How to use multiple logistic regression in retrospective database analyses

Accurately measuring the effectiveness of medical or pharmacological interv entions is essential in economic evaluations. Many economic evaluations are performed retrospectively using medical and pharmacy claims databases. A c hallenge to using this approach is that variation in treatment patterns pre vents simple comparisons between patient groups. Multiple logistic regressi on is a powerful technique that can overcome this challenge and is suited f or evaluating dichotomous outcomes (eg, disease/no disease). An example is presented that applies multiple logistic regression to determine the impact of different drug therapies on hospitalization risk for heart failure amon g patient with congestive heart failure.