Predictors of Length of Stay after Transcatheter Aortic Valve Implantation: Impact of Prognostic Nutritional Index.

Abstract

Pre-transcatheter aortic valve replacement (TAVR) nutritional status can potentially affect the length of hospital stay (LoS) after TAVR. The Prognostic Nutritional Index (PNI) is a widely recognised nutritional index. We aimed to determine the effect of PNI on LoS in patients undergoing TAVR. The study population (158 patients) was divided into two groups: early discharge (LoS ≤ 3 days) and late discharge (LoS > 3 days). PNI was calculated before TAVR. In the LoS > 3 days group, the median age, creatinine level, rate of surgical access site closure and rate of major complications were higher, whereas estimated glomerular filtration rate, albumin, haemoglobin, lymphocyte count and PNI were significantly lower. Receiver operating characteristic curve analysis revealed a PNI cutoff of 39 (area under the curve = 0.778, p < 0.001) with 86.8% sensitivity and 55.2% specificity for predicting extended LoS. The 30-day endpoint analysis revealed significantly higher rates of death and hospitalisation with LoS > 3 days and PNI ≤ 39. Multivariate binary logistic regression analysis identified several independent predictors of extended LoS: severe renal insufficiency [odds ratio: 3.951 (95% confidence interval: 1.281-12.191); p = 0.017], surgical access site closure [4.353 (1.701-11.141); p = 0.002), complications [7.448 (1.305-42.518); p = 0.024] and PNI < 39 [5.906 (2.375-14.684); p < 0.005]. Decreased PNI may be associated with LoS > 3 days after TAVR. Nutritional status assessed using PNI may be a useful independent predictor of LoS after TAVR.