Predictors of Successful Discharge of Patients on Postoperative Day 1 After Craniotomy for Brain Tumor

BACKGROUND: Shorter hospital stays have been associated with decreased complication rates, fewer hospital-acquired infections, and lower costs. We evaluated an optimized treatment paradigm for patients undergoing craniotomy allowing for postoperative day 1 (POD1) discharge if the criteria were met. We compared the complication and readmission rates between the POD1 patients and those with longer stays, and examined the patient and surgical variables for predictors of POD1 discharge. METHODS: We performed a retrospective review of craniotomies performed for tumor from 2011 to 2015. Craniotomies for tumors were included, and laser ablations and biopsies were excluded. RESULTS: A total 424 of patients were included, 132 (31%) of whom had been discharged on POD1. The mean length of stay was 6 days. The POD1 patients had had significantly better preoperative Karnofsky performance scale scores (P < 0.0001) and modified Rankin scale scores (P < 0.0001). Patient frailty, measured using the modified frailty index, was negatively predictive of POD1 discharge (P = 0.0183). Surgical factors predictive of early discharge were awake surgery (P < 0.0001) and supratentorial location (P < 0.0001). No POD1 patients experienced deep venous thrombosis (DVT), pulmonary embolus (PE), or urinary tract infections. However, of the patients with a length of stay >1 day, 4.4% and 2.7% developed DVT or PE (P = 0.0119) and urinary tract infections (P = 0.0202), respectively. Multivariate regression identified patient factors (male gender, low preoperative modified Rankin scale score), tumor factors (right-sided, supratentorial, smaller size), lower modified frailty index score, and operative factors (lack of a cerebrospinal fluid drain, awake surgery) as independent predictors of successful early discharge. CONCLUSIONS: Patients with good functional status can be safely discharged on POD1 after tumor craniotomy if the appropriate postoperative criteria have been met. Patients with early discharge had lower 30-day readmission and DVT/PE rates, likely owing to better baseline health status.