An algorithm for prospective individual matching in a non-randomized clinical trial

A method is described to achieve balance across prognostic factors in inter vention trials for which randomized allocation to treatment group is not po ssible. The method involves prospective individual matching of patients tha t have already been assigned to treatment groups. Data can be analyzed usin g methods appropriate for prospective matched cohort studies. Successful im plementation depends on the number and complexity of factors to be matched, and on the number of available control patients. Simulation studies sugges t that, in order to yield satisfactory match rates and to reduce costs asso ciated with screening unmatched controls, no more than three prognostic fac tors should generally be considered. Baseline prognostic indices, incorpora ting information from multiple variables, provide effective matching factor s. The implementation of the method in a successful clinical trial, the Del irium Prevention Trial, is discussed. In that study, treatment group was de termined by hospital admission to either an intervention floor or to one of two usual care hospital floors. The ratio of available control to interven tion patients was 1.3, and 95% of the eligible intervention floor patients were successfully matched to control floor patients. Excellent balance was demonstrated for non-matching factors, due in part to the use of a composit e baseline risk score as a matching factor. In addition, external validity is enhanced because most eligible intervention patients are enrolled as the y present. The methods outlined in this report provide a methodologically r igorous alternative for achieving balance across treatment groups, with res pect to important prognostic factors, in non-randomized clinical trials. an d will have broad applicability in the numerous situations in which randomi zation is not possible. (C) 2001 Elsevier Science Inc. All rights reserved.