OBJECTIVES: The NAVIGATOR trial will evaluate the impact of nateglinide and valsartan on the progression to diabetes in subjects with impaired glucose tolerance at high risk for a cardiovascular event. Investigators in this trial include endocrinologists, cardiologists and general internists. Although economic evaluations of clinical trials typically adjust for country, little is known about how physician-specific characteristics may affect ALOS. Our objective was to determine which physician characteristics would be independent predictors of ALOS for myocardial infarction (MI). METHODS: We analyzed responses from a survey of NAVIGATOR investigators to determine factors associated with self-reported ALOS in the treatment of MI. Candidate predictors in the multiple regression model were physician age, sex, years since graduation, medical specialty, practice environment, and geographic region. This relationship was examined using linear regression. RESULTS: 385 respondents with complete data were included in the model. Physician age, sex, years since graduation, specialty of internal medicine, practice setting (solo or group), and geographic region met the criterion for inclusion in the multivariable model. Male physicians reported an ALOS 3 days shorter than female doctors (p=<.0001). Physicians in group practice were associated with a 1-day longer ALOS as compared to hospital-affiliated physicians (p=<.0001). All geographic regions except for Africa and Canada were significantly associated with longer ALOS than the US (Asia Pacific 6 days, Eastern Europe 9, Europe 3.5, Latin America 2) (p=<.0001 for all). CONCLUSIONS: For NAVIGATOR investigators, physician sex, practice setting and geographic region are predictors of ALOS. We will adjust for physician characteristics in the economic analysis of the NAVIGATOR trial. We conclude that measuring physician characteristics should be incorporated into multinational clinical trials.