The popularity of meta-analysis in medical research can be attributed to the methodologic appeal of systematically identifying, reviewing, and combining individual studies, and explicitly reporting criteria used in their review. Meta-analysis also increases the chance of detecting effects of moderate size, which nevertheless have important public health implications, through minimization of biases and small random errors. The purpose of this report is to describe how a meta-analysis is performed, including protocol development, literature search strategy, considerations of study homogeneity, and combinability of studies. Methodologic issues concerning the conduct of meta-analysis are also discussed. A meta-analysis performed to evaluate the efficacy of verapamil usage in preventing secondary nonfatal myocardial infarctions among patients recovering from an acute myocardial infarction is presented to illustrate the issues.
HeldPYusufSFurbergC. Calcium channel blockers in acute myocardial infarction and unstable angina: an overview. Br Med J.1989;299:1187–1192.
6.
HeldPHYusufS. Effects of beta-blockers and calcium channel blockers in acute myocardial infarction. Eur Heart J.1993;14(supp F):18–25.
7.
HeldPYusufS. Calcium antagonists in the treatment of ischemic heart disease: myocardial infarction. Coronary Artery Dis.1994;5:21–26.
8.
YusufSHeldPFurbergC. Update of effects of calcium antagonists in myocardial infarction or angina in light of the second Danish verapamil infarction trial (DAVIT-2) and other recent studies. Am J Cardiol.1991;67:1295–1297.
9.
YusufSLessemJJhaPLonnE. Primary and secondary prevention of myocardial infarction and strokes: An update of randomly allocated, controlled trials. J Hypertension.1993;11(supp 4):S61–S73.
10.
PsatyBMHeckbertSRKoepsellTDThe risk of myocardial infarction associated with antihypertensive drug therapies. JAMA.1995;274:620–625.
11.
ChalmersTCSmithHJrBlackburnBA method for assessing the quality of a randomized control trial. Control Clin Trials.1981;2:31–49.
12.
HardyRJThompsonSG. Detecting and describing heterogeneity in meta-analysis. Stat Med.1998;17: 841–856.
13.
PetoR. Why do we need systematic overviews of randomized trials?Stat Med.1987;6:233–240.
14.
CollinsRPetoRMacMahonSBlood pressure, stroke, and coronary disease. Part 2, Short-term reductions in blood pressure: overview of randomized drug trials in their epidemiological context. Lancet.1990;335:827–838.
15.
HosmerDWLemeshowS. Applied Logistic Regression.New York, NY: John Wiley & Sons; 1989.
16.
BerlinJALairdNMSacksHSChalmersTC. A comparison of statistical methods for combining event rates from clinical trials. Stat Med.1989;8: 154–176.
17.
ThompsonSGPocockSJ. Can meta-analysis be trusted. Lancet.1991;338:1127–1130.
18.
DerSimonianRLairdN. Meta-analysis in clinical trials. Control Clin Trials.1986;7:177–188.
19.
WhiteheadAWhiteheadJ. A general parametric approach to the meta-analysis of randomized clinical trials. Stat Med.1991;10:1665–1677.
20.
BrandRKragtH. Importance of trends in the interpretation of an overall odds radio in the meta-analysis of clinical trials. Stat Med.1992;11:2077–2082.
21.
BerkeyCSHoaglinDCMostellerFColditzGA. A random-effects regression model for meta-analysis. Stat Med.1995;142:383–387.
22.
HardyRJThompsonSG. A likelihood approach to meta-analysis with random effects. Stat Med.1996;15:619–629.
23.
CookRJWalterSD. A logistic regression model for trend in 2 × 2 × K tables with applications to meta-analysis. Biometrics.1997;53:352–357.
24.
WalterSD. Variation in baseline risk as an explanation of heterogeneity in meta-analysis. Stat Med.1997;16:2883–2900.
25.
The Danish Study Group on Verapamil in Myocardial Infarction.Eur Heart J.1984;5:516–528.
26.
LauJSchmidCHChalmersTC. Cumulative metaanalysis of clinical trials builds evidence for exemplary medical care. J Clin Epidemiol.1995;48: 45–57.
27.
MantelNHaenszelW. Statistical aspects of the analysis of data from retrospective studies of disease. J Natl Cancer Inst.1959;22:719–748.
28.
HansenJFSigurdBMellemgaardKLyngbyeJ. Verapamil in acute myocardiol infarction. Dan Med Bull.1980;5:516–528.
29.
BussmanWDSeherWGruengrassM. Reduction of creatine kinase and creatine kinase—MB indexes of infarct size by intravenous verapamil. Am J Cardiol.1984;54:1224–1230.
30.
CreaFDeanfieldJCreanPSharonMDaviesGMaseriA. Effects of verapamil in preventing early post-infarction angina and reinfarction. Am J Cardiol.1984;55:900–904.
31.
The Danish Study Group On Verapamil in Myocardial Infarction.The effect of verapamil on mortality and major events after myocardial infarction (The Danish Verapamil Infarction Trial II—DAVIT II). Am J Cardiol.1990;66:779–785.
32.
RegnoFCarboninPPahorMA controlled trial of verapamil in patients after acute myocardial infarction: results of the Calcium Antagonist Reinfarction Italian Study (CRIS). Am J Cardiol.1996;77:365–369.
33.
HansenJF. Review of postinfarct treatment with verapamil: Combined experience of early and late intervention studies with verapamil in patients with acute myocardial infarction. Cardiovas Drugs Ther.1994;8:543–547.
