Preoperative frailty predicts adverse short-term postoperative outcomes in patients treated with radical nephroureterectomy

Background: To investigate the effect of frailty on short-term postoperative outcomes and total hospital charges (THCs) in patients with non-metastatic upper urinary tract carcinoma, treated with radical nephroureterectomy (RNU). Methods: Within the National Inpatient Sample (NIS) database we identified 11 258 RNU patients (2000-2015). We used the Johns Hopkins frailty-indicator to stratify patients according to frailty status. Time trends and multivariable logistic, Poisson and linear regression models were applied. Results: Overall, 1801 (16.0%) patients were frail, 4664 (41.4%) were older than 75 years and 1530 (13.6%) had Charlson comorbidity index ≥2. Rates of frail patients increased over time, from 7.3% to 24.9% (P <.001). Frail patients exhibited higher rates (all P <.05) of overall complications (62.6% vs 50.9%), in-hospital mortality (1.6% vs 1.0%), non-home–based discharge (22.7% vs 12.1%), longer length of stay (LOS) (6 vs 1 day) and higher THCs ($49 539 vs $39 644). Moreover, frailty independently predicted (all P <.05) overall complications (OR, 1.46), in-hospital mortality (OR, 1.52), non-home–based discharge (OR, 1.36), longer LOS (RR, 1.30) and higher THCs (RR, +$11 806). Conclusion: Preoperative frailty is important in RNU patients. One of four RNU patients is frail. Moreover, frailty predicts short-term postoperative complications, as well as longer LOS and higher THCs after RNU.