P3447New frailty assessment based on routine nurse anamnesis before discharge is a strong predictor of all-cause mortality in patients with myocardial infarction

Abstract

Background: Age and frailty are well-established risk factors in patients with acute myocardial infarction (AMI). Many frailty assessments are based on questionnaires and physical tests, which are not practical in daily clinical routine. At our department all patients are assessed daily by the nursing staf for their nursing demands, including activities of daily living (ADL; eat/drink, stool, mobility, hygiene) in 3 diferent grading (not, partly, fully self-sufcient). Purpose: We hypothesised that theses data is enough to calculate a predictive frailty score. Methods: We performed a retrospective analysis (inclusion of: Type-1 AMI, ≥65 years, years 2012-2014; exclusion: cardiogenic shock) and collected the patient's anamnesis and ADL assessments before discharge. Depending on the grading within each ADL category (0 points = not, 1 point = partly, 2 points = fully self-sufcient, respectively), patients were strati-fed into 3 diferent groups: severe (0-3 points), moderate (4-6 points) and not frail (7-8 points), respectively. Primary end-point was all-cause mortality on 31st December 2015. We performed a descriptive analysis of data. Diferences in mortality were assessed by log-rank test, independent predictors were investigated by linear and cox-regression models. Results: We identifed 396 patients (35.6% STEMI, 44.4% females, mean age 76.7 years). Of those, 5.6% were severe, 22.5% were moderate, and 72.0% were non-frail according to our score. Overall all-cause mortality was 25.3% (median follow-up 2.5 years). Frail patients were older (p = 0.001), had signif-cantly more often STEMI at presentation (p = 0.022), increased conservative therapy (p < 0.001), known heart failure (p = 0.025), atrial fbrillation (p < 0.001), a history of stroke (p = 0.043) and were more likely to be female (p = 0.012). Independent predictors for frailty score at discharge (adjusted for baseline characteristics) were age (OR-0.38;-0.064-0.012; p = 0.004), rep-erfusion therapy (OR 0.989; 0.548-1.43; p < 0.001) and atrial fbrillation (OR-1.242;-1.769 to-0.716; p < 0.001). Mortality rates for frailty cohorts are depicted in Fig. 1 | 15-3 (log-rank <0.0001 between groups). Independent predictors for long-term all-cause mortality were frailty score (p < 0.0001; HR 0.755; 0.687-0.830), age (p = 0,001; HR 1.05; 1.021-1.080), kidney function (MDRD) (p = 0.002; HR 0.986; 0.978-0.995), known heart failure (p = 0.017; HR1.911; 1.121-3.257) and reperfusion therapy (p = 0.032; HR 0.586; 0.36-0.955). Tere was no interaction between age and the frailty score (p = 0.636). Conclusions: Our frailty score is very easy to assess and to calculate. After adjustment for prognostic baseline characteristics, our score serves as an independent predictor for long-term all-cause mortality in patients after AMI ≥65 years.