| Summary: | 博士 === 靜宜大學 === 食品營養學系 === 101 === Background: The clinical parameters of patients undergoing clinical peritoneal dialysis (CPD) is associated with long-term clinical outcome. Previous studies had pointed out that some markers are independent risk factors of mortality in PD patients；however, which parameter would powerfully predict the dropout rate of PD is currently less clear, especially in longitudinal cohort. Objective: We compared the effects of different indicators on dropout rate (technique failure and death) in PD patients in a longitudinal cohort. Materials and Methods: The study enrolled 263 chronic PD patients; data were collected in PD outpatient clinics at the National Cheng Kung University Hospital from June 2003 to September 2011 and fellow up to September 2012, total fellow up 9 years. The data analysis included SGA 7-point score, anthropometry, hand grip strength (HGS), biochemical parameters, residual renal function, dialysis adequacy. Results:103 patients dropped out from PD during 9 years of follow-up. Univariate Cox regression analyses showed that serum albumin (HR = 0.40, P< 0.0001), total SGA score (HR = 0.74, P < 0.0001), and activity score in SGA (HR=0.73, P < 0.0001) can reduce the dropout rate. Kaplan-Meier analysis, patients with SGA score > 4 or blood albumin > 3.5 g / dl can significantly reduce the dropout rate in PD patients (P < 0.0001); furthermore, male with HGS > 28.5 kg or female with HGS > 20.0 kg had significantly lower dropout rates. Multivariate Cox's regression analyses with no cutoffs revealed that serum albumin level (HR=0.49, P = 0.015) and SGA score (HR= 0.78, P = 0.010) significant predicted the high risk of dropout. By multivariate Cox’s regression analyses with cutoffs in the three parameters revealed that the albumin 3.5 g/dl increased 2.14-fold (P = 0.005), SGA 4 score increased 1.86 times (P = 0.009), HGS 28.5 kg in male increase 2.61 times (P = 0.016), female HGS 20.0 kg increased 1.51 times of risk. By time-independent multivariate Cox’s regression analyses, patients maintained a higher BMI after dialysis can reduce the risk of dropout (HR= 0.83) and serum albumin significantly reduced the risk (HR = 0.399). Conclusion: The indicators of dropout rate including total SGA score and serum albumin can better predict the risk of dropout peritoneal dialysis than other markers. HGS only predicts the dropout rate in men, but not women. In the long-term analysis, only serum albumin can dependently predict the dropout of PD.
|
|---|
