Beyond the hospital gates: Elucidating the interactive association of social support, depressive symptoms, and physical function with 30-day readmissions

Objective The aims of this study were to design patient-oriented risk models that elucidate the interactions among depressive symptoms, physical function, and sociodemographic markers and delineate the association of these indicators with 30-day readmissions. Design This prospective longitudinal study examined readmissions within 30 days of an index operation among adults (>18 yrs) undergoing outpatient surgery. The primary exposures were measures of physical function, depression, and sociodemographic factors. Multivariable logistic regression and hierarchical generalized linear mixed models with fixed and random effects tested interactions and association of these factors with 30-day readmissions. Results Adjusting for covariates and interaction terms, depressive symptoms (adjusted odds ratio, 2.50; 95% confidence interval, 1.45-4.29) and low physical function (adjusted odds ratio, 1.68; 95% confidence interval, 1.15-2.13) increased the odds of 30-day readmissions. Stratified by socioeconomic status (SES), those in the lowest SES group with depressive symptoms had a 73% increase in probability of readmissions (adjusted odds ratio, 2.67; 95% confidence interval, 1.10-4.07). Stratified by depressive symptom severity, being married or having a life partner lowered the readmission odds (adjusted odds ratio, 0.52; 95% confidence interval, 0.31-0.88). Final models had good discrimination and fitness (area under the curve, 0.710; corrected Akaike Information Criterion, 541.1). Conclusions Within 30 days of index surgery, depressive symptom severity moderated the association of sociodemographic, social support and physical function indicators with 30-day readmissions.