Accounting for serious adverse events

Drs Miller and Humphries[1] raise important issues regarding serious adverse events (SAEs) that merit further discussion, particularly in relation to our Therapeutics Letter #48, “Do statins have a role in primary prevention?”[2]

However, Drs Miller and Humphries misrepresent a number of issues addressed in Letter #48:

1. We did not write that “statins may do more harm than good in patients without a myocardial infarction.” Rather, we interpreted the available SAE data as consistent with the possibility that unrecognized serious adverse events are increased by statin therapy and that the magnitude of the increase is similar to the magnitude of the reduction in cardiovascular serious adverse events in these populations. This hypothesis can be tested through full disclosure of RCT findings.

2. We did not state that total SAEs are the most “appropriate method(s) to evaluate RCTs.” SAEs are one of a number of important outcomes from RCTs that require full evaluation.

3. Miller and Humphries write that “Professor James Shepherd, the principal investigator of the PROSPER study, has confirmed that the primary outcomes in his study were not included in the SAE rate quoted in the publication.” However, in 2004 the statistician for PROSPER stated the opposite indicating that he felt it necessary to correct what Professor Shepherd had said. According to the statistician, “The results in the PROSPER Lancet manuscript relate to all SAEs including study endpoints” (personal communication) To us, the confusion between the lead author and the lead statistician illustrates the need to clarify SAEs and their utility in clinical trial interpretation, not to disregard the subject.

SAEs are data routinely collected during all clinical trials. In addition, manufacturers must document all SAEs of which they become aware after drug licensing and report such data to national regulators within 15 days. Unfortunately, much of this information never appears in published reports, particularly journal articles.

Recording and reporting of SAEs (according to standard definitions) automatically includes all deaths and life-threatening events and thus has the advantage of capturing both benefit and harm. Herein lies their greatest strength but also a common source of confusion. Consider mortality for example. A decrease in this SAE constitutes benefit, while an increase constitutes harm. Therefore, in contrast to what Drs Miller and Humphries imply, careful analysis of SAEs does not reflect a focus on adverse effects of therapy. In fact, it reflects an interest in attaining a comprehensive measure of overall benefit or harm from drug therapy. It is, of course, dependent on the proper collection, analysis, and reporting of all SAE data, which seldom occurs in the published literature—the primary conclusion of TI Letter #48.

We agree with Drs Miller and Humphries’ emphasis on the difficulties associated with analysis of uncommon events that display high variance. Thorough and proper analysis represents a significant challenge for clinicians, academics, and policy makers. The analytical issues raised by Miller and Humphries have been considered in depth in the academic literature.[3]

One issue worth highlighting relates to mortality from statin therapy. For more than a decade, researchers have raised concerns about the harmful effects of lipid lowering drugs.[4] The earlier trials for fibrates (clofibrate, gemfibrozil, bezafibrate, and fenofibrate) showed significant reduction in coronary deaths but no reduction in total mortality.[5] Meta-analyses showed a statistically significant increase in noncoronary deaths.[6] No single cause of death explained the increase in noncoronary deaths.

Concern regarding increased noncoronary deaths diminished with the Scandinavian Simvastatin Survival Study (4S).[7] Simvastatin, an HMG CoA reductase inhibitor (statin), reduced total deaths as well as coronary deaths. Similar coronary and total mortality benefit was found with pravastatin.[8] However, both of these early statin studies enrolled patients who had symptomatic coronary heart disease, termed “secondary” prevention populations.

In primary prevention trials, statins have not reduced coronary or total mortality either in individual trials[9-13] or in meta-analyses of all trials.[14] A total mortality benefit has not been demonstrated in primary prevention despite the fact that statins reduce nonfatal myocardial infarction and nonfatal stroke.[2] No explanation has been provided for the lack of observed mortality benefit. Two logical possibilities are that when used for primary prevention, statins:

• Reduce only the number of nonfatal coronary and cerebral vascular events, or

• Reduce some types of cardiovascular deaths but this is offset by an increase in noncardiovascular death.

In the case of mortality in primary prevention studies, our meta-analysis demonstrates a lack of power to detect a 5% relative difference in mortality as demonstrated by the confidence limits around the point estimate relative risk (RR) of 0.95. However, a quick analysis shows that there exists an 80% power to detect a 10% drop in relative risk. While this does not entirely rule out a mortality benefit for statins for primary prevention, it has not been demonstrated by experiments involving very large study populations (totaling approximately 40 000 people).

The intent and advantage of randomized clinical trials is to identify both expected and unexpected effects of treatment. Proper randomization creates an even distribution of characteristics between the experimental groups. Consequently, all observed differences in outcomes between groups (both favorable and unfavorable) must be evaluated as to whether they are likely due to chance or to real differences in treatment. To argue that clinical trials are designed to look at only one primary outcome misses the point of conducting a controlled trial. Analysis of a trial must also compare and weigh benefit against harm. When this is done and overall benefit is not demonstrated, the usefulness of a treatment is questionable.

In summary, while we agree with Drs Miller and Humphries that companies do report SAEs to regulatory bodies, we do not know how the regulatory bodies deal with that information and we feel that clinicians, scientists, and patients are entitled to this same information. Accurate reporting, full availability, and thorough analysis of SAE data are essential components to properly interpret the clinical significance of clinical trial findings.

—Ken Bassett, MD 
Centre for Health Services and Policy Research, UBC
General Practice, Surrey
Chair, Drug Assessment Working Group, Therapeutics Initiative, UBC
Acting Director, Therapeutics Initiative, UBC

—Keith Chambers, MD
Associate Clinical Professor
Health Care and Epidemiology, UBC
Clinical Epidemiologist, Vancouver Hospital

—James McCormack, BSc(Pharm) PharmD
Professor, Faculty of Pharmaceutical Sciences, UBC
Clinical Supervisor, St. Paul’s Hospital

—Thomas L. Perry Jr., MD 
Clinical Assistant Professor
Departments of Anesthesiology, Pharmacology & Therapeutics and Medicine, UBC
Clinical Pharmacologist, General Internist, Vancouver Hospital

—James M. Wright, MD
Professor, Departments of Anesthesiology, Pharmacology & Therapeutics and Medicine, UBC
Director, Therapeutics Initiative, UBC

References

1. Miller DB, Humphries KH. A new way to evaluate randomized clinical trials? New approach does more harm than good. BCMJ 2005;47:241-244. Full Text
2. Do statins have a role in primary prevention? Ther Lett 2003;48. www.ti.ubc.ca/PDF/48.pdf.
3. Brookes ST, Whitely E, Eggar M. Subgroup analyses in randomized trials: Risks of subgroup-specific analyses; power and sample size for the interaction test. J Clin Epidemiol 22004;57:229-236. PubMed Abstract Full Text
4. Muldoon MF, Manuck SB, Matthews KA. Lowering cholesterol concentrations and mortality: A quantitative review of primary prevention trials. BMJ 1990;301:309-314. PubMed Abstract
5. The Lipid Research Clinics Coronary Primary Prevention Trial results. I. Reduction in incidence of coronary heart disease. JAMA 1984;251:351-364. PubMed Abstract
6. Davey Smith G, Pekkanen J. Should there be a moratorium on the use of cholesterol lowering drugs? BMJ 1992;304:431-434. PubMed Citation
7. Randomised trial of cholesterol lowering in 4444 patients with coronary heart disease: The Scandinavian Simvastatin Survival Study (4S). Lancet 1994;344:1383-1389. PubMed Abstract
8. Prevention of cardiovascular events and death with pravastatin in patients with coronary heart disease and a broad range of initial cholesterol levels. The Long-Term Intervention with Pravastatin in Ischaemic Disease (LIPID) Study Group. N Engl J Med 1998;339:1349-1357. PubMed Abstract Full Text
9. Downs JR, Clearfield M, Weis S, et al. Primary prevention of acute coronary events with lovastatin in men and women with average cholesterol levels: Results of AFCAPS/TexCAPS. JAMA 1998;279:1615-1622. PubMed Abstract Full Text
10. Sever PS, Dahlof B, Poulter NR, et al. Prevention of coronary and stroke events with atorvastatin in hypertensive patients who have average or lower-than-average cholesterol concentrations, in the Anglo-Scandinavian Cardiac Outcomes Trial—Lipid Lowering Arm (ASCOT-LLA): A multicentre randomised controlled trial. Lancet 2003;361:1149-1158. PubMed Abstract Full Text
11. ALLHAT Officers and Coordinators for the ALLHAT Collaborative Research Group. Major outcomes in moderately hypercholesterolemic, hypertensive patients randomized to pravastatin vs usual care: The Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial (ALLHAT-LLT). JAMA 2002;288:2998-3007. PubMed Abstract Full Text
12. Shepherd J, Blauw GJ, Murphy MB, et al. Pravastatin in elderly individuals at risk of vascular disease (PROSPER): A randomised controlled trial. Lancet 2002;360:1623-1630. PubMed Abstract Full Text
13. Shepherd J, Cobbe SM, Ford I, et al. Prevention of coronary heart disease with pravastatin in men with hypercholesterolemia. N Engl J Med 1995;333:1301-1308. PubMed Abstract Full Text
14. Pignone M, Phillips C, Mulrow C. Use of lipid lowering drugs for primary prevention of coronary heart disease: Meta-analysis of randomised trials. BMJ 2000;321:983-986. PubMed Abstract Full Text