(398) Preoperative Fasting and Ventilator-Free Survival in Intubated Patients: A Retrospective Study

Introduction: Preoperative fasting is a potential risk factor for malnutrition in critical illness. To assess whether fasting to prevent pulmonary aspiration is necessary in intubated patients, we examined the association between duration of preoperative fasting and days alive and free from ventilator (DAFV) at post-operative day 28 (POD28).

Methods: We conducted a retrospective study of adult non-burn ICU patients at Massachusetts General Hospital from 06/2016 to 06/2022. Included patients underwent surgery during ICU stay, were mechanically ventilated within 12 hours (h), and received tube feeds within 48h preoperatively. Exclusion criteria were gastrointestinal, airway, and emergency procedures, non-supine positioning, and postoperative extubation. Duration of preoperative fasting was categorized as: short ( < 8h), intermediate (8-16h), or long (>16h). The primary outcome, DAFV on POD28, was defined as 28 minus last ventilation day. In patients who were ventilated or died at POD28, DAFV was 0. We assessed the effects of fasting duration on DAFV using count and zero-inflated multivariate regression models with age, sex, race, BMI, total and preoperative ICU length of stay, number of procedures, and ASA status as co-variables.

Results: Patients (n=424) in the short (n=79), intermediate (n=246), and long (n=99) fasting groups had comparable median age (58.5, 59.6, 60.9 years), % of women (40.5, 41.9, 43.4%), mean BMI (29.3, 28.5, 28.8 kg/m2), mean number of procedures (1.5, 1.4, 1.5), and median preoperative ICU stay (6.4, 8.8, 9.8 days). 28-day mortality was 26.2% (n=111). Median (IQR) fasting duration was 12.2h (8.7-15.5) for all patients: 4.3h (1.9-6.5) in the short; 12.0h (9.9-13.7) in the intermediate, and 24.0h (18.4-33.3) in long fasting group. Median (IQR) DAFV at POD28 was 10 (0-21) in the short, 9 (0-23) in the intermediate, and 7 (0-21) in the long fasting group. Fasting duration was not associated with DAFV by POD28 in multivariable models: (1) short vs intermediate fasting (count OR 1.06, 95% CI 0.92-1.23; zero-inflated OR 0.73, 95% CI 0.40-1.33); (2) short vs long fasting (count OR 1.05, 95% CI 0.88-1.24; zero-inflated OR 0.66, 95% CI 0.33-1.33).

Conclusions: Our findings do not support prolonged preoperative fasting to prevent perioperative pulmonary aspiration in intubated patients.