Evaluating the impact of frailty on periprocedural adverse events and mortality among patients with GI bleeding

Abstract

Frailty is a known predictor of mortality and adverse events in the inpatient setting; however, it has not been studied as a modality to assess risk among patients undergoing endoscopy for GI bleeding (GIB). We aimed to determine the association between frailty status and risk of adverse events in hospitalized patients with GIB who underwent endoscopy. We performed a cohort study using the 2016 and 2017 National Inpatient Sample database, using International Classification of Diseases diagnostic codes to identify adult patients with GIB who underwent endoscopic procedures within 2 days of admission and the Hospital Frailty Risk Score to classify patients as frail or nonfrail. Univariable and multivariable logistic regression models were constructed to assess the predictors of periprocedural adverse events, and marginal standardization analysis was performed to assess for possible interaction between age and frailty. A total of 757,920 patients met inclusion criteria, of which 44.4% (336,895) were identified as frail and 55.6% (421,025) as nonfrail; 49.2% of frail patients had composite periprocedural adverse events compared with 25.5% of nonfrail patients (P< .001). Frail patients notably had more cardiovascular (32.1% vs 17.1%, P< .001), pulmonary (18.5% vs 4.3%, P< .001), GI (10.1% vs 6.1%, P< .001), and infectious (9.9% vs .7%, P< .001) adverse events compared with nonfrail patients. Frail patients also had higher all-cause inpatient mortality rates (4.8% vs .5%, P< .001). On multivariable analysis, positive frailty status was associated with a 2.13 times increased likelihood of having composite periprocedural adverse events. In hospitalized patients undergoing endoscopy for GIB, frailty status is associated with increased periprocedural adverse events including all-cause mortality. The use of frailty assessments can thus further guide clinical decision-making when considering endoscopy and risk of adverse events in adult patients with GI hemorrhage.