily separated from the adverse 22 event. Each adverse event was not only 23 identified by what patient had this adverse event 24 but also at what point during their therapy did Page 217 1 they have this event. So that's not necessarily 2 as far as when or what visit or so forth that was 3 reported. And actually, in this particular 4 report, if I understand this correctly, the 5 numbers across the tops are probably -- well, I 6 don't know. If the numbers across the tops are 7 weeks on therapy, that would have to come from a 8 combination of when they started the study, when 9 they started the drug, when the adverse event was 10 reported. There are numerous factors that would 11 have to be taken into consideration to determine 12 where the report -- whether this was a first 13 report and whether this was a hit in column five 14 or column two or whatever. 15 Q. Okay. But to a certain extent, 16 computers work on logic; correct? 17 A. Yes. 18 Q. And could you tell the computer 19 to start with looking at Exhibit Number 5, and 20 then ask it for all of this other information and 21 get something this comprehensive or would you 22 have to start with as broad a term as possible? 23 In other words, look at all adverse experiences 24 for biopsies, and then find the ones that have Page 218 1 this other additional information? 2 A. I'm not sure I'm understanding 3 your question. 4 Q. See where the dots are at the 5 top of the page, like sort of a broken line? 6 A. Right. 7 Q. It says drug, Fluoxetine body 8 system, body as a whole. Does this refresh your 9 recollection as to whether or not -- 10 A. There must have been a body 11 system classification in the dictionary, yes. 12 Q. So would you have searched just 13 a body system classification for this 14 information? 15 MR. MYERS: Are you asking is that what 16 they did when they did this report or if she were 17 doing a -- similar to what she would do? 18 Q. First of all, does this say to 19 you that just this body system classification was 20 searched for this information that's listed in 21 this particular printout? 22 A. This particular report, I have 23 no indication from the papers that are here that 24 this is the complete report. Page 219 1 Q. Right, I mean just for this 2 one. 3 A. For these five pages, given the 4 fact that it starts at A and ends with W, I would 5 presume that this is all of the adverse events 6 that were classified as body as a whole adverse 7 events. 8 Q. Okay. 9 A. Just given the information that 10 I have here. 11 Q. Okay. So you could start as 12 broad a body as a whole this category and then 13 exclude adverse events if you wanted to, couldn't 14 you? 15 A. I'm not sure what you're 16 asking. 17 Q. For instance, in here there's 18 no adverse event of overdose listed, okay, I'll 19 just use that as an example. So say if under the 20 category of body as a whole, the adverse event of 21 overdose is usually listed and it's not in here, 22 would that mean that it had been excluded? 23 A. That would -- 24 MR. MYERS: I don't mean to interrupt Page 220 1 you, Nancy, but it does say drug overdose on the 2 first page, see, so -- I'm not telling you what 3 to ask her. 4 MS. ZETTLER: This must be for ELECT. 5 MR. MYERS: I don't know, but that word 6 is there. 7 Q. Well, let's use suicide, we'll 8 use suicide instead. 9 A. I don't know what the body 10 system classification for suicide is. 11 Q. Assuming that -- is it 12 possible, assuming that suicide was listed under 13 body as a whole, just hypothetically, if it's not 14 in here would it be possible to exclude suicide? 15 MR. MYERS: If you're doing a search, 16 could you rule something out? 17 Q. Right. Could you say I want 18 all the adverse events listed under body as a 19 whole except, you know, nosebleed, suicide 20 attempt, depression, things that are normally 21 would be listed, if you were to list all the 22 adverse events. Could you intentionally exclude 23 them from this analysis? 24 A. It is possible to program that Page 221 1 way, if that's what you're asking. 2 Q. Yes. Is there any way to tell 3 from this whether or not these were SSAI terms or 4 ELECT terms? 5 A. No. 6 (PLAINTIFFS' EXHIBIT NO. 6 WAS 7 MARKED FOR IDENTIFICATION AND 8 RECEIVED IN EVIDENCE.) 9 Q. This is obviously a page from 10 table forty-eight; correct? 11 A. Yes. 12 Q. And this is the table 13 forty-eight that we were referring to in Exhibit 14 Number 2, I believe? 15 A. Okay. 16 Q. Does that make sense that this 17 is a page from the table forty-eight or one 18 example of the table forty-eight that is 19 represented in Exhibit 2 of the safety audit? 20 A. Yes. 21 Q. Is there any way you can tell, 22 Lori, from Exhibit Number 6, if this is the 23 patient that's discussed in Exhibit Number 2? 24 A. I can't tell from what is here Page 222 1 whether it's anything that would have been in 2 here. 3 Q. Okay. Would you have a better 4 reference of being able to tell whether or not a 5 patient listed in Exhibit 6 is also discussed in 6 Exhibit 2 if you had the rest of the PIP number? 7 A. No. 8 Q. Okay. 9 A. There were no references in 10 Exhibit 2 to specific patient numbers. 11 Q. What about if you look at 12 Exhibit Number 2, the second page, it says under 13 specific observations, about halfway down the 14 page, it says HCAH, and then something that's 15 blanked out? 16 A. This refers to table forty-one. 17 MR. MYERS: Forty-eight's at the end of 18 the report. 19 MS. ZETTLER: I remember, sorry, you're 20 right. 21 Q. Okay. So as far as table 22 number forty-eight, they were not talking 23 specific patients, okay. Would you be able to 24 tell if, say, this was -- say that this was Page 223 1 Exhibit Number 6, okay, was a representation of 2 table forty-one as opposed to table forty-eight -- 3 would you be able to tell without the patient 4 numbers whether or not an individual patient was 5 discussed in Exhibit 2? 6 A. I could not tell that, I would 7 have to have the entire -- 8 Q. PIP number. 9 A. -- PIP number, or I would have 10 to have the clinical report form to determine 11 whether they were the same. 12 Q. Patients are referenced in the 13 clinical report form by PIP numbers also, aren't 14 they? 15 A. Yes. 16 Q. Okay. So if the clinical 17 report form, the patient number and investigator 18 number and study names were blanked out, you 19 wouldn't be able to tell either, would you? 20 A. No. 21 Q. Let me ask you this about 22 Exhibit 6: Would something like this be easier 23 or harder to construct than what's represented in 24 Exhibit 5? Page 224 1 A. I'm not sure I could give you 2 an answer off the top of my head on that. The 3 complexity of tables -- the table number 4 forty-eight comes from many more places than the 5 information that's in Exhibit 5, but the way 6 table five is constructed is very complex as 7 well. It wouldn't be comparing apples to apples, 8 in other words, and I'm not sure if I can give 9 you that kind of an estimate. 10 Q. Okay. I guess actually you did 11 answer my question in that the information pulled 12 together for table forty-eight is not something 13 that is found in one single data base? 14 A. No. 15 Q. What kinds of documents did you 16 keep in your files when you were working on 17 Fluoxetine? 18 A. I do not remember. 19 Q. Would you keep office 20 correspondence, memos, things of that nature? 21 A. If they pertained to the 22 projects, yes. 23 Q. Would you have communicated 24 with anybody outside of Lilly -- corresponded Page 225 1 with anybody outside of Lilly in your position? 2 A. I did not, as far as I can 3 remember, initiate any conversation with anyone 4 outside of Lilly. 5 Q. Were you ever contacted by 6 anybody outside of Lilly? 7 A. Not directly. 8 Q. How about indirectly? 9 A. Yes. 10 Q. What types of people would 11 contact you indirectly? 12 A. There would be requests from 13 either investigators or regulatory agencies for 14 data, that information would come through the 15 Lilly medical community, either the physician or 16 the CRA. 17 Q. So for instance the approvable 18 letter would be -- 19 A. That would be an example of 20 indirect communication. 21 Q. It may not be directed towards 22 you, but it would implicate you in some way? 23 A. Right. 24 Q. Would you have copies of those Page 226 1 letters generally, I mean we talked about you 2 having approvable letters in your file, but would 3 you generally have copies of that type of 4 correspondence in your file? 5 A. Currently or at that time? 6 Q. At the time you were working on 7 the drug. 8 A. At that time, yes. 9 Q. Okay. And all of that stuff 10 would have been turned over to the legal 11 department with the rest of your file? 12 A. It I had it in my possession 13 when the legal department asked for it, I would 14 have turned it over. 15 Q. Or it would have been 16 transferred to whoever took over your position 17 when you left? 18 A. I believe so. 19 Q. On how many occasions have you 20 had to turn documents over to the legal 21 department related to Fluoxetine? 22 A. I've done that once. 23 Q. Okay. And do you remember, was 24 that after the project regarding the approvable Page 227 1 letter? 2 A. Yes. 3 Q. Do you recall how long after 4 that they asked you to do that? 5 A. No. It was after I left 6 medical, but I don't remember the exact time 7 frame. 8 Q. And you left medical when? 9 A. I left medical in January of 10 1989. 11 Q. Do you remember if it was in 12 the first year after you left medical? 13 A. I do not remember. 14 Q. Did they ask you to record in 15 any way the types of documents you were turning 16 over or categories of documents you were turning 17 over? 18 A. I don't remember the original 19 request. 20 Q. Did you have to fill out any 21 forms when you transferred the document? 22 A. I don't remember that I did. 23 Q. Do you remember how much you 24 turned over, did it comprise more than a box or Page 228 1 two? 2 A. From what I personally touched 3 and turned over to the legal department would 4 have been less than a box. 5 Q. Okay. It would have been 6 transferred in a box, but didn't fill the whole 7 box? 8 A. I don't remember, I don't 9 believe it would have but I don't remember how 10 much there was. 11 Q. You talked about pulling 12 information from buckets earlier. What did you 13 mean by buckets? 14 A. When Paul was talking earlier 15 about, I believe the subject was, working with 16 other analysts to write coding conventions and I 17 was referring to variables that are used in 18 programs, and used the term buckets as a term to 19 try to explain what a variable was. 20 Q. So bucket is interchangeable 21 with variables? 22 A. Yes, and it's not something 23 that's a generally used systems term, it was an 24 attempt on my part to explain what a variable is. Page 229 1 Q. So a variable is a container, 2 so to speak, for information? 3 A. Yes. 4 Q. And you would have to set up 5 the containers and decide which containers you 6 wanted to pull information from and then figure 7 out how to pull that information from the various 8 containers? 9 A. As it related to the discussion 10 we were having, a variable is a temporary holding 11 place for the information that's in a record 12 during the running of the program. 13 Q. Okay. Refresh my recollection 14 as to what the end result or type of end result 15 would have been if you used the temporary holding 16 places, what would that have been done for? 17 A. What that is is in writing a 18 program or running a program, as the program is 19 processing a record of data, the program needs to 20 keep track of the pieces of data that it has, 21 that's what the variables are as a particular 22 record is being processed. 23 Q. Okay. 24 A. And in the case of producing Page 230 1 one of these reports, I'm writing the record of 2 data. What prints out from that record on the 3 page, there may be, in the case of one of these 4 exhibits that you had -- in the case of Exhibit 5 Number 4, for example, every record would have 6 contained at least these four variables that were 7 being printed out, and I would have said to print 8 each of those four variables that I had been 9 writing the portion that printed this. And in 10 the case of Exhibit 6, I would hate to estimate 11 at this point how many variables there were. But 12 it's basically for a given record, how many 13 pieces of information am I keeping track of on 14 that record, and how do I put that out on a piece 15 of paper or to a file. 16 Q. So if you're going to construct 17 something like Exhibit 4 and you're asking the 18 computer to go to a data base to find the SSI 19 term, okay, once it found that it would 20 temporarily hold that information until you 21 gathered all the other containers, so to speak, 22 and then decided how to put it all together? 23 A. Yes. 24 Q. And for something like Exhibit Page 231 1 6 it could be literally dozen or even hundreds of 2 variables? 3 A. Yes. 4 Q. And even Exhibit 4, like the 5 visit number would be considered a variable? 6 A. Yes. 7 Q. And that's the same thing as 8 standards or is standards something different? 9 A. No. What we were discussing at 10 the time that issue came up was what -- I'm not 11 sure I remember exactly how the question was 12 phrased to me, but what opportunity I might have 13 had to impact how that language was used. The 14 language itself is standard, it's written by a 15 software vendor, we purchased the language from 16 the software vendor. How we choose to use that 17 language within Lilly to make it easier for 18 another analyst to understand is what the 19 standards were for. If I name this -- for 20 example, on Exhibit Number 6, you have HCAE in 21 the upper left side. 22 Q. Under project? 23 A. Right. And how I named the 24 variable that contains that information, I may Page 232 1 have determined that across all clinical trials 2 we would always refer to that as PROJ versus PROJ 3 on one place and project on another place. 4 That's the type of standards I was referring to 5 when we were discussing this earlier. 6 Q. So every time you talk about 7 HCAE, you called it a project or PROJ? 8 A. Each time that is reported 9 within our system, the variable name that would 10 be used for it would be the same. 11 Q. For instance, adverse 12 experience would always be adverse experience, it 13 wouldn't be adverse event, adverse reaction, 14 adverse whatever? 15 A. That would be an example. 16 Q. Okay. Is the ELECT dictionary 17 data base that we're talking about that held the 18 actual dictionary, is that something that was on 19 the System 38 or would that be on the main frame? 20 A. It would have to have been in 21 both places. It would have to be in the System 22 38 to be referenced during data entry, and it 23 would have to be on the main frame to be 24 referenced during reporting. Page 233 1 Q. Okay. How about something like 2 the DEN dictionary, would that be on a System 38 3 or would that be on the main frame? 4 A. I don't know because I don't 5 know anything about the DEN dictionary. 6 Q. Do you do data entry on the 7 main frame? 8 A. It was policy at the time I was 9 in the department to not do data entry on the 10 main frame. 11 Q. So it would be in something -- 12 System 38 or something similar to that where if 13 you were going to do a data entry into DEN, you 14 would work with something outside the main frame? 15 A. I'm not sure I understand the 16 question. 17 Q. I'm trying to figure out if -- 18 I mean I guess it's possible for you to transfer 19 or copy the DEN information into the main frame, 20 correct, from wherever it existed originally? 21 MR. MYERS: Let me object to the form 22 before she answers. She's told you that she 23 knows little or nothing about the DEN system, so 24 I don't know where this is going. Page 234 1 Q. Do you understand the question? 2 A. In general theory, in systems, 3 you can take a file from one system and transfer 4 it to a format on another system. What that 5 would entail with the DEN data base, I don't know 6 because I don't know that data base. 7 Q. You don't have any specific 8 recollection or even general recollection as to 9 whether or not the DEN data base was maintained 10 on a System 38? 11 A. General recollection would be 12 that it was on the System 38 because that's where 13 all the other data entry was done. That's not 14 something I would be able to attest to for 15 certain. 16 Q. I'm sorry I didn't mean to cut 17 you off. 18 A. That's okay. 19 Q. Besides Entrex, are you aware 20 of any other data entry systems that were used 21 besides the System 38 while you were working on 22 Fluoxetine? 23 A. While I was working on 24 Fluoxetine, no. Page 235 1 Q. Do you have any recollection of 2 operating any program for any reason while you 3 were working on Fluoxetine that related to 4 suicidality in any way? 5 A. I don't remember writing any 6 program that would have dealt specifically with 7 suicidality, only programs that would have 8 reported all adverse events in which that would 9 have been one of them. 10 Q. Okay. How about violent 11 aggressive behavior or hostility? 12 A. No. 13 Q. That would be the same thing as 14 suicidality that may be encompassed in adverse 15 events that you wrote other programs regarding? 16 A. That is correct. 17 Q. When you were pulled back into 18 Fluoxetine to help respond to the approvable 19 letter, after that letter was responded to you 20 went back to where you are now working -- or, I'm 21 sorry, you went back to DEA, right? 22 A. Right. 23 Q. Did anybody tell you why they 24 needed you to specifically come back and work on Page 236 1 Fluoxetine for that project? 2 A. Specifically that I had 3 knowledge of the data base, I had worked on it up 4 until about three or four months before that 5 time, and that the knowledge was still fresh 6 enough that I could be of value in responding to 7 the approvable letter. 8 Q. During that two or three month 9 period of time, did you work exclusively on 10 responding to the approvable letter? 11 A. I do not recall whether I got 12 any other requests on my other job or not, I 13 don't know. 14 Q. But unless you had other 15 requests here and there, you would have worked 16 exclusively on responding to the approvable 17 letter during that period of time? 18 A. Yes. 19 Q. Have you ever heard of Tess 20 events, T-E-S-S? 21 A. Yes, I have heard the term. 22 Q. Do you remember what it means? 23 A. No. 24 Q. Do you remember what Tess Page 237 1 events are? 2 A. I don't remember what it means. 3 Q. Is Mrs. Krupa still with Lilly? 4 A. Yes, she is. 5 Q. Were the SSAI and ELECT event 6 terms given number codes? 7 A. The SSAI dictionary, I believe, 8 did have number codes, and the ELECT dictionary, 9 I don't recall whether they did or not, I don't 10 know. 11 Q. So each of the terms that were 12 listed in the dictionary in the SSAI were they 13 entered by numbers? In other words when the 14 information was taken from the CRF and put into 15 the computer, it was put into the computer by 16 number as opposed to the actual written term? 17 MR. MYERS: Before she answers it, let 18 me just object to the form to the extent I think 19 she recalled about number codes for SSAI and she 20 said she didn't know about ELECT, and you're 21 assuming they used them for ELECT. 22 MS. ZETTLER: If I said ELECT, I meant 23 SSAI. 24 MR. MYERS: I think you said both. Page 238 1 A. Can we start the question over? 2 Q. Sure. I'm just trying to get 3 an idea of how -- say I'm in data processing or 4 data entry, okay, and I've got a CRF and on the 5 CRF the doctor has written in depression as an 6 event term under the SSAI. Would I then look at 7 the event term and look it up in the dictionary 8 and see that there was a number code and enter 9 that number code into the computer or would I put 10 depression into the computer? 11 A. I don't remember how it was 12 done. 13 Q. Do you recall retrieving 14 adverse event information when you were still 15 using the SSAI? 16 A. I know I would have done it, 17 but I don't remember any specifics about it. 18 Q. Do you remember if you would do 19 queries, construct queries, using a code number 20 for an adverse event or actual event term? 21 A. That, I don't remember. 22 Q. Earlier when you were talking 23 to Paul, I believe about Exhibit 2 in the audit, 24 I believe you said that it mentions in here that Page 239 1 the errors were corrected, the errors discussed 2 regarding table forty-one. Could you show me 3 where it says that these were corrected? 4 A. At the bottom of the first 5 page, last statement, all problems identified 6 during this audit were corrected before the 7 safety update was submitted to the FDA. 8 Q. Okay. And you take that to 9 mean that all of these problems or these errors 10 that were listed in the following pages were 11 changed or were corrected? 12 A. Yes. 13 MS. ZETTLER: Let's take a quick break. 14 MR. MYERS: Sure. 15 (A SHORT RECESS WAS TAKEN.) 16 Q. I can't remember what you said 17 when Paul asked you when this switch was made 18 generally in the medical community at Lilly from 19 SSAI to ELECT. My understanding from what you 20 said earlier, and if I'm wrong, tell me, is that 21 when you ran the program or wrote the program and 22 ran it to change Fluoxetine from SSAI and ELECT 23 other drugs were already using ELECT? 24 A. Yes, that is true. Page 240 1 Q. There were other drugs besides 2 Fluoxetine that were using the SSAI? 3 A. Yes, that is true. 4 Q. Was there a time when all the 5 drugs were switched to ELECT? 6 A. Yes. 7 Q. Do you remember about when that 8 happened? 9 A. I don't know when that would 10 have been completed. 11 Q. The program that you wrote to 12 switch from SSAI to ELECT, was that used on other 13 drugs besides Fluoxetine? 14 A. Yes. 15 Q. So eventually, because of the 16 program that you wrote, everything was switched? 17 A. I'm not sure if I would make 18 the cause and effect claim, but that may have 19 assisted. 20 Q. Okay. So that was part of the 21 process of changing everything over? 22 A. Yes. 23 Q. And I believe your testimony 24 earlier was that was done for consistency? Page 241 1 A. Consistency across the company. 2 Q. Do you know if SSAI was a Lilly 3 created dictionary or if that was something 4 created by another entity? 5 A. I don't know what the source 6 was. 7 Q. Did you get a 8 first-out-of-the-chute award, have you ever heard 9 of that? 10 A. I don't recall it. 11 Q. Did you get any kind of 12 recognition for the work you did on Prozac? 13 A. There was no official 14 recognition of any type. I know I received 15 electronic messages from both the physician and 16 the CRA that I worked with thanking me for 17 efforts on particular projects, but I don't know 18 of anything official. 19 Q. Did you ever get a certificate 20 or a plaque or anything like that? 21 A. I think there was a 22 certificate. 23 Q. That was in relationship to 24 your work on Fluoxetine or Prozac? Page 242 1 A. I think so. 2 Q. Do you remember if that was 3 before or after the special projects you worked 4 on on the approvable letter? 5 A. I don't know when it would have 6 been. 7 Q. Have you ever heard of a 8 dancing elephant pin? 9 A. Dancing elephant pin, no. It 10 sounds interesting, though. 11 Q. Have you ever heard of a CC 12 check? 13 A. Yes. 14 Q. What is a CC check? 15 A. CC stands for contingent 16 compensation, which is a standard part of 17 compensation for most Lilly employees. 18 Q. Like a bonus? 19 A. Yes. 20 Q. Is that a year-end bonus or is 21 it done by projects? 22 A. No, it's a year-end based on 23 company results. 24 Q. Profit sharing type of thing? Page 243 1 A. I don't know the definition of 2 profit sharing in this context. Very loosely, 3 yes, it's a way of profit sharing. 4 Q. Is there any sort of incentive 5 program involved? In other words is it a 6 merit-based type of thing? 7 A. No. 8 Q. What time of year do you 9 usually get the CC checks, is it around Christmas 10 time? 11 A. It's received after the annual 12 results have been tabulated. 13 Q. Annual results of what? 14 A. Of the company, the income, 15 sales results, basically the annual report. 16 Q. What time of year is that? 17 A. It's usually in February. 18 Q. Okay. 19 A. I think it's always in 20 February. 21 Q. Is that a consistent amount 22 throughout the years, has that been a consistent 23 amount or has it fluctuated? 24 A. What time frame are you talking Page 244 1 about? 2 Q. Have you consistently gotten a 3 CC check throughout the years that you've worked 4 for Lilly? 5 A. I have gotten a CC check every 6 year. 7 Q. Has the amount been consistent 8 or has it gradually increased or decreased in 9 various years? 10 A. It has, as far as I can 11 remember, always increased based on increasing 12 sales and revenues. 13 Q. Okay. Was there ever a time 14 where it increased substantially? 15 A. I don't know what you mean by 16 substantially. 17 Q. Where you have gotten your 18 check and you've thought to yourself, wow, this 19 is pretty big compared to last year. 20 A. Oh, my first one out of 21 college, but then every check I got was big. 22 They were all consistent with what people who had 23 been at the company for years had explained was a 24 general trend or general pattern. Page 245 1 Q. Is it usually a percentage 2 increase? 3 A. A percentage point a year or 4 so. 5 Q. Does it increase depending on 6 your position with the company? In other words 7 were people who are higher than you on the food 8 chain get a larger CC check? 9 A. The CC is calculated based on 10 base salary. 11 Q. Is it your understanding that 12 Prozac was approved for use in the treatment of 13 depression in late 1987? 14 A. Yes. 15 Q. After that year, 1987, did any 16 of your CC checks increase higher than years 17 before? In other words, a larger percentage? 18 A. I don't remember what the 19 percentage increase was from year to year, so I 20 don't think I can answer the question. 21 Q. Did you receive a monetary 22 bonus for your work on Fluoxetine? 23 A. No. 24 Q. Did you receive a bonus in the Page 246 1 form of anything material, like a car or anything 2 of that nature? 3 A. No. 4 Q. Earlier you were talking, I 5 think, when you were looking at Exhibit 2, and 6 you were talking with Paul about the -- on page 7 four, about halfway down, where it says table 8 forty-one lists depression and self-inflicted 9 injury as adverse events causing termination. 10 CRF lists both suicidal thoughts and 11 self-inflicted injury. Suicidal thoughts is 12 mapped to depression in the SSAI dictionary. Was 13 it your understanding that that was Lilly's 14 policy to map suicidal thoughts to depression? 15 A. Can you explain what you mean 16 by Lilly's policy? 17 Q. Sure. Other than looking at 18 the SSAI dictionary and seeing suicidal thoughts 19 and seeing it mapped to depression, were you ever 20 told that that was the way that Lilly wanted it 21 mapped? 22 A. It's my understanding that any 23 of the mappings in those dictionaries were 24 medical decisions. Page 247 1 Q. When you say medical decisions, 2 you mean the medical research physicians at 3 Lilly? 4 A. Yes. 5 Q. Outside the dictionary itself, 6 were you ever given anything in writing that said 7 that that is what they wanted, suicide thoughts 8 to map to depression? 9 A. I don't remember that there 10 would have been anything of that nature. 11 Q. Earlier when we were talking 12 about computer edits that were done, the batch 13 edits for variables or things like vital signs, 14 type information, temperature, blood pressure, 15 things of that nature, was there an edit set up 16 to kick out any adverse event terms that weren't 17 found in the SSAI dictionary? So in other words 18 if somebody had written down or entered into the 19 computer suicidal thoughts, would the computer 20 flag that so you could figure out what was going 21 on? 22 A. I don't know whether there was 23 anything like that or not. Again, that gets back 24 to I'm not sure whether a number was entered or a Page 248 1 term was entered, so I don't know how it would 2 have been edited. 3 Q. So if a number was to be 4 entered, it would be a matter of somebody down 5 the line physically looking up the number and 6 writing that in so that data entry people could 7 enter that number? 8 A. I don't know because I don't 9 know how it was entered. 10 Q. Were you responsible for doing 11 any systems work with regards to articles that 12 were published by Lilly? 13 A. I think I might have been, I 14 can't recall any specific instances. 15 Q. If you were to do that type of 16 work or to the best of your recollection, would 17 those requests have come from somebody at Lilly 18 or would they have come from clinical 19 investigators? 20 A. Those would have come from the 21 physician at Lilly. 22 Q. The research physician? 23 A. The research physician. 24 Q. Do you have any recollection, Page 249 1 generally, of fulfilling a request by a clinical 2 investigator for data analysis or reporting with 3 regards to Fluoxetine? 4 A. I don't remember any specific 5 cases. I believe I mentioned before that at the 6 conclusion of a study, it was practice to report 7 the data from that study as a whole and then to 8 break it down by investigator so that those 9 reports could be sent back to the investigators. 10 So they had a summary of what their patients' 11 data was. 12 Q. Because earlier you said that 13 you, with regards to correspondence coming from 14 outside of Lilly, that you may have been charged 15 with responding to investigator questions. Can 16 you give me some examples of what types of 17 information they would have been asking for? 18 A. Any of that would have gone 19 through the Lilly research physician in both 20 directions, so there wouldn't have been -- as far 21 as I know, there was no direct communication with 22 an outside investigator. I don't recall whether 23 there were any specific analysis requests made by 24 any specific investigators. Page 250 1 Q. Okay. 2 A. Nothing comes to mind, I can't 3 remember anything -- 4 Q. So -- 5 A. -- in particular. 6 Q. So it would have been a matter 7 of the research physician coming to you and 8 saying Investigator A wants this information, and 9 then you set up a program to retrieve it and get 10 it back to the research physician? 11 A. Yes. 12 Q. Who decides what different 13 variables are going to be entered into the data 14 bases? 15 A. From the clinical report form? 16 Q. Right. 17 A. That would be all data from the 18 clinical report form -- anything that's recorded 19 on the clinical report form is entered into the 20 data base. That's, as far as I know, a policy. 21 Q. Who decides what information 22 will be recorded on the clinical report form? 23 A. That would be -- when I was 24 involved with that, it was a team of people. It Page 251 1 would have been a systems analyst involved, a 2 CRA, a physician, I don't recall whether or not 3 statisticians were involved in that activity. 4 Q. Would that be done on a project 5 by project basis? 6 A. I believe so, yes. 7 Q. So depending on the study that 8 was being run, you could use information or 9 clinical report forms that had already been 10 constructed with some minor modifications or some 11 extensive modifications? 12 A. That is correct. 13 Q. Who would decide what to call 14 the variable? 15 A. In what part, in what files? 16 Q. Let's take, for instance, 17 something like Exhibit 4, where we talked about 18 the drug name, the PIP visit number, SSI terms. 19 Who would decide to call the PIP the PIP? 20 A. The way it is referenced in the 21 programs, for the most part, is at the discretion 22 of the analyst, based on, as I mentioned before, 23 some coding standards and coding guidelines. How 24 that appears and how that's titled on a report Page 252 1 was usually done with the systems analyst talking 2 to the person who made the original request. 3 Q. So, for instance, the clinical 4 research physician? 5 A. Yes. 6 Q. Who decided on the standards? 7 A. The standards are coding 8 guidelines that were written, were written by a 9 number of analysts. 10 Q. At Lilly? 11 A. At Lilly who were working in 12 the medical area at the time. 13 Q. Is S-A-S a medical specific 14 software? 15 A. No. 16 Q. Other than the safety updates, 17 did you work on any of the final reports from the 18 clinical trials? 19 A. Yes. 20 Q. Do you have any specific 21 recollection of any of the clinical trials you 22 worked on, does anything stick out in your mind? 23 A. None in particular, no. 24 Q. Would your responsibilities Page 253 1 with regards to the trial, the final reports from 2 the clinical trials, be the same as the safety 3 updates and the articles, you would be given 4 queries and -- you would be given requests for 5 information and you would construct a query to 6 gather that information? 7 A. Yes. 8 Q. Did you work with 9 statisticians? 10 A. Yes. 11 Q. Did they give you requests? In 12 other words, if you want to know what percentage 13 of people in Study A suffered the adverse event 14 of depression. 15 A. I don't remember that type of 16 request coming from statisticians. 17 Q. What types of requests came 18 from statisticians? 19 A. Most of the requests from 20 statisticians were to assist them in 21 understanding the data bases or formatting the 22 data bases so that they could use the data base 23 for statistical analysis. 24 Q. Did they give you the Page 254 1 information regarding the analysis that they 2 wanted to run and have you enter that or 3 construct a program to run a specific analysis? 4 A. I don't remember that it was 5 ever done that way, I can't remember any cases 6 where that would have been handled that way. 7 Q. Yet, like earlier, you said 8 like you're not a statistician; correct? 9 A. Yes. 10 Q. So if they wanted to find out -- 11 if they wanted to run a complicated statistical 12 analysis of information that was in the data 13 base, like for instance comparing Ham-D-3 scores 14 of people who were on Fluoxetine in a study with 15 Ham-D-3 scores of people who were on a comparitor 16 drug than Fluoxetine, you wouldn't know how to 17 set up that request, would you? 18 A. It depended on what the 19 comparison was. If they were trying to do a 20 statistical analysis of the differences including 21 E values and so forth, I would not know how to do 22 that request. 23 Q. How would you do that with the 24 S-A-S then, would somebody give you that Page 255 1 information to construct a program with or would 2 they write the program? 3 A. In a situation like that where 4 a statistical analysis and decision was being 5 made, the statistician would write the program. 6 Q. Were there statisticians who 7 were also system analysts, would they come to you 8 for advice on how to write the program and set it 9 up a certain way? 10 MR. MYERS: There were two questions, 11 weren't there. Wasn't one were there 12 statisticians who were also analysts? 13 MS. ZETTLER: Yes. 14 A. The systems analyst's role was 15 really more of a job classification. In general, 16 one who was doing -- was a statistician, was a 17 statistician, and someone with my educational 18 background or experience level was a systems 19 analyst. 20 Q. Okay. Let me ask you this: 21 Were there statisticians with enough experience 22 and knowledge of the computers and the data bases 23 to be able to construct or write a program to be 24 able to do a statistical analysis of information Page 256 1 using the S-A-S? 2 A. Yes. 3 Q. Who would that be? 4 A. Basically any of the 5 statisticians in the department, to my knowledge, 6 had that knowledge and capability. 7 Q. They would use the S-A-S 8 software to do that, though? 9 A. Yes. 10 Q. How about Greg Enas, to your 11 knowledge would he have that kind of experience? 12 A. To the best of my knowledge, he 13 would. 14 Q. The DB2 data base management 15 system, is that what it's called, a data base 16 management system? 17 A. Yes, referred to as DB2. 18 Q. What role does this play in the 19 System 38 and the main frame and all that other 20 stuff, how does that work into the whole picture? 21 A. That was not in the picture 22 when I was working on Prozac. 23 Q. When did that come into the 24 picture? Page 257 1 A. I don't know. I know it was 2 installed in the main frame at Lilly while I was 3 still in medical, I don't remember exactly when. 4 Q. Do you know what the purpose of 5 DB2 is? 6 A. It's a way of storing the data, 7 it's an organization for storing the data. 8 Q. Okay. So data that is copied 9 into the main frame is stored now using the DB2? 10 A. I don't know. 11 Q. But is that how it can be used? 12 A. It can be used that way, yes. 13 Q. When we were talking earlier 14 about the data base files, and then there were 15 subfiles underneath that, like in other words you 16 have the CRF and there were some files that 17 contained the demographic data like age and 18 things like that? 19 A. Yes. 20 Q. Were those data bases and 21 subfiles created at the time the CRF was created 22 or was that something that was created after all 23 the data was entered into the system? 24 A. On the main frame? Page 258 1 Q. System 38. 2 A. On the System 38 those files 3 had to be in place before data could be entered 4 from the case report form or clinical report 5 form. 6 Q. Okay. When you constructed 7 that computer case report form, for lack of a 8 better phrase, would that be considered -- would 9 you actually have on the screen, for instance, 10 patient name, age, you know, date of birth, 11 whatever, would you actually have a -- could you 12 look on the screen and take the patient name from 13 the hard copy and type it into a specific area? 14 In other words was there a space left for that 15 said patient name and a space so you type it in? 16 A. As far as I can recall, data 17 entry screens were designed to look as much as 18 possible like the clinical report form, and that 19 would include text on the screen that would tell 20 the person who is keying what belongs in that 21 field, and that might be the patient number or 22 the patient date of birth or that type of thing. 23 Q. Would each of those be 24 considered -- if you're looking at the screen and Page 259 1 you see a place for a patient date of birth, 2 would that be considered a file? 3 A. No. 4 Q. Okay. It's the information 5 after it's entered into the computer that's 6 separated into a separate file? 7 A. As it's being entered, data 8 would go into separate files. Something like the 9 date of birth would be a field within that file. 10 Q. Okay. So you may have a 11 demographic file that would include the date of 12 birth, the age, the location, things of that 13 nature? 14 A. That sort of information. 15 Q. Okay. So that screen 16 containing the demographic data would be a screen 17 related to a specific file, if it just contained 18 the demographic data? 19 A. I don't recall how that was set 20 up, I don't remember. 21 Q. Are you familiar with Word 22 Perfect? I'm getting it down to my level now. 23 A. No. I use another word 24 processing package, though, and there may be Page 260 1 similarities. 2 Q. Well, let me ask you this: I'm 3 trying to get a visual idea of if I were to look 4 at a file on the computer without going into the 5 file, I would be able to recognize that there was 6 a file, okay. So, in WordPerfect you have 7 different directories, so if I wanted to start a 8 directory of my particular cases, I can name a 9 Case A directory. And under that directory I may 10 have an indication that I have a couple of 11 letters and each letter would have its own name. 12 Would that be something similiar to what we're 13 talking about here? 14 A. Not really. The file structure 15 on a System 38 is very different from what you 16 would see on a PC. It's certainly not as visual 17 as what you would have on a PC or what you could 18 bring up within Word or WordPerfect or a PC 19 package. 20 Q. Would it be done more codes and 21 things of that nature? 22 A. Files. I don't know that 23 there's a subdirectory type of structure on the 24 System 38. It's -- I know there's not on the Page 261 1 main frame, it's a little different than that. 2 Q. If I wanted to look at all 3 information from study HCCD, for instance, how 4 would I go about pulling up that information or 5 would I have to print it all out in hard form? 6 A. How would you pull up that 7 information? 8 Q. Like how would I be able to get 9 that information out of the System 38 if I wanted 10 to look at -- say, it's a small study, there's 11 only six patients or something like that, and I 12 wanted to get all that information on that 13 certain study, how would I go about doing that? 14 A. Along with the ability to enter 15 data on the System 38, it's possible to bring 16 that data up on screens that would look like the 17 same way it was entered on the System 38. 18 Q. So on the computer CRF you 19 could actually see? 20 A. You could bring those back up, 21 yes. 22 Q. Could you print them out in 23 hard copy? 24 A. I honestly do not remember Page 262 1 whether there was even a screen print capability 2 at that time, I don't know. We did, on some 3 studies, have programs that produced reports that 4 just looked like the screen with the data, they 5 didn't have any editing value in them or 6 anything, they just printed the data out. 7 Q. Is it your understanding that 8 after the information from the CRFs were 9 collected and edited and verified, that the data 10 bases were locked at a certain point so that 11 nobody could get in there without somebody else's 12 knowledge and change anything around? 13 A. I remember there being a 14 procedure for locking the data base, I don't 15 remember what the procedure was and whether it 16 related to the System 38 or the main frame. 17 That's something that I don't remember that 18 anymore. 19 Q. Would you have had the type of 20 clearance, for lack of a better word, to be able 21 to go in and look at that information without 22 having to have it unlocked, so to speak? 23 A. I think once data was locked, 24 whatever that meant, it was still available for Page 263 1 viewing on the System 38. 2 Q. Okay. You couldn't make any 3 changes, though, without -- 4 A. Again, I don't know because I 5 don't know what locking -- what the impact of 6 that was or the procedures for it. 7 MS. ZETTLER: Until we resolve the 8 issue of the documents, I don't have anything 9 further. 10 MR. CLEMENTI: Let me ask a question or 11 two. It shouldn't take too long. 12 * * * * * * * * * * 13 CROSS EXAMINATION 14 BY MR. CLEMENTI: 15 Q. My name is Paul Clementi and I 16 represent some doctors in a case in Chicago and 17 some other cases, but that's not important here. 18 I want to ask you a couple of questions. 19 MR. SMITH: That's not important for 20 her to -- 21 MR. CLEMENTI: That's not important for 22 her to know. 23 Q. Do you know what a Dear Doctor 24 letter is, have you ever heard that term before? Page 264 1 A. No. 2 Q. Do you have any involvement -- 3 do you know what a package insert is, let me ask 4 you that first? 5 A. Yes. 6 Q. Do you have any involvement in 7 writing package inserts? 8 A. No. 9 Q. Did you have any involvement in 10 giving data to people who were writing package 11 inserts, in other words assisting in some way in 12 gathering information that would end up on a 13 package insert for a product? 14 A. Do I now or did I ever? 15 Q. Did you ever regarding Prozac? 16 A. Yes. 17 Q. Okay. And when you did that, 18 was there one person who was in charge of working 19 on these package inserts for Prozac? 20 A. I don't know. The request 21 would have come through the physician or CRA, I'm 22 not sure I know who would have been writing the 23 package inserts. 24 Q. You don't remember specifically Page 265 1 anyone's name who asked you for information to 2 assist on that, do you? 3 A. No. 4 Q. Have you ever spoken to any 5 non-Eli Lilly doctor who was prescribing Prozac 6 and called you up and asked you for information -- 7 or you said that a series of queries came in, and 8 I want to know if any of that type of document 9 would have gone to you? 10 A. No, I never received requests 11 directly from anyone outside Lilly for 12 information. 13 Q. If a request came to -- has any 14 requests ever come to you through someone else 15 and then that information had to get back to that 16 person that you know of? 17 A. Not that I know of, I don't 18 know of any cases where that would have happened. 19 Q. And you're not a statistician, 20 right? 21 A. That's right. 22 Q. You would take data and work 23 with it within the system, you wouldn't decide 24 what P values were or what variables were, et Page 266 1 cetera; is that correct? 2 A. I'm not sure I know the 3 question. 4 Q. Would you ever decide for a 5 certain -- let me phrase this another way. Did 6 anyone ever ask you, while working on questions 7 with Fluoxetine, to look into statistical 8 significance of different events that occur while 9 people are on Fluoxetine? 10 A. No. 11 Q. While working with Fluoxetine 12 or any other drug, were you ever the person who 13 would decide what the search would be to get 14 certain data or would a search come to you? In 15 other words, would someone say to you please find 16 this out for me, and then you would run the 17 search, or would you say I'm going to run the 18 search on this drug? 19 MR. MYERS: Are you asking her is she 20 the one that formulated the issue or question? 21 MR. CLEMENTI: Right. 22 A. No, I did not. 23 Q. But at some point someone 24 formulated an issue or question? Page 267 1 A. Yes. 2 MR. CLEMENT: That's all I have. 3 MS. ZETTLER: I have a couple more. 4 * * * * * * * * * * 5 RECROSS EXAMINATION 6 BY MR. ZETTLER: 7 Q. To your knowledge, was there 8 ever a project that was done where information 9 from the main frame to include with prescribing 10 information where things were to be sent to 11 doctors? 12 MR. MYERS: Are you asking if she ever 13 did such a project? 14 MS. ZETTLER: Right. 15 A. I don't recall doing anything 16 like that. 17 Q. Okay. With regards to the 18 package inserts, what type of responsibility did 19 you have? 20 A. My responsibility would have 21 been to report occurrences of adverse events, 22 number and percent of occurrences, basically. 23 Q. Was that -- did you do that on 24 more than one occasion? Page 268 1 A. I did that type of request 2 frequently. Are you asking whether I did that 3 more than once with a package insert? 4 Q. Right. 5 A. Not to my knowledge. 6 Q. Was it your understanding that 7 the package insert has been changed on a number 8 of occasions? 9 A. I have no idea, I don't know 10 anything about that. 11 Q. Was that the same type of 12 information that was accumulated for the safety 13 update? 14 A. Yes, it would have been the 15 same kind of information. 16 Q. To your knowledge, was the 17 information that was used in a safety update the 18 same information that was used, not generally, 19 but specifically the same information that was 20 then used in the package insert? 21 A. I do not recall whether the 22 safety update data was used in the package 23 insert. 24 MS. ZETTLER: That's all I have. Page 269 1 MR. SMITH: Let me ask a couple of 2 questions. 3 * * * * * * * * * * 4 REDIRECT EXAMINATION 5 BY MR. SMITH: 6 Q. When you came to Lilly, did 7 they give you any training in psychiatric 8 terminology or psychological terms or anything of 9 that nature? 10 A. No. 11 Q. There are a lot of terms, 12 psychiatric terms, that have been used in 13 connection with Prozac since it's an 14 anti-depressant; correct? 15 A. Yes. 16 Q. How did you pick up the meaning 17 of those various terms? 18 A. If it were necessary for me to 19 understand the meaning of a term, I would go to 20 the physician, the research physician, and ask 21 for help in understanding the term. 22 Q. Would there be occasions when 23 you would do that? 24 A. I can't think of any specific Page 270 1 cases where I needed to make any judgment based 2 on the term where that would have been necessary. 3 Q. In connection with designing a 4 program in response to a new protocol, I would 5 assume that you would want to have a pretty good 6 working knowledge of the protocol? 7 A. Yes. 8 Q. And would you go talk with the 9 research physicians or the medical monitor 10 concerning that particular protocol -- 11 A. Yes. 12 Q. -- to get some input from him. 13 What's the name of some of the research 14 physicians or medical monitors you sought 15 assistance from? 16 A. Joe Wernicke would have been 17 the research physician I worked with during that 18 time. 19 Q. Did you ever discuss any of 20 this information or any of these protocols with 21 Dr. Beasley or Dr. Heilegenstein? 22 A. I believe I did meet with Dr. 23 Heilegenstein. I don't remember whether it would 24 have been in reference to specific protocols or -- Page 271 1 I know I would have met with and worked with Dr. 2 Heilegenstein, but I don't know the specific 3 instances. 4 Q. Generally, what would be the 5 occasion for those meetings? 6 A. Those meetings would come from 7 requests for reports as we described earlier, 8 where the physician, the research physician, 9 would make a request for some type of report. I 10 was involved in working on reviews of protocols 11 and clinical report forms. We would have had 12 occasion to work together in those instances. 13 Q. Am I dreaming this or did I see 14 your name in connection with Dr. Beasley's 15 meta-analysis paper? 16 MR. MYERS: You mean as an author? 17 A. I have no idea. 18 Q. As an individual providing 19 assistance from a systems analysis standpoint? 20 A. I have no idea, I haven't seen 21 any document with my name on it in reference to 22 that. 23 Q. Are you familiar with Dr. 24 Beasley's meta-analysis? Page 272 1 A. No, I'm not. 2 Q. Do you know whether or not he 3 used any of your work in providing either data or 4 statistics with respect to that piece? 5 A. No, I do not know. 6 Q. Were there any on-site, 7 on-the-job training facilities at all at Lilly in 8 connection with psychiatric terms or psychology 9 or how to operate a clinical trial? 10 MR. MYERS: For systems people or for 11 anybody? 12 MR. SMITH: Or for anybody, that she 13 attended. 14 A. At the time I was working in 15 medical, there were no courses or classes 16 specifically in medical psychology or medical 17 psychology terminology. There was one seminar 18 provided by a CRA to give systems people a base 19 understanding of medical terminology, in general, 20 in the case of something like otitis media, 21 breaking down the term to understand where the 22 term came from, but there was nothing 23 specifically related to psychiatry or psychology. 24 Q. Did you take any psychology at Page 273 1 Iowa State? 2 A. I took one psychology course at 3 Iowa State. 4 MR. SMITH: That's all I have. 5 MS. ZETTLER: Nothing further. 6 MR. CLEMENTI: Nothing further. 7 MR. MYERS: No questions. 8 (THE WITNESS WAS EXCUSED.) Page 274 1 COMMONWEALTH OF KENTUCKY ) 2 : ss COUNTY OF JEFFERSON ) 3 4 I, MARY KATHLEEN NOLD, A NOTARY PUBLIC IN 5 AND FOR THE STATE OF KENTUCKY AT LARGE, DO HEREBY 6 CERTIFY THAT THE FOREGOING TESTIMONY OF 7 LORI SHELDON 8 WAS TAKEN BEFORE ME AT THE TIME AND PLACE AS 9 STATED IN THE CAPTION; THAT THE WITNESS WAS FIRST 10 DULY SWORN TO TELL THE TRUTH, THE WHOLE TRUTH, 11 AND NOTHING BUT THE TRUTH; THAT THE SAID 12 PROCEEDINGS WERE TAKEN DOWN BY ME IN STENOGRAPHIC 13 NOTES AND AFTERWARDS TRANSCRIBED UNDER MY 14 DIRECTION; THAT IT IS A TRUE, COMPLETE AND 15 CORRECT TRANSCRIPT OF THE SAID PROCEEDINGS SO 16 HAD; THAT THE APPEARANCES WERE AS STATED IN THE 17 CAPTION. 18 WITNESS MY SIGNATURE THIS THE 19TH DAY OF 19 SEPTEMBER, 1993. 20 MY COMMISSION EXPIRES MARCH 10, 1994. 21 22 23 _________________________ MARY KATHLEEN NOLD 24 COURT REPORTER AND NOTARY PUBLIC STATE OF KENTUCKY AT LARGE Page 275 1 E R R A T A S H E E T 2 3 COMMONWEALTH OF KENTUCKY ) : SS 4 COUNTY OF JEFFERSON ) 5 6 I, LORI SHELDON, THE UNDERSIGNED 7 DEPONENT, HAVE THIS DATE READ THE FOREGOING PAGES 8 OF MY DEPOSITION AND WITH THE CHANGES NOTED 9 BELOW, IF ANY, THESE PAGES CONSTITUTE A TRUE AND 10 ACCURATE TRANSCRIPTION OF MY DEPOSITION GIVEN ON 11 THE SEPTEMBER 15, 1993 AT THE TIME AND PLACE 12 STATED THEREIN. 13 PAGE NO. LINE NO. CHANGE REASON Page 276 1 2 PAGE NO. LINE NO. CHANGE REASON 3 4 5 6 7 8 9 _____________________________ 10 LORI SHELDON 11 SWORN TO AND SUBSCRIBED BEFORE ME THIS 12 _____ DAY OF __________, 1992. 13 _____________________________ NOTARY PUBLIC, STATE OF 14 KENTUCKY AT LARGE Page 277 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Page 278 1 DIRECT EXAMINATION BY MR. SMITH:....................2 2 CROSS EXAMINATION BY MS. ZETTLER:.................105 3 CROSS EXAMINATION BY MR. CLEMENTI:................207 4 RECROSS EXAMINATION BY MR. ZETTLER:...............210 5 REDIRECT EXAMINATION BY MR. SMITH:................212 6 COMMONWEALTH.....................................216 7 PLAINTIFFS' EXHIBIT NO. 1 ........................74 8 PLAINTIFFS' EXHIBIT NO. 2.........................78 9 PLAINTIFFS' EXHIBIT NO. 3.........................94 10 PLAINTIFFS' EXHIBIT NO. 4........................153 11 PLAINTIFFS' EXHIBIT NO. 5........................167 12 PLAINTIFFS' EXHIBIT NO. 6........................172 13 14 15 16 17 18 Page 279 1 COMES LORI SHELDON................................14 2 DIRECT EXAMINATION BY MR. SMITH:...................15 3 CROSS EXAMINATION BY MS. ZETTLER:.................141 4 CROSS EXAMINATION BY MR. CLEMENTI:................264 5 RECROSS EXAMINATION BY MR. ZETTLER:...............268 6 REDIRECT EXAMINATION BY MR. SMITH:................270 7 COMMONWEALTH.....................................275 8 (QUESTION CERTIFIED.).............................32 9 PLAINTIFFS' EXHIBIT NO. 1 .......................103 10 PLAINTIFFS' EXHIBIT NO. 2........................109 11 PLAINTIFFS' EXHIBIT NO. 3........................127 12 PLAINTIFFS' EXHIBIT NO. 4........................198 13 PLAINTIFFS' EXHIBIT NO. 5........................216 14 PLAINTIFFS' EXHIBIT NO. 6........................222 15 16 17 18 Page 280