S. E. Roberts1, C. J. Wirtalla1, P. Dowzicky1, R. Hoffman1, C. Aarons1, R. R. Kelz1 1Perelman School Of Medicine At University Of Pennsylvania,Department Of Surgery,Philadelphia, PA, USA
Background:
Patient and provider factors contribute to known racial disparities in length of stay (LOS) and readmission rates (RR). However, little is known about the role of LOS on RR in black patients. To gain a better understanding of the association between race and readmissions, we examined the odds of readmission for black patients based on the discharge efficiency.
Methods:
Discharge claims from California and New York were used to identify white and black patients undergoing a general surgical operation (2010-2011). Discharge efficiency (DE) was defined at the patient-level based on the distribution of LOS at the treating hospital for all patients undergoing the same operation and approach (laparoscopic vs open). Early discharge (ED) was assigned if LOS<25th%ile, routine discharge (RD) for LOS=25-75th%ile and late discharge (LD) for LOS>75th%ile. Multivariable mixed-effects logistic regression was used to examine the association between patient race and RR with control for potential confounders including DE and hospital level clustering. The analysis was stratified by operation category and DE. Bonferroni correction was used to correct for multiplicity.
Results:
Among 350,019 patients included in the study, 44,156 were black (12.6%). While the majority of patients were RD, a greater proportion of black patients were LD compared to white patients and a greater proportion of white patients were ED (B: 8.1%ED, 48.6%RD, 43.3%LD compared to W: 9.3%ED, 51.5%RD, 39.2%LD; p<0.001). The unadjusted RR for white patients was higher following general surgery operations than for black patients (W: 4.9 % and B: 4.1%; p<0.001). The adjusted odds of readmission were greater for black patients when controlling for potential confounders including DE and hospital-level clustering (OR: 1.11; 95%CI: 1.04-1.18). When examining the adjusted odds of readmission by DE group, the disparity was most pronounced among RD patients (OR: 1.13; 95%CI: 1.03 -1.23). When examining the adjusted odds of readmission by operation category, the odds were greatest for black patients who underwent complex operations (OR: 1.22; 95%CI: 1.04-1.43).
Conclusions:
Black patients are significantly more likely than white patients to be readmitted to the hospital following general surgical operations regardless of the hospital in which they receive their care. The risk is most pronounced among patients who undergo complex operations. Furthermore, the disparity is greatest for RD black patients. This suggests that targeted interventions in RD black patients may be able to ameliorate this racial disparity in surgical care.