PHARMACOEPIDEMIOLOGY serious infections and malignancies, it is gen- formed using selected patients who are closely erally considered preferable, or even mandatory, observed. Rarely are there any significant over to have studies with placebo controls. There are doses in this population. Thus, the study of the a number of reasons for this. First, it is easier to effects of a drug when ingested in extremely high show that a new drug is more effective than a doses is rarely possible prior to drug marketing placebo than to show it is more effective than Again, this must await postmarketing pharma another effective drug. Second, one cannot actu- coepidemiologic studies ally prove that a new drug is as effective as a Finally, it is only recently that our society has standard drug. A study showing a new drug is become more sensitive to the costs of medical no worse than another effective drug does not care, and only recently have the techniques of provide assurance that it is better than a placebo; health economics been applied to evaluate the one simply could have failed to detect that it was cost implications of drug use. y It is clear that in fact worse than the standard drug. One could the exploration of the costs of drug use requires require a demonstration that a new drug is more consideration of much more than just the costs effective than another effective drug, but this is a of the drugs themselves. The costs of a drug's standard which does not and should not have to adverse effects may be much more than the costs be met. Yet, optimal medical care requires infor- of the drug, if these adverse effects result in addi- mation on the effects of a drug relative to the tional medical care and possibly even hospitaliza- alternatives available for the same indication. tions. 0 Conversely, a drugs beneficial effects This information must often await studies con- could reduce the need for medical care, resulting ducted after drug marketing in savings which can be much larger than the cost of the drug itself. As with ' studies of drug NEW TYPES OF INFORMATION NOT utilization, the economic implications of drug use AVAILABLE FROM PREMARKETING STUDIES an be predicted prior to marketing, but can only be rigorously studied after marketing(see Chap. As mentioned above, premarketing studies are ter 32) necessarily limited in size. The additional sample size available in postmarketing studies permits the study of drug effects which may be uncom- GENERAL CONTRIBUTIONS O mon,but important, such as drug-induced agra- PHARMACOEPIDEMIOLOGY nulocytosis Lastly, it is important to review the general con Premarketing studies are also necessarily lim- tributions that can be made by pharmacoepide ited in time; they must come to an end, or the miology. As an academic or a clinician, one is drug could never be marketed! In contrast, post- most interested in the new information about marketing studies permit the study of delayed drug effects and drug costs that can be gained drug effects, such as the unusual clear cell adeno- from pharmacoepidemiology. Certainly, these are carcinoma of the vagina and cervix which occur- the findings that receive the greatest public and red two decades later in women exposed in utero political attention. However, often no new infor to diethylstilbestrol. 9 mation is obtained, particularly about new The patterns of physician prescribing and adverse drug effects. This is not a disappointing patient drug utilization often cannot be predicted outcome, but in fact a very reassuring one,and prior to marketing, despite pharmaceutical manu- this reassurance about drug safety is one of the facturers'best attempts to do so, in order to plan most important contributions that can be made for drug marketing. Studies of how a drug is by pharmacoepidemiology studies. Related to actually being used, and determinants of changes this is the reassurance that the sponsor of the in these usage patterns, can only be performed study, whether manufacturer or regulator, is ful- after drug marketing(see Chapters 27 and 28) filling its organizational duty ethically and In most cases, premarketing studies are per- responsibly by looking for any undiscovered 第176页
第 176 页
WHAT IS PHARMACOEPIDEMIOLOGY? problems which may be there. In an era of pro- 16. Joyce CRB, Last JM, Weatherall M. Personal fac duct liability litigation, this is an important tors as a cause of differences in prescribing by assurance cannot change whether a dr general practitioners. Br J Prev Soc Med 1968: 22 es an adverse reaction, and the fact that it 17. Muller C. Outpatient drug prescribing related to does will hopefully eventually become evident. clinic utilization in four New York City hospitals What can be changed is the perception about Health Sery Res 1968: 3: 142-54 whether a manufacturer did everything possibl 18. Kono r. Trends and lessons of smon research to detect it and, so, whether it was negligent in In: Soda T, ed, Drug-Induced Suferings. Prince- ton: Excerpta Medica, 1980; 11 Herbst AL, Ulfelder H, Poskanzer DC. Adeno- carcinoma of the vagina: association of maternal stilbestrol therapy with tumor appearance REFERENCES 20 young women. N Engl J Med 1971: 284: 878-81 Wright P. Untoward effects associated with prac toll administration. Oculomucocutaneous syn- 1. Moore RD, Smith CR, Lietman PS. Increased risk drome. Br Med J 1975: 1: 595-8 of renal dysfunction due to interaction of liver dis- 21. Ticrynafen recalled. FDA Drug Bull 1980; 10: 3-4. ease and aminoglycosides. Am J Med 1986: 80: 22. Suspension of benoxaprofen (Opren). Br Med J 1093-7. 1982;285:51920 2. Rawlins MD, Thompson JW. Pathogenesis of 23. Strom BL, Carson JL, Morse ML, et al. The adverse drug reactions. In: Davies DM, ed, Text of indication on hypersensitivity book of Adverse Drug Reactions. Oxford University ciated with zomepirac sodium effect Pres,1977;44 steroidal antiinflammatory drugs Rheum 3. Strom BL. Integrating pharmacoepidemiology into 1987;30:1142-8. the design and conduct of clinical trials. J Clin Res 24. Inman WHW. Study of fatal bone marrow depres- Drug Dev1988;2:161-8 sion with special reference to phenylbutazone and 4. Geiling EMK, Cannon PR. Pathogenic effects of oxyphenbutazone. Br Med J 1977: 1: 1500-5. elixir of sulfanilimide(diethylene glycol) poisoning 25. Committee on Safety of Medicines. Osmosin JAMA1938;111:91926. (Controlled release indomethacin ) Current Prob- 5. Wallerstein RO, Condit PK, Kasper CK, et al. lems, No. 11. London: Committee on Safety of Statewide study of chloramphenicol therapy and Medicines, 1983. atal aplastic anemia. JAMA 1969: 208: 2045-50. 26. Barash CL, Lasagna L. The Bendectin saga: 6. Meyler L. Side Efects of Drugs. Amsterdam: Voluntary "discontinuation. J Clin Res Drug Dev Elsevier. 1952 1987;1:277-92. 7. Erslev A, Wintrobe MM. Detection and preven 27. Strom BL, West SL, Sim E, Carson JL. The epi- ion of drug induced blood dyscrasias. JAMA demology of the acute flank pain syndrome from 1962;181:114-9 suprofen. Clin Pharmacol Ther 1989; 46: 693-9 8. Cluff LE, Thornton GF, Seidl'LG. Studies on the 28. FDa ponders approaches to curbing adverse epidemiology of adverse drug reactions. I. Meth effects of drug used against cystic acne. JAMA ods of surveillance. JAMA 1964: 188: 976-83. 1988:259:3225-30 9. Miller RR, Greenblatt DJ. Drug Efects in Hospi- 29. Stern RS. When a uniquely effective drug is ter- alized Patients. New York: Wiley, 1976. atogenic: The case of isotretinoin. N Engl J Med 10. Caranasos G, Stewart RB, Cluff LE. Drug- 1989;320:1007-9 nduced illness leading to hospitalization. JAMA 30. Hertzman PA, Falk H, Kilbourne EM, Page S, 1974;228:713-7. Shulman LE. The eosinophilia-myalgia syn- 11. Lenz w. Malformations caused by drugs in preg- drome: The Los Alamos Conference. J Rheumatol ancy. Am J Dis Child 1966: 112: 99-106 1991:18:867-73 2. Wardell WM. Therapeutic implications of the drug 31. Meyboom RHB. The triazolam exp in1979 Clin pharmacol Ther 1974: 15: 73-96 in The Netherlands, a problem of enera 3. Lee JAH, Draper PA, Weatherall M. Prescribin tion and verification. In: Strom BL in three English towns. Milbank Mem Fund Epidemiology and Postmarketing Drug Sur- 1965:43:285-90 veillance. New York: Plenum Press 14. Muller C. Medical of prescribing, J Chronic 32. Triazolam's status in European Community Di1965;18:6899 Lancet1991:338:1586-7 15. Meade Tw. Presc of chloramphenicol in 33. Brahams D. Triazolam suspended. Lancet 1991 B J1967;1:671-4. 338:938 77页
第 177 页
PHARMACOEPIDEMIOLOGY 34. Triazolam hearing. Lancet 1992; 339 35. Randall T. Penile, testicular acetate (DMPA) and risk of invasive squamous implants silicone cell cervical cancer. Contraception 1992: 45:299- JAMA1992;267:2578-9 Egger M, Smith GD, Imhoof H, Teuscher A. risk 50. Nightingale SL. Warnings issued on nonsedating antihistamines terfenadine and astemizole. JAMA of severe h patients transferred to human insulin: a case con- 51. Ahmad SR. Antihistamines alert. Lancet 1992: = 340:542 Smith GD, Teuscher AU, Teuscher A. 52. Rothman K, Funch DP, Dreyer NA. Bromo- of human insulin on symptoms and awareness of hypoglycaemia: a randomised double criptine and puerperal seizures. Epidemiology 1990 1:232-8. blind crossover trial. Br Med J 1991: 303: 622-6 38. Colagiuri S, Miller JJ, Petocz P. Double-blind ross TP. Bromocriptine and puerperal seizures Epidemiology 1991; 2: 234-5 crossover comparison of human and porcine insu- 54. Finch RG. The withdrawal of Temafloxacin: Are lins in patients reporting lack of hypoglycaemia awareness. Lancet 1992; 339: 1432-5 here implications for other quinolones? Drug Saf 1993:8:9-11 39. Jick H, Hall GC, Dean AD, Jick SS, Derby LE A 55. Slone D, Shapiro S, Miettinen OS. Case-control omparison of the risk of hypoglycemia between surveillance of serious illnesses attributable t users of human and animal insulins. 1. Experience in the United Kingdom. Pharmacotherapy 1990 ambulatory drug use. In: Colombo F, Shapiro S Slone D, Tognoni G, eds, Epidemiological Evalua- 10:395-7 tion of Drugs. Massachusetts: PSG, 1977: 59-82 40. Jick SS, Derby LE, Gross KM, Jick H. Hospitali- 56. Joint Commission on Prescription Drug Use. Final Report. Washington, Joint Commission on Pre and human insulins. 2. Experience in the United scription Drug Use, 1980 States. Pharmacotherapy 1990; 10: 398-9 57. Strom BL, Carson JL, Morse ML, Leroy AA.The 41. Teicher MH, Glod C, Cole J. Emergence of intense suicidal Computerized Online Medicaid Analysis and Sur treatment. Am J Psychiatry 1990; 147: 207. veillance System: a new resource for post-market- 42. Beasley CM, Dormseif BE, Bosomworth JC, et al. ing drug surveillance. Clin Pharmacol Ther 1985 trolled trials of treatment for depression. Br Med Inman WHW. Prescription Event Monitoring 1991;303:685-92 59. Mattison N, Richard BW. Postapproval research 43. Crane J, Flatt A, Jackson R, et al. Prescribed requested by the FDa at the time of Nce fenoterol and death from asthma in New zealand 1981-83: case-control study. Lancet 1989:1:917- 3oa019gl019321309 44. Pearce N, Grainger J, Atkinson M, ef al. Case- 60. Rosendorff C. Prazosin: severe side effects are dose-dependent. Br Med J 1976; 2: 508 ontrol study of prescribed fenoterol and death 61. Graham RM, Thornell IR, Gain JM, et al. Prazo- from asthma in New Zealand, 1977-81. Thorax 1990;45:917-22 sin: the first-dose phenomenon. Br Med J1976: 2 1293-4 45. Grainger J, Woodman K, Pearce N, et al. Pre- Kaitin KI. Case studies of expedited review: AZT scribed fenoterol and death from asthma in New and L-Dopa. Law Med Health Care 1991: 19: 242 Zealand. 1981-7: a further case-control study 246 Thorax1991;46:105-11 46. Spitzer wO, Suissa S, Ernst P, et al. The use of B- 63. Wastila L, Lasagna L. The history of zidovudine (AZT). J Clin Res Pharmacoepidemiol 1990: 4: 25- agonists and the risk of death and near death from asthma. N Engl J Med 1992: 326: 501-6 47. Rosenfield A. Maine D, Rochat R, Shelton J 64. McKenzie MW, Marchall GL, Netzloff ML, et al. Adverse drug reactions leading to hospitalization Hatcher RA. The Food and Drug Administration and medroxyprogesterone acetate. What are the 65. May FE, Stewart RB, ClufT LE. Drug interactions Issues?JAMA1983;249:2922-8 and multiple drug administration. Clin Pharmacol 48. WHO Collaborative Study of Neoplasia and Ster Ther1977;22:322-8 oid Contraceptives. Breast cancer and depot- 66. Nelson WL, Fraunfelder FT, Sills JM medroxyprogesterone acetate: a multinational Adverse respiratory and cardiovascular tudy. Lancet1991;338:833-8. 49. WHO Collaborative Study of Neoplasia and Ster- attributed to timolol ophthalmic solution, 1985. Am J Ophthalmol1986;102:606-11 oid Contraceptives. Depot-medroxyprogesterone The International Agranulocytosis and Aplastic 第178页
第 178 页
WHAT IS PHARMACOEPIDEMIOLOGY? Anemia Study f agranulocytosis and 69. Eisenberg JM. New drugs and clinical econe aplastic anemia report of their relation to analysis of cost-effectiveness analysis in the dn reference to analgesics JAMA1986;256:1749-57 ment of pharmaceutical innovations. Rev Dis19846( Suppl4):S905-8 68. Stewart RB, Forgnone M, May FE, et al. Epide- 70. Morse ML, Leroy AA, Gaylord TA, Kellenberger miology of acute drug intoxications: patient char T. Reducing drug therapy-induced hospitalization acteristics, drugs, and medical complications. Clin Toxicol1974;7:513-30 mpact of drug utilization review. Drug InfJ1982 16:199-202 第179页
第 179 页
2 Study Designs Available for Pharmacoepidemiology Studies BRIAN L. STROM University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania, USA Pharmacoepidemiology applies the methods of principles of clinical pharmacology, the content epidemiology to the content area of clinical phar- area of pharmacoepidemiology, in a similar macology. Therefore, in order to understand the manner. approaches and methodological issues specific to he field of pharmacoepidemiology, the basic principles of the field of epidemiology must be OVERVIEW OF THE SCIENTIFIC METHOD understood. To this end, this chapter will begin with an overview of the scientific method, in gen- The scientific method is a three-stage process(see eral. This will be followed by a discussion of the Figure 2. 1). In the first stage one studies a sample different types of errors one can make in design- of study subjects. Second, one generalizes the ing a study. Next the chapter will review the information obtained from this sample of study Criteria for the Causal Nature of an Associa- subjects, drawing a conclusion about a popula tion, "which help guide the decision of whether tion in general. This conclusion is referred to as an association demonstrated in a particular study an association. Third, one generalizes again is, in fact, a causal association. Finally, the drawing a conclusion about scientific theory or specific study designs available for epidemiologic causation. Each will be discussed in turn studies, or in fact for any clinical studies, will be Any given study is performed on a selection of reviewed. The next chapter discusses a specific individuals, who represent the study subjects methodological issue which needs to be addres- These study subjects should theoretically repre- sed in any study, but which is of particular sent a random sample of some defined popula importance for pharmacoepidemiology studies: tion. For example one might perform a the issue of sample size. These two chapters are randomized clinical trial of the efficacy of intended to be an introduction to the field of epi- methyldopa in lowering blood pressure,ran- demology for the neophyte. More information domly allocating a total of 40 middle-aged on these principles can be obtained from any hypertensive men to receive either methyldopa or textbook of epidemiology or clinical epidemiol- placebo and observing their blood pressure six ogy. -13 Finally, Chapter 4 will review basic weeks later. One might expect to see the blood Pharmacoepidemiology( Second edition). Edited by Brian L.Strom 第180页
第 180 页