Readmission Rates and Gastrointestinal Surgeries

Orlando—After controlling for patient and operative factors, the authors in a study of patients undergoing major gastrointestinal surgeries found numerous factors were associated with an increased risk of 30-day unplanned readmissions. The factors included patients who took steroids for chronic conditions (odds ratio [OR], 1.67), an operation time of at least 4 hours (OR, 1.45), discharge to a facility other than the home (OR, 1.37), predischarge complications (OR, 1.28), an American Society of Anesthesiologists Physical Status Classification of 3, 4, or 5 (OR, 1.27), dependent functional status (OR, 1.22), and pancreatic resection (OR, 1.15).

Kristin N. Kelly, MD, the study’s lead author, presented the results at DDW during a plenary session. She suggested that telephone follow-up, early postoperative clinic appointments, visiting nurse services, and outpatient treatment of minor complications could help target complications in the early days after patients undergo surgery.

Dr. Kelly noted that several studies have shown high readmission rates lead to billions of dollars in excess healthcare spending. The soaring costs have contributed to decreasing reimbursements for excessive rates of readmission for pneumonia, myocardial infarction, and coronary heart failure.

By identifying the etiology of readmissions, such as complications, lack of preparation for discharge, or a social situation at home, Dr. Kelly said surgeons and physicians could improve care and decrease the percentage of patients requiring readmission. She added that most gastrointestinal surgeries are performed electively. Readmission rates after major gastrointestinal operations are between 10% and 20%, with patients typically returning to the hospital after experiencing gastrointestinal problems, electrolyte or metabolic issues, infections, or obstructions.

Most of the studies examining readmissions have been small and from single center data or from large administrative databases, according to Dr. Kelly. This study was the first to collect data on unplanned readmissions following gastrointestinal surgeries. The authors used the 2011 American College of Surgeons National Surgical Quality Improvement Program (ACS NSQIP^®) database. They searched for major gastrointestinal resections and grouped patients by the type of resections, including foregut (81% of patients), liver (6%), pancreas (10%), and intestinal (3%). Patients were excluded if they died before discharge or stayed in the hospital for >30 days.

The authors analyzed data from 42,609 patients undergoing a major gastrointestinal resection: 12.3% of the patients had an unplanned readmission. The groups were similar in terms of age and sex, but patients who had an unplanned readmission were significantly more likely to have a comorbidity, a higher American Society of Anesthesiologists Physical Status Classification, and a body mass index of at least
30 kg/m².

More than 80% of patients underwent elective surgeries. Patients who were readmitted to the hospital within 30 days were significantly more likely to have emergency surgery or laparoscopic surgery, a longer operative time, postoperative length of stay, and a predischarge minor or major complication. The mean postoperative length of stay was 8.0 days in the unplanned readmission group and 7.4 days for patients who were not readmitted (P<.0001).

The readmission rates based on type of resections were 11.8% for intestinal, 12.7% for liver, 13.5% for foregut, and 16.3% for pancreas. Overall, major complications within 30 days of discharge were found in 26.0% of patients: 23.7% for intestinal, 30.8% for liver, 36.1% for foregut, and 38.7% for pancreas.

Dr. Kelly said the study had a few limitations. The authors did not have access to the causes or days to readmission for all of the institutions and did not know some of the variables that could determine other readmissions for these patients, including gastrointestinal complications, dehydration, and social factors. In addition, the findings represented hospitals in the NSQIP database and may not be generalizable to other institutions, according to Dr. Kelly. Also, a retrospective database does not allow the authors to examine information about the appropriateness or quality of care or the patients’ knowledge, attitudes, and beliefs regarding their readmissions.