Unplanned hospital admission after ambulatory surgery: a retrospective, single cohort study.

Abstract

We estimated the rate of unplanned hospital and intensive care unit (ICU) admissions following ambulatory surgery centre (ASC) procedures, and identified factors associated with their occurrence. This retrospective cohort included adult patients who underwent ASC procedures within a large community practice from January 2010 to December 2014. Patients were categorized into two groups: unplanned postoperative hospital/ICU admission within 24 hr of procedure or uneventful discharge. Demographics, comorbidities, anesthesia type, procedure type, procedure group, and ASC facility were assessed. Of the 211,389 patients included, there were 211,147 uneventful discharges (99.89%) and 242 unplanned hospital admissions (0.11%), of which 75 were ICU admissions (0.04%). The multivariable logistic regression model for hospital admission showed an increased risk associated with age > 50 yr (odds ratio [OR], 1.53); American Society of Anesthesiologists (ASA) physical status (III vs II: OR, 1.45; IV vs II: OR, 1.88), comorbidity (chronic obstructive pulmonary disease: OR, 2.63; diabetes mellitus: OR, 1.62; transient ischemic attack: OR, 2.48) procedure (respiratory: OR, 2.92; digestive: OR, 2.66; musculoskeletal system: OR, 2.53), anesthetic management (general anesthesia [GA] and peripheral nerve block vs GA: OR, 1.79), and ASC facility (189BB: OR, 2.29; 30E9A: OR, 7.41; and BD21F: OR, 1.69). The multivariable logistic regression model for ICU admission showed increased risk of unplanned ICU admission associated with ASA physical status (ASA III vs II: OR, 3.0; ASA IV vs II: OR, 8.52), procedure (musculoskeletal system: OR, 2.45), and ASC facility (00E6C: OR, 3.14; 189BB: OR, 2.77; 30E9A: OR, 2.59; and BD21F: OR, 3.71). While a small percentage of adult patients who underwent ASC procedures required unplanned hospital admission (0.07%), approximately one-third of these admissions were to the ICU (0.04%). Facility was at least as strong a predictor of hospital admission as the patient- and/or procedure-specific variables.