Jeffrey D Howard1, Mickey S Ising2, Megan E Delisle3, Robert C G Martin4. 1. University of Louisville, Hiram C. Polk Dept. of Surgery, Louisville, KY, 40202, United States. Electronic address: jeffrey.howard@ky.gov. 2. University of Louisville, Hiram C. Polk Dept. of Surgery, Louisville, KY, 40202, United States. Electronic address: mickeyising@gmail.com. 3. University of Manitoba, Dept. of Surgery, 2009-311 Hargrave St, Winnipeg, MB, R3B 0V8, Canada. Electronic address: delislem@myumanitoba.ca. 4. University of Louisville, Hiram C. Polk Dept. of Surgery, Louisville, KY, 40202, United States. Electronic address: robert.martin@louisville.edu.
Abstract
BACKGROUND: Appropriate postoperative readmission rates and modifiable risk factors for readmission have yet to be defined for many operations. This systematic review and meta-analysis attempt to define these parameters for pancreaticoduodenectomy. MATERIALS AND METHODS: The main outcomes were readmission rate, risk factors, and reasons for readmission. Meta-analyses were performed when data was homogeneous, otherwise, a qualitative review was performed. RESULTS: The 30-day, 90-day, and overall readmission rates were 17.63%, 26.14%, and 27.18%, respectively. In the meta-analysis, chronic pancreatitis (OR, 1.44, p = 0.04), operative length (MD, 26.1; p < 0.01), wound infection (OR, 1.9, p < 0.01), intra-abdominal abscess (OR, 3.79, p < 0.01), VTE (OR, 2.27, p = 0.01), and LOS (MD, 1.66, p < 0.01) where associated with readmission. CONCLUSION: Hospital readmission will continue to be a quality metric and will influence reimbursement models. Thirty and 60-day readmission data underestimate the true readmission rate. Chronic pancreatitis, operative length, and several post-operative complications were associated with greater readmission. More uniform reporting is necessary to identify modifiable risk factors associated with readmission.
BACKGROUND: Appropriate postoperative readmission rates and modifiable risk factors for readmission have yet to be defined for many operations. This systematic review and meta-analysis attempt to define these parameters for pancreaticoduodenectomy. MATERIALS AND METHODS: The main outcomes were readmission rate, risk factors, and reasons for readmission. Meta-analyses were performed when data was homogeneous, otherwise, a qualitative review was performed. RESULTS: The 30-day, 90-day, and overall readmission rates were 17.63%, 26.14%, and 27.18%, respectively. In the meta-analysis, chronic pancreatitis (OR, 1.44, p = 0.04), operative length (MD, 26.1; p < 0.01), wound infection (OR, 1.9, p < 0.01), intra-abdominal abscess (OR, 3.79, p < 0.01), VTE (OR, 2.27, p = 0.01), and LOS (MD, 1.66, p < 0.01) where associated with readmission. CONCLUSION: Hospital readmission will continue to be a quality metric and will influence reimbursement models. Thirty and 60-day readmission data underestimate the true readmission rate. Chronic pancreatitis, operative length, and several post-operative complications were associated with greater readmission. More uniform reporting is necessary to identify modifiable risk factors associated with readmission.
Authors: Alexandra W Acher; James R Barrett; Patrick B Schwartz; Chris Stahl; Taylor Aiken; Sean Ronnekleiv-Kelly; Rebecca M Minter; Glen Leverson; Sharon Weber; Daniel E Abbott Journal: J Gastrointest Surg Date: 2020-07-15 Impact factor: 3.452