Exploring the burden of inpatient readmissions after major cancer surgery

Abstract

Purpose: Travel distances to care have increased substantially with centralization of complex cancer procedures at high-volume centers. We hypothesize that longer travel distances are associated with higher rates of postoperative readmission and poorer outcomes.

Methods: SEER-Medicare patients with bladder, lung, pancreas, or esophagus cancer who were diagnosed in 2001 to 2007 and underwent extirpative surgery were included. Readmission rates and survival were calculated using Kaplan-Meier functions. Multivariable negative binomial models were used to examine factors associated with readmission.

Results: Four thousand nine hundred forty cystectomies, 1,573 esophagectomies, 20,362 lung resections, and 2,844 pancreatectomies were included. Thirty- and 90-day readmission rates ranged from 13% to 29% and 23% to 43%, respectively, based on tumor type. Predictors of readmission were discharge to somewhere other than home, longer length of stay, comorbidities, higher stage at diagnosis, and longer travel distance (P < .001 for each). Patients who lived farther from the index hospital also had increased emergency room visits and were more likely to be readmitted to a hospital other than the index hospital (P < .001). Of readmitted patients, 31.9% were readmitted more than once. Long-term survival was worse and costs of care higher for patients who were readmitted (P < .001 for all).

Conclusion: The burden of readmissions after major cancer surgery is high, resulting in substantially poorer patient outcomes and higher costs. Risk of readmission was most strongly associated with length of stay and discharge destination. Travel distance also has an impact on patterns of readmission. Interventions targeted at higher risk individuals could potentially decrease the population burden of readmissions after major cancer surgery.

Fig 1.

(A) Travel distances by SEER region, disease site, and index hospital volume. (B) Travel distances in the Southwest region are presented separately because of the large difference in range of values compared with other regions. Outlier values are not shown. Horizontal reference line designates 25 miles.

Fig 2.

Readmission rates by disease site.

Fig 3.

Kaplan-Meier estimates of survival by readmission group for (A) bladder cancer, (B) lung cancer, (C) pancreas cancer, and (D) esophagus cancer.

Fig A1.

Patient cohort selection. HMO, health maintenance organization.