Perioperative Predictors of Discharge to an Inpatient Rehabilitation Facility Following Coronary Artery Bypass Grafting: A Single-Center Retrospective Review

Background and/or Objectives: Patients undergoing coronary artery bypass grafting (CABG) can experience a significant post-operative decrease in cardiopulmonary capacity with associated limitations in ambulation and loss of functional independence. This decline in a patient’s functional level can be significant enough to warrant admission to an inpatient rehabilitation facility (IRF) prior to going home. This study will aim to assess specific patient characteristics that can provide early indications for which patients are likely to require admission to an IRF in hopes of optimizing efficiency and safety with transitions of care.

Design: This is a retrospective cohort study with data obtained through electronic medical record queries and manual chart review. The variables analyzed included sex, age, body mass index (BMI), acute care length of stay (LOS), number of vessels operated on, presence of a concomitant valve replacement, Charlson Comorbidity Index (CCI) score, initial AM-PAC 6-clicks mobility score, and insurance provider (Medicare vs other).

Setting: Tertiary-Care University Hospital

Participants: The study population consisted of 1281 patients who underwent CABG from 2018-2021 and had a documented AM-PAC 6-clicks mobility score upon admission.

Interventions: Logistic regression analysis

Main Outcome Measures: Discharge to an IRF, a skilled nursing facility (SNF), or home.

Results: Based on the adjusted logistic regression model, age (OR=1.5, p=0.0125), acute care LOS (OR=1.2, p= < 0.0001), having Medicare (OR=2.0, p=0.0076), and lower initial mobility scores were significantly associated with higher odds for discharge to IRF. These same variables along with female gender (OR=2.445, p= < 0.0001) and a higher CCI score (OR 1.181, p=0.001) were significantly associated with higher odds for discharge to SNF. When comparing the odds of discharge to IRF instead of SNF there were two significant variables: female gender (OR=2.361, p=0.0052) and longer acute care LOS (OR=0.933, p= < 0.0001).

Conclusions: Identification of these perioperative predictors has the potential to assist clinicians and physiatrists with early discharge planning and improved continuity of care.