Factors influencing readmission after curative gastrectomy for gastric cancer

Abstract

Background: The incidence of, and associated risk factors for, readmission after potentially curative gastrectomy for patients with gastric cancer has not been well studied. We sought to determine the 30-day readmission rate as well as the potential risk factors for readmission at our institution in patients undergoing gastrectomy for gastric cancer with curative intent.

Study design: We performed a retrospective analysis of all patients undergoing potentially curative gastrectomy for gastric cancer from 1995 to 2011. The 30-day hospital readmission rate was determined, and potential clinicopathologic risk factors for readmission were examined.

Results: Readmission to the hospital within 30 days occurred in 14.6% (61 of 418) of patients, including 6 patients who were readmitted more than once. The most common reasons for readmission included nutritional difficulties (n =12, 20%), intra-abdominal fluid collections (n = 11, 18%), and small bowel obstruction (n = 6, 10%). Factors associated with a higher 30-day readmission rate included type of resection (total gastrectomy, 23% vs subtotal gastrectomy, 13% vs esophagogastrectomy, 9%, p = 0.016), pre-existing cardiovascular disease (17%, p = 0.05), and history of a major postoperative complication (24%, p < 0.001). Factors not associated with a higher readmission rate included advanced age, pre-existing pulmonary disease, T or N stage, extent of lymph node dissection, receipt of neoadjuvant chemotherapy or radiotherapy, length of stay of the index hospitalization, and destination and level of support on discharge.

Conclusions: Readmission after potentially curative gastrectomy for gastric cancer is common. Patients with pre-existing cardiovascular disease, those who suffer major postoperative complications, and those undergoing total gastric resections are at especially high risk for readmission, and strategies designed to support these high-risk patients on discharge are warranted.