Pharmacoepidemiology

(Epidemiology 221b)

Instructor

Alec Walker Kresge 908 432-4565

Meetings

Kresge 201

Tuesdays and Thursdays 3:30 – 5:20

(No class December 17 and January 21 – HSPH Faculty Meeting)

Office hours Wednesdays 2:30 – 4:00 and by appointment

Purpose

The goal of this course is to introduce students to the most important issues of pharmacoepidemiology. To this end, we will emphasize the ways in which the observational study of drugs can draw on standard epidemiologic technique, and explore the ways in which drugs present unique research problems and opportunities.

Structure

Material will be taught through seminars, case studies, and group projects. Readings must be done in advance of the class.

Grading

There will be a term paper, to be done in teams. (60%)

Students will chosen in class to evaluate assigned papers for that class session. (10%)

Students will be expected to participate in class discussions. (30%)

Each activity is graded on a four-point scale, corresponding roughly to

1) work that would be fully acceptable from a professional colleague

2) work that shows evidence of a colleague-in-training

3) work that has some merit but would be insufficient for ordinary scientific interchange

4) unacceptable or incomplete work.

In the past, most students have performed at level 2 above. The grade for level 2 performance is a B+.

Term Papers

The term paper should be about 2000 words, and should cover some narrow topic in drug epidemiology. Students may co-author papers, in which case the expected length is 1500 words per author. Any class topic is acceptable for development into a term paper. A historical review of the development of evidence for an adverse effect of a drug or medical device is a very good topic. A straight current-evidence review would be acceptable if there is some interesting complication or uncertainty. If you can see applications of methodologic work that you are doing elsewhere, you may wish to consider a methods topic, but this must reflect a substantial adaptation of the methods and ideas to the special needs of pharmacoepidemiology. Each year one or two of the term papers makes its way into the published literature, and one or two become the basis of thesis proposal.

Staying in Touch

You can see some of the pharmacoepidemiology program's activities, and read some previous student papers at the Drugs and Devices Information Line website. You can get there by entering "Harvard Pharmacoepidemiology" on your web search engine, or by entering the following URL http://www.hsph.harvard.edu/Organizations/DDIL/ddilhpge.html . Papers are available under "Harvard Reviews in Pharmacoepidemiology." DDIL also provides links to the home pages of a number of regulatory authorities and research groups. You should make an effort to follow the links, so that you can get a sense of the "real" world of drug safety.

 

My e-mail address is amwalker@epi.harvard.edu .

Schedule

Date

Topic

November 10

Introduction: From Events to Epidemiology in Drug Safety

12

Spontaneous Reports

17

Population Case Series

19

Bayesian and Recognition-Primed Causality Assessments

24

Quantifying Risk and Benefit by National Surveillance.

December 1

Follow-up studies. Confounding by Indication.

3

Follow-up studies with sampling

8

Case-Control Studies

10

Determinants of Prescribing

15

Duration of Therapy

17

No Class

January 5

Large linked data resources. Defining exposure

7

Large linked data resources. Defining outcomes

12

Roles for epidemiologists in the pharmaceutical industry

14

The tension between clinical trialists and observational investigators

19

Term papers due

Denial: Some problems of institutional recognition of adverse drug effects in government and industry.

21

No Class

Class Synopsis

Introduction: From Events to Epidemiology in Drug Safety

November 10

Purpose

Provide an epidemiologist’s overview of the data and standard analyses in drug safety studies.

Readings

Feenstra J, van Drie-Pierik RJHM, Lacié CF, Stricker BHCh. Acute myocardial infarction associated with sildenafil. Lancet 1998;352:957-8

Assal F, Spahr L, Hadengue A, Rubbici-Brandt L, Burkhard PR. Tolcapone and fulminant hepatitis. Lancet 1998;352:958

Spontaneous Reports

November 12

Purpose

Spontaneous reports to regulatory authorities are the backbone of safety surveillance. We will review the structure and challenges of these programs.

Reading

Baum C, Kweder S, Anello C. The spontaneous reporting system in the United States. Chapter 10 of B. Strom, ed., Pharmacoepidemiology (Second Edition)

Wiholm BE, Olsson S, Moore N, Wood S. Spontaneous reporting systems outside the United States. Chapter 11 of Strom, op cit.

Lippman JS, Shangold GA. A review of post-marketing safety and surveillance data for oral contraceptives containing norgestimate and ethinyl estradiol. Int J Fertil 1997;42:230-239

Population Case Series

November 17

Purpose

Differentiate between selected and unselected case series. Review case-crossover design as a model for analysis of unselected case series.

Readings

Maclure M. The case-crossover design: A method for studying transient effects on the risk of acute events. Am J Epidemiol 1991;133:144-53

Connolly HM, Crary JL, McGoon MD, Hensrud DD, Edwards BS, Edwards WD, Schaff HV. Valvular heart disease associated with fenfluramine-phenteramine. N Engl J Med 1997;337:581-8

Adams PC, Robinson A, Reid MM, Vishu MC, Livingston M. Blood dyscrasias and mianserin. Postgrad Med J 1983;59:31-3

Bayesian and Recognition-Primed Causality Assessment

November 19

Purpose

To understand Bayesian inference both as a technique for allocating cause in individual cases and as a paradigm for inference.

To examine more naturalistic approaches to causality assessment.

Readings

Lane, DA. Causality assessment for unintended drug effects. Chapter 6 in A.G. Hartzema, M.S. Porta and H.H. Tilson (Eds.) Pharmacoepidemiology: An Introduction Harvey Whitney Books Cincinnati 1991

Klein G. The Power of Intuition. From Sources of Power: How People Make Decisions. The MIT Press, Cambridge, 1998, pp. 31-44

Quantifying Risk and Benefit by Means of National Surveillance

November 24

Purpose

A few drug effects that can be assessed and quantified by national surveillance. There are however many pitfalls.

Readings

Coulter DM, Edwards IR. Mianserin and agranulocytosis in New Zealand. Lancet 1990; 336:785-7

Lindberg G, Bingegfors K, Ranstam J, Råstam L, Melander A. Use of calcium channel blockers and risk of suicide: ecological findings confirmed in population based cohort study. BMJ 1998;316:741-5

Follow-up Studies. Confounding by Indication.

December 1

Purpose

The association between cimetidine and gastric cancer has been the object of a long-standing discussion, one that has only recently been resolved.

Readings

Colin-Jones DG, Langman MJS, Lawson DH, Vessey MP. Cimetidine and gastric cancer: preliminary report from post-marketing surveillance study. Brit Med J 1982;285:1311-1313

Colin-Jones DG, Langman MJS, Lawson DH, Vessey MP. Postmarketing surveillance of the safety of cimetidine: mortality during the second third and fourth years of follow up. Brit Med J 1985;291:1084-88

Johnson AG, Jick SS, Perera DR, Jick H. Histamine-2 receptor antagonists and gastric cancer. epidemiology 1996;7:434-436

Walker AM. Confounding by indication. Epidemiology 1996;7:335-336

Follow-up Studies with Sampling

December 3

Purpose

In order to define confounding factors and time-varying factors in a cohort study it may be necessary to sample the population.

Readings

Jick H, Vasalakis C, Weinrauch LA, Meier CR, Jick SS, Derby LE. A population-based study of appetite suppressant drugs and the risk of cardiac-valve regurgitation. N Engl J Med 1998;339:719-24

Baird IM. Risks of heart-valve abnormalities with appetite suppressants. Lancet 1998;353:1403-4

Case-Control Studies

December 8

Purpose

To appreciate the role of definition of study subjects in the assessment of causal associations in non-randomized investigations

Readings

WHO Collaborative Study of Cardiovascular Disease and Steroid Hormone Contraception. Effects of different progestagens in low oestrogen oral contraceptives on venous thromboembolic disease. Lancet 1995;346:1582-1588

Walker AM. Newer oral contraceptives and the risk of venous thromboembolism. Contraception 1998;57:169-181

Determinants of Prescribing

December 10

Purpose

Examine the ways in which prescribing practices may affect the population composition of users, and the methods for assessing the impact of these on observational studies.

Readings

Dunn N, White I, Freemantle S, Mann R. The role of prescribing and referral bias in studies of the association between third generation oral contraceptives and increased risk of thromboembolism. Pharmacoepidemiol Drug Safety 1998;7:3-14

Lidegaard Ø. The influence of thrombotic risk factors when oral contraceptives are prescribed. Acta Obstet Gynecol Scand 1997;76:252-260

Duration of Therapy

December 15

Purpose

Risks in a cohort defined by current use of a drug may change with increasing duration of use. These changes could be etiologic, or they may reflect changing cohort composition.

Readings

Lewis MA, Heinemann LAJ, MacRae KD, Bruppacher R, Spitzer WO. The increased risk of venous thromboembolism and the use of third generation progestagens: Role of bias in observational research. Contraception 1996;54:5-13

Weiss NS. Bias in studies of venous thromboembolism in relation to the use of new formulations of oral contraceptives. (letter) Contraception 1997;55:189-190

Suissa S, Blais L, Spitzer WO, Cusson J, Lewis M, Heinemann L. First-time use of newer oral contraceptives and the risk of venous thromboembolism. Contraception 1997;56:141-146

Large Linked Data Resources

January 5 and 7

Purpose

Learn the mechanics and organization of large linked databases

Review the breadth of study options in these resources

Readings

Lanza LL, Walker AM, Bortnichak EA, Dreyer NA. Peptic ulcer and gastrointestinal hemorrhage associated with nonsteroidal anti-inflammatory drug use in patients younger than 65 years. Arch Intern Med 1995;155:1371-1377

Lanza LL, Dreyer NA, Schultz NJ, Walker AM. Use of insurance claims in epidemiologic research: Identification of peptic ulcers, GI bleeding, pancreatitis, hepatitis and renal disease. Pharmacoepidemiology Drug Safety 1995;4:239-248

Roles for Epidemiologists in the Pharmaceutical Industry

January 12

Purpose

Review functions of drug development from the viewpoint of an epidemiologist’s contribution.

Reading

None. Note that term papers are due in one week, on January 19.

The Tension between Clinical Trialists and Observational Investigators

January 14

Purpose

The attitude toward what constitutes a valid basis for causal inference is long-standing point of friction between epidemiologists and clinical experimenters. At the heart of the controversy are issues of psychology as much as logic.

Readings

None

Denial: Some problems of institutional recognition of adverse drug effects in government and industry

January 19

Purpose

More psychology. What seems to be true is commonly a function of how closely you identify with the history and outcomes of a question.

Reading

And the Band Played On. Randy Shilts. Excerpts.