POS-654 ASSESSMENT OF OVERHYDRATION IN CHILDREN ON CONTINUOUS AMBULATORY PERITONEAL DIALYSIS BY BIOIMPEDANCE VECTOR ANALYSIS – A LONGITUDINAL OBSERVATIONAL STUDY

Abstract

Detecting overhydration in children on chronic ambulatory peritoneal dialysis (CAPD) is critical as it contributes to cardiovascular morbidity and mortality. However, precise clinical assessment of overhydration is challenging. Bio-impedance vector analysis (BIVA) is a bedside body composition tool that can estimate overhydration with a potential role of guiding the clinician on fluid management. This study aimed to longitudinally assess and compare the overhydration status detected using clinical assessment and BIVA in children on CAPD. A longitudinal observational study was undertaken in clinically stable children (age 4-18years) on CAPD (minimum duration of 6 months) from May 2017 to January 2020 at the pediatric chronic kidney disease clinic of a tertiary hospital. Presence of overhydration was assessed by clinical signs, bioimpedance measures and BIVA at enrollment, 3 and 6 months of follow up. Modifications to CAPD prescriptions and antihypertensive medications were based only on clinical assessment. Children were categorized as having clinical overhydration if they fulfilled 2 of 3 criteria of weight gain (>7% of body weight), uncontrolled hypertension (office BP ≥95thcentiles despite being on anti-hypertensive medications) and presence of signs of fluid overload. Based on a normative reference R-Xc graph, derived from z-scores of 20 healthy controls, bioimpedance vectors were plotted. The position of the vector in the left lower quadrant beyond the 95% tolerance ellipse was regarded as overhydration. Differences in the longitudinal vector displacement were analyzed using Hotelling’s T2 test. Twenty six children [18 boys, median age 9.5(4,16)years] on CAPD for 21(12,72) months were enrolled. Non-glomerular etiology was noted in 69% and 58% had residual kidney function. At enrollment, the proportion of children with overhydration assessed clinically and by BIVA was 34.6% and 30.7% respectively. There was poor agreement between measures of overhydration by clinical assessment and BIVA at all three time points. Among those detected to be overhydrated at enrollment, 9/11 (82%) and 4/9 (44%) remained overhydrated at 6 months by BIVA and clinical assessment respectively. The BIVA plots on the reference RXc graph (Figure) revealed a significant difference in the vector displacement towards the quadrant of overhydration at all time points, with the maximum difference seen between enrollment and 6 months (T2 = 32.58, p <0.001). In children on CAPD, there is poor agreement between measures of overhydration detected clinically and by BIVA. Longitudinal vector plots over 6 months reflect persistence of overhydration in a majority, as compared to clinical assessment. Further studies to explore the utility of BIVA in guiding the clinician with fluid management in these children are warranted.