Medical research commonly relies on the combination of 2 x 2 tables of counted data for making inferences about treatment effects or about the causes of disease. This article reviews point estimation and interval estimation for a common odds ratio. Traditional methods for providing these estimates face special challenges, and sometimes break down, when the data are sparse. Recent research provides practical alternatives to the traditional methods, and uses new computer algorithms to make them readily available in microcomputer software packages. This article illustrates and compares the various methods and offers recommendations for using the newer methods when analysing data sets of moderate size.
Get full access to this article
View all access options for this article.
References
1.
Mantel N., Haenszel W.Statistical aspects of the analysis of data from retrospective studies of disease. Journal of the National Cancer Institute1959; 22: 719-48.
2.
Hauck WWThe large-scale variance of the Mantel-Haenszel estimator of a common odds ratio. Biometrics1979; 35: 817-19.
3.
Gart JJPoint and interval estimation of the common odds ratio in the combination of 2 x 2 tables with fixed marginals. Biometrika1970; 57: 471-75.
4.
Fleiss JLStatistical methods for rates and proportions, 2nd Ed. New York: Wiley, 1981.
5.
Tarone RE, Gart JJ, Hauck WWOn the asymptotic inefficiency of certain noniterative estimators of a common relative risk or odds ratio. Biometrika1983 ; 70: 519-22.
6.
Hauck WW, Donner A.The asymptotic relative efficiency of the Mantel-Haenszel estimator in the increasing-number-of-strata case. Biometrics1988; 44: 379-84.
7.
Breslow N.Odds ratio estimators when the data are sparse. Biometrika1981; 68: 73-84.
8.
Donner A., Hauck WWThe large-sample relative efficiency of the Mantel-Haenszel estimator in the fixed-strata case. Biometrics1986;42:537-45.
9.
Liang KYOdds ratio inference with dependent data. Biometrika1985; 72: 678-82.
10.
Donald A., Donner A.A simulation study of the analysis of sets of 2 x 2 contingency tables under cluster sampling: Estimation of a common odds ratio. Journal of the American Statistical Association1990; 85: 537-43.
11.
McKinlay SMThe effect of nonzero second-order interaction on combined estimators of the odds ratio. Biometrika1978; 65: 191-202.
12.
Woolf B.On estimating the relation between blood groups and disease. Annals of Human Genetics1955;19: 251-53.
13.
Yusuf S., Peto R., Lewis J., Collins R., Sleight P.Beta blockade during and after myocardial infarction: An overview of the randomized trials. Progress in Cardiovascular Diseases1985; 27: 335-71.
14.
Berlin JA, Laird NM, Sacks HS, Chalmers TCA comparison of statistical methods for combining event rates from clinical trials. Statistics in Medicine1989; 8: 141-51.
15.
Greenland S. , Salvan A.Bias in the one-step method for pooling study results. Statistics in Medicine1990; 9: 247-51.
16.
Gart JJOn the combination of relative risks. Biometrics1962;18: 601-10.
17.
Cox DRThe analysis of binary data. London: Methuen, 1970.
18.
Zelen M.The analysis of several 2 x 2 contingency tables. Biometrika1971; 58: 129-37.
19.
Greenland S.On the logical justification of conditional tests for two-by-two contingency tables. The American Statistician1991; 45: 248-51.
20.
Hauck WWA comparative study of conditional maximum likelihood estimation of a common odds ratio. Biometrics1984; 40: 1117-23.
21.
Ejigou A.Small-sample properties of odds ratio estimators under multiple matching in case-control studies. Biometrics1990; 46: 61-69.
22.
Pigeot I.A class of asymptotically efficient noniterative estimators of a common odds ratio. Biometrika1990; 77: 420-23.
23.
Pigeot I.A jacknife estimator of a combined odds ratio. Biometrics1991; 47: 373-81.
24.
Gart JJThe comparison of proportions: A review of significance tests, confidence intervals and adjustments for stratification. Review of the International Statistical Institute1971; 39: 148-69.
25.
Gart JJ, Zweifel JROn the bias of various estimators of the logit and its variance with applications to quantal bioassay. Biometrika1967; 54: 181-87.
26.
Guilbaud O.On the large-sample distribution of the Mantel-Haenszel odds-ratio estimator. Biometrics1983; 39: 523-25.
27.
Ury HKHauck's approximate large-sample variance of the Mantel-Haenszel estimator (letter to the editor). Biometrics1982; 38: 1094-95.
28.
Cornfield J.A statistical problem arising from retrospective studies. In Neyman J ed, Proceedings of the Third Berkeley Symposium1956; 4: 135-48.
29.
Breslow NE, Liang KYThe variance of the Mantel-Haenszel estimator. Biometrics1982; 38: 943-52.
30.
Flanders WDA new variance estimator for the Mantel-Haenszel odds ratio. Biometrics1985; 41:637-42.
31.
Robins J., Breslow N., Greenland S.Estimators of the Mantel-Haenszel variance consistent in both sparse data and large-strata limiting models. Biometrics1986; 42: 311-23.
32.
Robins J., Greenland S., Breslow NEA general estimator for the variance of the Mantel-Haenszel odds ratio. American Journal of Epidemiology1986; 124: 719-23.
33.
Phillips A. , Holland PWEstimators of the variance of the Mantel-Haenszel log-odds-ratio estimate. Biometrics1987; 43: 425-31.
34.
Mehta CR, Walsh SJComparison of exact, mid-P, and Mantel-Haenszel confidence intervals for the common odds ratio across several 2 x 2 contingency tables. The American Statistician1992; 46: 146-50.
35.
Sato T.Confidence limits for the common odds ratio based on the asymptotic distribution of the Mantel-Haenszel estimator. Biometrics1990; 46: 71-80.
36.
Mehta CR, Patel NR, Gray R.Computing an exact confidence interval for the common odds ratio in several 2 x 2 contingency tables. Journal of the American Statistical Association1985;80:969-73.
37.
Vollset SE, Hirji KF, Elashoff RMFast computation of exact confidence limits for the common odds ratio in a series of 2 x 2 tables. Journal of the American Statistical Association1991; 86:404-409.
38.
Agresti A.A survey of exact inference for contingency tables (with discussion). Statistical Science1992; 7: 131-77.
39.
Fisher RAThe logic of inductive inference. Journal of the Royal Statistical Society, A 1935; 98: 39-54.
40.
Lancaster HOSignificance tests in discrete distributions . Journal of the American Statistical Association1961; 56: 223-34.
41.
Pratt JW, Gibbons JDConcepts of nonparametric theory. New York: Springer-Verlag, 1981.
42.
Barnard GAOn alleged gain in power from lower P-values. Statistics in Medicine1989; 8: 1469-77.
43.
Thomas D.Exact and asymptotic methods for the combination of 2 x 2 tables. Computers and Biomedical Research1975; 8: 423-46.
44.
Mehta CR, Patel NRA network algorithm for the exact treatment of the 2 x k contingency table . Communications in Statistics—Part B1980; 9: 649-64.
45.
Mehta CR, Patel NRA network algorithm for performing Fisher's exact test in r x c contingency tables. Journal of the American Statistical Association1983; 78: 427-34.
46.
Mehta CR, Patel NRA hybrid algorithm for Fisher's exact test on unordered r x c contingency tables. Communications in Statistics1986; 15: 387-403.
47.
Mehta CR, Patel NR, Tsiatis AAExact significance testing to establish treatment equivalence with ordered categorical data. Biometrics1984; 40: 819-25.
Barr CE, Colditz G., Chalmers TCEfficacy of alpha-interferon therapy of chronic carriers of the hepatitis B virus: a meta-analysis, 1990. (Unpublished report. Technology Assessment Group, Harvard School of Public Health, Boston, Massachusetts.)
50.
Antczak-Bouckoms A., Tullock Jfc, Adams M.Ambulatory cardiac monitoring for the evaluation of antiarrhythmic agents. International Journal of Technology Assessment in Health Care1993; 9: 124-38.
51.
Chalmers TC (ed). Data analysis for clinical medicine. Rome: International University Press, 1988.
52.
Innes JR, Ulland MG et al. Bioassay of pesticides and industrial chemicals for tumorigenicity in mice: A preliminary note. Journal of theNational Cancer Institute1969; 42: 1101-14.
53.
Fleiss JLConfidence intervals for the odds ratio in case-control studies: The state of the art. Journal of Chronic Diseases1979 ; 32: 69-77.
54.
Davis LJGeneralization of the Mantel-Haenszel estimator to nonconstant odds ratios . Biometrics1985; 41: 487-95.
55.
Laird NM, Mosteller F.Some statistical methods for combining experimental results. International Journal of Technology Assessment in Health Care1990; 6: 5-30.
