RSS Feeds
-
Validity of composite end points
Submit responseI read with great interest the paper by Lim et al. The authors showed that the estimates of effect based on composite outcomes were sometimes inconsistent with those for the outcomes’ individual components and that composite outcomes were frequently used when the assumption that each individual event was equally important was not met. They also demonstrated that commonly occurring but less clinically important end points contributed more to composite estimates of effect and that less commonly occurring but more clinically serious end points contributed less (1).
I would like to emphasize that the use of composite end points is extended in randomized controlled trials and especially in meta-analyses. Various combinations of end points like death, myocardial infarction, stroke and others are used in cardiothoracic research like those evaluating, for example, off-pump coronary artery bypass grafting versus on pump. It is very important to focus on this kind of meta-analysis as the results might prove if one surgical technique is superior to the other (2).
Another example is the use of composite end points in cardiology research like MACE (major adverse cardiac events). By definition, MACE and other composite end points include multiple types of clinical events. Kip et al. showed that there was significant heterogeneity in the individual outcomes used to define composite end points such as MACE and widely different results might be obtained even within a single study according to the sets of clinical outcomes that were used. Unfortunately, the use of MACE has extended to reporting of outcomes in which both effectiveness and safety outcomes are put together, even when one is not necessarily interested in assessing the ‘net’ effect of a particular intervention. They also showed that the use of MACE could lead to different conclusions about the ‘effect’ of a particular intervention. Another interesting issue might be the potential for certain sponsors to “game” their trials by construction and use of particular composite outcomes. Besides, if different definitions were used for even one composite end point, rates of the composite end point might vary widely (3).
The validity of composite end points is important. Three questions have to be answered if the clinicians decide to base their decision on them.
Are the component end points of similar importance to patients? Did the more and less important end points occur with similar frequency? Are these end points likely to have similar relative risk reductions(4)?
References
1. Lim E, Brown A, Helmy A, Mussa S, Altman D. Composite Outcomes in Cadiovascular Research:A Survey of Randomized Trials. Ann Intern Med. 2008; 149: 612-617.
2. Parolari A, Alamanni F, Cannata A, Naliato M, Bonati L, Rubini P, et al. Off-pump versus on-pump coronary artery bypass: meta-analysis of currently available randomizes trials. Ann Thorac Surg 2003, 76 (1): 37-40.
3. Kip K, Hollabaugh K, Marroquin O, Williams D. The Problem with Composite End Points in Cardiovascular Studies. J Am Coll Cardiol 2008; 51: 701-707.
4. Montori V, Permanyer-Miralda G, Ferreira-Gonzalez I, Busse J, Pachero-Huergo V, Bryant D. Validity of composite end points in clinical trials. BMJ 2005; 330: 594-96.
Conflict of Interest:
No competing interests.
-
COMPOSITE OUTCOMES: A POTENTIAL STRATEGY FOR MASKING “P-VALUES DRIVEN ENDPOINTS”?
Submit responseIn their article, Lim and colleagues contribute important insights into the use of composite outcomes in clinical trials (1). In doing so, their report contributes to a debate initiated a few years ago (2-6). Confirming our earlier work (5), Lim’s study shows that composite outcomes in cardiovascular trials are common, that components usually vary in importance to patients, that events occur more frequently in the less important components than the most important ones, and that less important outcomes (i.e. angina) usually drive the effect of the therapy on the composite.
What are the new contributions that Lim’s work makes? We believe there are two. First, Lim and colleagues suggest that the smaller the sample size of a trial, the more the components that are included in composite outcomes. This is consistent with the primary motivation for most of the use of composite outcomes: to reduce sample size requirements to achieve the longed for statistical significance (4). Second, a plot of the P values of the composite outcome revealed marked asymmetry around the midpoint of the plot, thus suggesting reporting bias: trialists may have selected components to include in the composite that favour statistical significance.
Overall, Lim’s study supports views we have expressed in previous writings (3-6). While there are instances in which the use of composite endpoints are appropriate (4) they run a serious risk of providing misleading impressions about the nature and magnitude of treatment benefits. This occurs when the composite outcome includes components of widely varying importance to patients and the less important components are primarily responsible for the observed effect (6). These instances provide serious challenges for the interpretation of results and their application in clinical practice, and run a high risk of misleading both patients and physicians.
Reference List
(1) Lim E, Brown A, Helmy A, Mussa S, Altman DG. Composite outcomes in cardiovascular research: a survey of randomized trials. Ann Intern Med. 2008;149:612-17.
(2) Freemantle N, Calvert M, Wood J, Eastaugh J, Griffin C. Composite outcomes in randomized trials: greater precision but with greater uncertainty? JAMA. 2003;289:2554-59.
(3) Montori VM, Permanyer-Miralda G, Ferreira-Gonzalez I, Busse JW, Pacheco-Huergo V, Bryant D et al. Validity of composite end points in clinical trials. BMJ. 2005;330:594-96.
(4) Ferreira-Gonzalez I, Permanyer-Miralda G, Busse JW, Bryant DM, Montori VM, Alonso-Coello P et al. Methodologic discussions for using and interpreting composite endpoints are limited, but still identify major concerns. J Clin Epidemiol. 2007;60:651-57.
(5) Ferreira-Gonzalez I, Busse JW, Heels-Ansdell D, Montori VM, Akl EA, Bryant DM et al. Problems with use of composite end points in cardiovascular trials: systematic review of randomised controlled trials. BMJ. 2007;334:786.
(6) Guyatt GH, Montori VM, Ferreira-Gonzalez I, Busse JW, Schunemann HJ, Jaeschke R et al. Composite Endpoints. In: Guyatt GH, Rennie D, Meade MO, Cook DJ, eds. Users' guides to the medical literature. Second ed. McGrawHill; 2008: 237-48.
Conflict of Interest:
None declared