Analytic strategies for recurrent events in epidemiologic studies: background and application to hospitalization risk in the elderly

Due to the intraindividual dependence, specific analytic strategies are nee ded to assess risk factors for recurrent events. Although well established in the biostatistics literature, applications of these techniques are almos t nonexistent in the field of epidemiology. The authors applied four differ ent regression approaches for recurrent events (logistic, Poisson, and two different Cox proportional hazards regressions) to derive rate ratios of ho spitalizations for various prognostic factors in a cohort of 2424 frail eld erly. Over a median follow-up of 670 days, 3299 hospitalizations were obser ved in 1564 persons. Estimated rate ratios were similar in all four approac hes and virtually identical in three. With all methods, confidence interval s of the rate ratios were considerably wider than with naive Poisson regres sion neglecting intraindividual dependence of events. Appropriate analysis of recurrent events is feasible with minor modifications of multivariable m odels familiar to epidemiologists and should no longer be neglected in epid emiologic research. In our setting, Poisson regression was the most conveni ent approach. (C) 2000 Elsevier Science Inc. All rights reserved.