Predictive Model for the Assessment of Preoperative Frailty Risk in the Elderly

Abstract

Adequate preoperative evaluation of frailty can greatly assist in the efficient allocation of hospital resources and planning treatments. However, most of the previous frailty evaluation methods, which are complicated, time-consuming, and can have inter-evaluator error, are difficult to apply in urgent situations. Thus, the authors aimed to develop and validate a predictive model for pre-operative frailty risk of elderly patients by using diagnostic and operation codes, which can be obtained easily and quickly from electronic records. We extracted the development cohort of 1762 people who were hospitalized for emergency operations at a single institution between 1 January 2012 and 31 December 2016. The temporal validation cohort from 1 January 2017 to 31 December 2018 in the same center was set. External validation was conducted on 6432 patients aged 75 years or older from 2012 to 2015 who had emergency surgery in the Korean national health insurance database. We developed the Operation Frailty Risk Score (OFRS) by assessing the association of Operation Group and Hospital Frailty Risk Score with the 90-day mortality through logistic regression analysis. We validated the OFRS in both the temporal validation cohort and two external validation cohorts. In the temporal validation cohort and the external validation cohort I and II, the c-statistics for OFRS to predict 90-day mortality were 0.728, 0.626, and 0.619, respectively. OFRS from these diagnostic codes and operation codes may help evaluate the peri-operative frailty risk before emergency surgery for elderly patients where history-taking and pre-operative testing cannot be performed.