IMPACT OF MARITAL-STATUS ON OUTCOMES IN HOSPITALIZED-PATIENTS - EVIDENCE FROM AN ACADEMIC-MEDICAL-CENTER

Background: Prior studies have described the importance of social supp ort on long-term patient outcomes. Few studies have investigated the i mpact of social support on outcomes in hospitalized patients. Objectiv e: To examine the relationship between marital status, an important as pect of social support, and several hospital outcomes. Methods: Patien ts included 40 820 adult medical and surgical patients discharged from a midwestern academic medical center during 1988 through 1991, of who m 21 291 were unmarried and 19 529 were married. Using multivariable r egression analyses, we compared the following outcomes in married and unmarried patients: rate of in-hospital death, rare of nursing home di scharge, length of stay, and hospital charges. Severity of illness was measured using a previously validated commercial method. Results: Adm ission severity of illness was higher in unmarried than married patien ts; 40% of unmarried patients had moderate or high severity compared w ith 32% of married patients. In a series of multivariable analyses, co ntrolling for severity of illness, age, gender, race, and diagnosis, t he risk of nursing home discharge was more than 2.5 times greater for unmarried than for married patients (multivariable odds ratio, 2.67; 9 5% confidence interval, 2.33 to 3.06), while the risk of in-hospital d eath for unmarried compared with married patients was higher among sur gical patients (odds ratio, 1.30; 95% confidence interval, 1.06 to 1.5 8) but not among medical patients (odds ratio, 0.98; 95% confidence in terval 0.84 to 1.15). In additional analyses, multivariable models est imated that hospital charges and length of stay were 5% and 8% higher (P<.001), respectively, for unmarried than for married patients. In a series of stratified analyses, the above differences among unmarried p atients tended to be greater for patients who were never married than for patients who were widowed, divorced, or separated. Conclusions: Th e findings suggest that marital status was an independent risk factor for several important hospital outcomes. This adds to our understandin g of the importance of social support and other nonbiological factors on outcomes in hospitalized patients. This also has implications for t he design of hospital-based interventions to improve patient outcomes and for the development of equitable prospective and capitated hospita l payment formulas.