Data Collection
Data was collected retrospectively from the electronic medical record. Demographic and baseline data collected include sex, age, body mass index (BMI), race, tobacco use, and past medical history including diabetes mellitus, hypertension, and hyperlipidemia. Predicted operative mortality was calculated for each patient, accounting for patient risk factors, using the European System for Cardiac Operative Risk Evaluation II (EuroSCORE II) 22. Intraoperative data collected include LVAD device type, transesophageal echocardiography (TEE) time, cardiopulmonary bypass (CPB) time, total operative time, and total intubation time including during the postoperative period. The intent of therapy as bridge to transplant or destination therapy was also recorded. Total intubation time was recorded for the primary intubation event from the immediate preoperative intubation time to postoperative extubation time.
Dysphagia variables collected include whether a speech-language pathologist (SLP) was consulted, whether a barium swallow study was performed and time to swallow study following surgery, and dysphagia status and severity. Dysphagia status was recorded as a binary variable, and severity was recorded as described in the electronic medical record as mild, moderate, or severe.
Health-related outcomes recorded include total time NPO postoperatively, intensive care unit (ICU), postoperative and total length of stay (LOS), incidence of pneumonia and sepsis, 30-day readmission, and 30-day, 1-year, 2-year, and 3-year mortality. Total time NPO was defined as the time from surgery to the first diet order by mouth excluding orders for ice chips. Pneumonia was recorded when the medical record documented clinical suspicion, supporting chest x-ray findings and clinical features, and a plan of management.