Venous thromboembolism: regional differences in the nationwide inpatient sample, 1993 to 2000.

Venous thromboembolism (VTE) is a costly complication of hospitalization. The sequelae make it a concern for public health planners. The Nationwide Inpatient Sample (NIS) contains data for hospital discharges in the United States. These data were reviewed to determine their suitability for health policy planning. International Classification of Diseases, Ninth Revision, Clinical Modification codes for VTE were applied to the NIS data. The sample was queried for demographic information, mortality, length of hospital stay, diagnosis, and treatment. The rates were standardized for geographic region and disease acuity. Statistical analysis included descriptive reporting of means and event rates; analysis of variance and logistic regression were used for regional effects and modeling of mortality. Between 1993 and 2000, 636,814 discharges involved VTE (1.2%). This rate was consistent over time and within regions. Regional differences existed in the acceptance of new technology and hospital charges. Mortality varied from 6.3% (Midwest) to 7.9% (Northeast) and was associated with admission type, comorbidities, pulmonary embolism, and discharge from the Northeast region. White race, chronic venous insufficiency, and female gender were protective variables. The NIS data report a consistent mortality rate despite improved therapy. Regional diagnostic, treatment, and economic differences exist. The data are useful for the purposes of public health care planning and stimulating clinical trial questions.