STUDY DESIGN

Prospective analysis of resource utilization (RU) after lumbar surgery.

OBJECTIVE

To determine contributions of patient psychosocial variables to RU, specifically length of stay (LOS) and discharge destination.

SUMMARY OF BACKGROUND DATA

Associations between demographic/clinical variables and RU are well-established; less is known about associations between psychosocial variables, such as social support and depressive/anxiety symptoms, and RU.

METHODS

Preoperatively 532 patients provided standard ratings of depressive and anxiety symptoms and amount of social support. Clinical variables included calculation of a composite score reflecting invasiveness of surgery. Two outcomes were considered: LOS ≤ 3 or > 3 days; and discharged home without external services or discharged with external services. Independent variables in multivariable logistic models were demographic/clinical and psychosocial variables.

RESULTS

Mean age was 56 years, 55% were men, 87% had degenerative conditions, 40% had a positive depression screen, 62% had anxiety greater than population norms, and 77% reported as much social support as wanted. Variables that remained associated with a LOS > 3 days (53%) in multivariable analysis were: older age (OR 1.8, confidence interval, CI, 1.2-2.8); female sex (OR 1.9, CI 1.2-2.9), less social support (OR 1.4, CI 1.1-1.7); obesity (OR 1.8, CI 1.1-2.8), worse functional status (OR 1.0, CI 1.0-1.0), and more invasive surgery (OR 3.0, CI 2.4-3.7). Variables that remained associated with discharged with external services (17%) in multivariable analysis were: older age (OR 3.3, CI 1.9-5.9); less social support (OR 1.3, CI 1.0-1.7); positive depression screen (OR 1.9, CI 1.1-3.4); more anxiety (OR 2.4, CI 1.3-4.2); major comorbidity (OR 1.7, CI 1.0-2.8); and more invasive surgery (OR 1.9, CI 1.5-2.5).

CONCLUSION

This study confirms the importance of demographic/clinical variables as predictors of greater RU after lumbar surgery and provides preliminary evidence for the importance of concurrent psychosocial variables, particularly less social support, and worse psychological well-being.

LEVEL OF EVIDENCE