Can We Predict Unplanned Intensive Care Unit Admission in Hip and Knee Arthroplasty?

Abstract

Preoperative risk stratification remains important for both patient safety and cost-effective allocation of resources in total joint arthroplasty. Limited literature exists investigating risk factors for unplanned intensive care unit (ICU) admission following arthroplasty. Therefore, the purpose of this study was to assess a broader spectrum of factors that may increase the risk of ICU admission across multiple surgical settings. We retrospectively reviewed 18,169 patients who underwent a hip (40.7%) or knee (59.3%) arthroplasty procedure (primary: n = 13,384, 90%; revision: n = 1,485, 10%) between August 1, 2015, and January 31, 2019, across nine sites at a single institution. Patient demographics, laboratory data, surgical parameters, and preoperative scores on the Hip Disability and Osteoarthritis Outcome Score (HOOS) or Knee Injury and Osteoarthritis Outcome Score (KOOS) and the Veterans RAND 12-Item Health Survey (VR-12) were recorded. A univariate analysis was conducted to identify differences between patients who required an ICU admission and those with no ICU involvement. Logistic regression was then used to generate predicted risk of ICU Involvement. A total of 147 patients (0.99%) required an ICU admission. Increasing age, higher Charlson's comorbidity index, American Society of Anesthesiologists (ASA) class 3 or greater, elevated blood urea nitrogen (BUN), elevated creatinine, decreased preoperative hemoglobin level, general anesthesia, surgical setting, revision procedure, and lower baseline scores on the KOOS Quality of Life (QOL) subscale (17 vs. 22.5, p = 0.032), VR-12 physical component score (PCS; 24.4 vs 27.1, p = 0.001), and VR-12 mental component score (MCS; 45.9 vs. 49.6, p = 0.003) were associated with a statistically significant increase in the risk of ICU admission. As performance of arthroplasty shifts in both the setting of care and the complexity of patients, identifying patients most at risk of requiring higher acuity of care is important. Preoperative assessment of comorbidities, kidney function, surgical setting, KOOS QOL subscale score, and VR-12 scores may provide guidance for the improvement of value-based care pathways.