Five-Year Hospital Readmission After Coronary Artery Bypass Surgery and the Association With Off-Pump Surgery and Sex

Abstract

Background Limited data exist on long-term readmission and its association with patient and procedural characteristics after coronary artery bypass grafting. We aimed to investigate 5-year readmission after coronary artery bypass grafting and specifically focus on the role of sex and off-pump surgery. Methods and Results We performed a post hoc analysis of the CORONARY (Coronary Artery Bypass Grafting [CABG] Off or On Pump Revascularization) trial, involving 4623 patients. The primary outcome was all-cause readmission, and the secondary outcome was cardiac readmission. Cox models were used to investigate the association of outcomes with sex and off-pump surgery. Hazard function for sex was studied over time using a flexible, fully parametric model, and time-segmented analyses were performed accordingly. Rho coefficient was calculated for the correlation between readmission and long-term mortality. Median follow-up was 4.4 years (interquartile range, 2.9-5.4 years). The cumulative incidence rates of all-cause and cardiac readmission were 29.4% and 8.2% at 5 years, respectively. Off-pump surgery was not associated with either all-cause or cardiac readmission. The hazard for all-cause readmission in women over time was constantly higher than the hazard for men (hazard ratio [HR], 1.21 [95% CI, 1.04-1.40]; P=0.011). Time-segmented analyses confirmed the higher risk for all-cause (HR, 1.21 [95% CI, 1.05-1.40]; P<0.001) and cardiac (HR, 1.26 [95% CI, 1.03-1.69]; P=0.033) readmission in women after the first 3 years of follow-up. All-cause readmission was strongly correlated with long-term all-cause mortality (Rho, 0.60 [95% CI, 0.48-0.66]), whereas cardiac readmission was strongly correlated with long-term cardiovascular mortality (Rho, 0.60 [95% CI, 0.13-0.86]). Conclusions Readmission rates are substantial at 5 years after coronary artery bypass grafting and are higher in women but not with off-pump surgery. Registration URL: http://www.clinicaltrials.gov/; Unique identifier: NCT00463294.

Keywords: clinical trial; coronary artery bypass grafting; epidemiology; readmission; rehospitalization.

Figure 1. Cumulative incidence of all‐cause (A) and cardiac (B) readmission in the overall cohort during the follow‐up.

Figure 2. Cumulative incidence of all‐cause (A) and cardiac (B) readmission in the on‐ and off‐pump groups.

HR indicates hazard ratio.

Figure 3. Variation in the hazard ratio for all‐cause readmission in women vs men during the follow‐up years.

The dashed line indicates the no‐effect line, where the hazard ratio=1. Values >1 indicate a higher risk for women, and values <1 indicate a higher risk for men. The gray shaded area represents the 95% CI.

Figure 4. Time‐segmented analysis for all‐cause (A) and cardiac (B) readmission in women vs men.

HR indicates hazard ratio.