Frailty, Length of Stay, and Mortality in Kidney Transplant Recipients: A National Registry and Prospective Cohort Study

Abstract

Objective: To test whether frailty, a novel measure of physiologic reserve, is associated with longer kidney transplant (KT) length of stay (LOS), and modifies the association between LOS and mortality.

Background: Better understanding of LOS is necessary for informed consent and discharge planning. Mortality resulting from longer LOS has important regulatory implications for hospital and transplant programs. Which recipients are at risk of prolonged LOS and its effect on mortality are unclear. Frailty is a novel preoperative predictor of poor KT outcomes including delayed graft function, early hospital readmission, immunosuppression intolerance, and mortality.

Methods: We used registry-augmented hybrid methods, a novel approach to risk adjustment, to adjust for LOS risk factors from the Scientific Registry of Transplant Recipients (n = 74,859) and tested whether (1) frailty, measured immediately before KT in a novel cohort (n = 589), was associated with LOS (LOS: negative binomial regression; LOS ≥2 weeks: logistic regression) and (2) whether frailty modified the association between LOS and mortality (interaction term analysis).

Results: Frailty was independently associated with longer LOS [relative risk = 1.15, 95% confidence interval (CI): 1.03-1.29; P = 0.01] and LOS ≥2 weeks (odds ratio = 1.57, 95% CI: 1.06-2.33; P = 0.03) after accounting for registry-based risk factors, including delayed graft function. Frailty also attenuated the association between LOS and mortality (nonfrail hazard rate = 1.55 95% CI: 1.30-1.86; P < 0.001; frail hazard rate = 0.97, 95% CI: 0.79-1.19, P = 0.80; P for interaction = 0.001).

Conclusions: Frail KT recipients are more likely to experience a longer LOS. Longer LOS among nonfrail recipients may be a marker of increased mortality risk. Frailty is a measure of physiologic reserve that may be an important clinical marker of longer surgical LOS.

Figures

Figure 1

Distribution of Kidney Transplantation Length of Stay Using SRTR Data between 1995-2014 (n=173,868). The figure shows the distribution of KT length of stay within 30 days. 1% had LOS between 30-365 days. Those with a LOS >365 days (n=59) were excluded from all analyses.

Figure 2A

Distribution of Kidney Transplant (KT) Length of Stay By Year of KT Using SRTR Data between 1995-2014 (n=173,868). Note: Outliers (data points that fall outside the lower quartile-1.5*IQR, upper quartile+1.5*IQR) were excluded from the figure.

Figure 2B

The Prevalence of Length of Stay Exceeding 2 Weeks, by Year of Kidney Transplant Using SRTR Data between 1995-2014 (n=173,868).

Figure 3

Mortality by Kidney Transplant (KT) Length of Stay Using SRTR Data between 2008-2014 (n=74,859). A KT LOS exceeding 2 weeks is associated with an increased risk of mortality (Log-rank P-value