Development and Validation of a Multivariable Prediction Model for Postoperative Intensive Care Unit Stay in a Broad Surgical Population

Abstract

IMPORTANCE Despite limited capacity and expensive cost, there are minimal objective data to guide postoperative allocation of intensive care unit (ICU) beds. The Surgical Risk Preoperative Assessment System (SURPAS) uses 8 preoperative variables to predict many common postoperative complications, but it has not yet been evaluated in predicting postoperative ICU admission. OBJECTIVE To determine if the SURPAS model could accurately predict postoperative ICU admission in a broad surgical population. DESIGN, SETTING, AND PARTICIPANTS This decision analytical model was a retrospective, observational analysis of prospectively collected patient data from the 2012 to 2018 American College of Surgeons (ACS) National Surgical Quality Improvement Program (NSQIP) database, which were merged with individual patients' electronic health record data to capture postoperative ICU use. Multivariable logistic regression modeling was used to determine how the 8 preoperative variables of the SURPAS model predicted ICU use compared with a model inputting all 28 preoperatively available NSQIP variables. Data included in the analysis were collected for the ACS NSQIP at 5 hospitals (1 tertiary academic center, 4 academic affiliated hospitals) within the University of Colorado Health System between January 1, 2012, and December 31, 2018. Included patients were those undergoing surgery in 9 surgical specialties during the 2012 to 2018 period. Data were analyzed from May 29 to July 30, 2021.

Biased Study Design and Statistical Analysis in a Need for Intensive Care Unit Admission Surgical Prediction Model—Reply

“…In Reply We wish to thank Roepke and Besen for their interest in our recent publication and reply to their comments. The purpose of our study was to determine if we could accurately predict need for postoperative intensive care unit (ICU) use at our institution.…”

“…Finally, the authors thought that the statistical analysis introduced bias by use of stepwise forward variable selection, Hosmer-Lemeshow analysis to assess calibration, and absence of internal validation on the training sample and shrinkage techniques. The model of interest (Surgical Risk Preoperative Assessment System) selected all variables, so the odds ratios reported in Table 3 of our article would be identical to a model created not using stepwise forward selection. We find the Hosmer-Lemeshow approach useful particularly to show which groups might be considered as high risk for the potential application of risk-mitigating processes of care .…”

Biased Study Design and Statistical Analysis in a Need for Intensive Care Unit Admission Surgical Prediction Model

“…To the Editor We read with concern the article published in JAMA Surgery by Rozeboom and colleagues . The report is compliant with Transparent Reporting of a Multivariable Prediction Model for Individual Prognosis or Diagnosis, but it is a good example where this does not equate with study risk of bias .…”

Daily Load Leveling in Surgical Critical Care—The Tip of the Utilization Iceberg

“…With the COVID-19 pandemic placing tremendous strain on intensive care unit (ICU) capacity, optimizing critical care utilization has faced increasing scrutiny. A predictive model for ICU admission such as the one presented by Rozeboom et al may help hospital leadership with daily elective surgical schedule smoothing and reduce undesirable downstream effects of planned ICU admissions on emergency department diversion.…”

“…Reducing variability in complex surgical caseload, or load leveling, decreases strain on ICU capacity and may improve throughput . Rozeboom and colleagues developed a tool to predict postoperative ICU admission with excellent performance using just 7 clinical variables. The online calculator and parsimony of required data provide a convenient instrument for practical use, but we noted the composite variable missed 85% of patients who ultimately were admitted to the ICU after surgery.…”