Matching is occasionally used in cohort studies; examples include studies of twins and some studies of traffic crashes. Analysis of matched cohort data is not discussed in many textbooks or articles and is not mentioned in the Stata manuals. Risk ratios can be estimated using matched-pair cohort data with Stata's mcc command. We describe a new command, csmatch, which can produce these risk ratios and is often more convenient. We briefly review flexible regression methods that can estimate risk ratios in matched cohort data: conditional Poisson regression and some versions of Cox regression.
CummingsP., McKnightB., and GreenlandS.2003. Matched cohort methods in injury research. Epidemiologic Reviews25: 43–50.
2.
CummingsP., McKnightB., and WeissN.S.2003. Matched-pair cohort methods in traffic crash research. Accident Analysis and Prevention35: 131–141.
3.
KoepsellT. D., and WeissN.S.2003. Epidemiologic Methods: Studying the Occurrence of Illness.Oxford: Oxford University Press.
4.
LynskeyM. T., HeathA. C., BucholzK. K., SlutskeW. S., MaddenP. A. F., NelsonE. C., StathamD. J., and MartinN.G.2003. Escalation of drug use in early-onset cannabis users vs. co-twin controls. Journal of the American Medical Association289: 427–433.
5.
MantelN., and HaenszelW.1959. Statistical aspects of the analysis of data from retrospective studies. Journal of the National Cancer Institute22: 719–748.
6.
McNemarQ.1947. Note on the sampling error of the difference between correlated proportions or percentages. Psychometrika12: 153–157.
7.
NurminenN.1981. Asymptotic efficiency of general noniterative estimation of common relative risk. Biometrika68: 525–530.
8.
RothmanK., and GreenlandS.1998. Modern Epidemiology.2nd ed. Philadelphia: Lippincott–Raven.
