Abstract
OBJECTIVES: To provide nursing home (NH)-specific estimates to assess whether venous thromboembolism (VTE) risk factors identified for the general population apply to NH residents. DESIGN: Population-based case-control study. SETTING: Olmsted County, Minnesota. PARTICIPANTS: All county residents with symptomatic objectively diagnosed incident VTE while resident in a NH from 1988 through 2000 (N = 182) and two age-, sex-, calendar-year-matched non-VTE Olmsted County NH residents per case (N = 364). MEASUREMENTS: Provider-linked medical records were reviewed to obtain information on active malignancy and recent hospitalization, surgery, trauma, or fracture as of index date (case's VTE date; respective provider registration date for controls). Risk factor prevalence and VTE-associated odds ratios (OR) were estimated and compared with previously obtained data for all Olmsted County residents from 1988 through 2000. For analyses, both groups were limited to individuals aged 65 and older. RESULTS: In NH residents, active malignancy, recent hospitalization, and recent surgery significantly increased VTE risk, but the magnitude of risk appeared much lower than general population estimates (e.g., for major surgery, OR = 2.5, 95% confidence interval (CI) = 1.4-4.3 for NH residents vs OR = 11, 95% CI = 7.0-17 for general population). In general, the prevalence of all evaluated VTE risk factors appeared much higher in NH controls than in general population controls. Thromboprophylaxis rates appeared higher for NH cases and controls than in the general population; disconcertingly, 47% of NH cases received prophylaxis. CONCLUSION: Although general population VTE risk factors (active cancer and recent hospitalization or surgery) can identify NH residents at higher risk for VTE, these exposures do not adequately stratify VTE risk for thromboprophylaxis recommendations. Further research into NH-specific risk factors and prophylaxis effectiveness is required.