Longitudinal bioimpedance vector plots add little value to fluid management of peritoneal dialysis patients

Abstract

Bioimpedance (BI) has the potential to enable better management of fluid balance, which can worsen over time on peritoneal dialysis (PD) due to loss of residual kidney function and progressive muscle wasting. We undertook a prospective, randomized, open-label, blinded end-point controlled trial to determine whether availability of longitudinal BI measures as vector plots helped clinicians maintain stable fluid status over 12 months in 308 peritoneal dialysis patients from the United Kingdom and Shanghai, China. Patients were recruited into 4 groups nested within a single trial design according to country and residual kidney function. Nonanuric subjects from both countries demonstrated stable fluid volumes irrespective of randomization. Hydration worsened in control anuric patients in Shanghai with increased extracellular/total body water (ECW/TBW) ratio (0.04; 95% CI: 0.01, 0.06) and reduced TBW (−1.76 L 95% CI: −2.70, −0.82), but was stable in the BI intervention group whose dialysate glucose prescription was increased. However, multilevel analysis incorporating data from both countries showed worsening ECW/TBW in active and control anuric patients. Clinicians in the United Kingdom reduced target weight in the nonanuric BI intervention group causing a reduction in TBW without beneficial effects on ECW or blood pressure. Thus, routine use of longitudinal BI vector plots to improve clinical management of fluid status is not supported. Bioimpedance (BI) has the potential to enable better management of fluid balance, which can worsen over time on peritoneal dialysis (PD) due to loss of residual kidney function and progressive muscle wasting. We undertook a prospective, randomized, open-label, blinded end-point controlled trial to determine whether availability of longitudinal BI measures as vector plots helped clinicians maintain stable fluid status over 12 months in 308 peritoneal dialysis patients from the United Kingdom and Shanghai, China. Patients were recruited into 4 groups nested within a single trial design according to country and residual kidney function. Nonanuric subjects from both countries demonstrated stable fluid volumes irrespective of randomization. Hydration worsened in control anuric patients in Shanghai with increased extracellular/total body water (ECW/TBW) ratio (0.04; 95% CI: 0.01, 0.06) and reduced TBW (−1.76 L 95% CI: −2.70, −0.82), but was stable in the BI intervention group whose dialysate glucose prescription was increased. However, multilevel analysis incorporating data from both countries showed worsening ECW/TBW in active and control anuric patients. Clinicians in the United Kingdom reduced target weight in the nonanuric BI intervention group causing a reduction in TBW without beneficial effects on ECW or blood pressure. Thus, routine use of longitudinal BI vector plots to improve clinical management of fluid status is not supported. Optimal fluid management is one of the primary objectives of dialysis treatment, and there is significant concern that peritoneal dialysis (PD) patients can become progressively fluid-loaded with time on treatment, especially as residual kidney function declines.1Van Biesen W. Williams J.D. Covic A.C. et al.Fluid status in peritoneal dialysis patients: the European Body Composition Monitoring (EuroBCM) study cohort.PLoS One. 2011; 6: e17148Crossref PubMed Scopus (195) Google Scholar, 2Tan B.K. Chan C. Davies S.J. et al.Achieving euvolemia in peritoneal dialysis patients: a surprisingly difficult proposition.Semin Dial. 2010; 23: 456-461Crossref PubMed Scopus (17) Google Scholar There is a growing body of evidence that bioimpedance (BI) analysis has a role to play in assisting the clinician in managing fluid status;3Davies S.J. Davenport A. The role of bioimpedance and biomarkers in helping to aid clinical decision-making of volume assessments in dialysis patients.Kidney Int. 2014; 86: 489-496Abstract Full Text Full Text PDF PubMed Scopus (210) Google Scholar this is primarily based on observational studies showing that overhydration, as determined from BI, predicts worse survival4Pillon L. Piccoli A. Lowrie E.G. et al.Vector length as a proxy for the adequacy of ultrafiltration in hemodialysis.Kidney Int. 2004; 66: 1266-1271Abstract Full Text Full Text PDF PubMed Scopus (99) Google Scholar, 5Paniagua R. Ventura M.-J. Avila-Díaz M. et al.NT-proBNP, fluid volume overload and dialysis modality are independent predictors of mortality in ESRD patients.Nephrol Dial Transplant. 2010; 25: 551-557Crossref PubMed Scopus (189) Google Scholar and the demonstration that BI can detect changes following interventions designed to improve fluid status.6Davies S.J. Woodrow G. Donovan K. et al.Icodextrin improves the fluid status of peritoneal dialysis patients: results of a double-blind randomized controlled trial.J Am Soc Nephrol. 2003; 14: 2338-2344Crossref PubMed Scopus (306) Google Scholar, 7Machek P. Jirka T. Moissl U. et al.Guided optimization of fluid status in haemodialysis patients.Nephrol Dial Transplant. 2010; 25: 538-544Crossref PubMed Scopus (175) Google Scholar BI data comprise the following two components: resistance to an electrical current, typically passed through the body from the wrist to the ankle, which is inversely proportional to total body water (TBW), and reactance, which is the impedance to this alternating current, also measured in Ohms, as it passes through tissues with cell membranes and thus it is proportional to cell mass. These two components can be plotted as a two-dimensional vector and used to track changes in fluid status that could support clinical decisions (Figure 1). However, few clinical trials have been conducted that clearly demonstrate a benefit of BI over and above good-quality standard clinical management. Longitudinal studies of body composition indicate that progressive overhydration is usually associated with a decline in muscle mass and a potential failure to adjust the dialysis prescription so as to reduce the extracellular water (ECW) volume down in parallel with this.8Chertow G.M. Johansen K.L. Lew N. et al.Vintage, nutritional status, and survival in hemodialysis patients.Kidney Int. 2000; 57: 1176-1181Abstract Full Text Full Text PDF PubMed Scopus (161) Google Scholar, 9Johansen K.L. Kaysen Ga Young B.S. et al.Longitudinal study of nutritional status, body composition, and physical function in hemodialysis patients.Am J Clin Nutr. 2003; 77: 842-846Crossref PubMed Scopus (145) Google Scholar, 10Chan C. Smith D. Spanel P. et al.A non-invasive, on-line deuterium dilution technique for the measurement of total body water in haemodialysis patients.Nephrol Dial Transplant. 2008; 23: 2064-2070Crossref PubMed Scopus (21) Google Scholar, 11Davies S. Engel B. Chan C. et al.Breath analysis and the measurement of total body water using isotope dilution—applications in the dialysis clinic.Curr Anal Chem. 2013; 9: 593-599Crossref Scopus (11) Google Scholar We hypothesized that the longitudinal application of BI alongside clinical evaluation would help the clinician identify this problem and thus make appropriate adjustments to the prescription. To test this hypothesis, we undertook a randomized controlled trial to determine whether the additional information available from longitudinal BI over 12 months could assist in maintaining stable or improved fluid status. Using the same basic design, we included four independent randomization groups comprising nonanuric and anuric patients from three UK dialysis centers (Stoke-on-Trent, Leeds, and Sheffield) and one Chinese center (Shanghai), respectively. Our aim was to determine whether routine clinical management supported by the longitudinal plot of the BI vector, which shows the direction in which fluid status is changing, resulted in more stable fluid status than control subjects. The outcome, to which the clinicians were blinded, was fluid volumes, ECW, TBW, and their ratio (ECW/TBW), as determined from the BI measurements after the trial was completed. Recruitment, randomization, and dropout to and from the four study groups are summarized in the consort diagram (Figure 2). With the exception of the UK anuric group, recruitment was sufficient to test our primary outcome with 80% power to detect a 1-kg change (in ECW) in the Shanghai (nonanuric and anuric) patients and a 0.8-kg change in the UK nonanuric (UK nonanuric) patients. Failure to achieve power in the UK anuric group was because of a combination of lack of recruitment indicative of the low proportion of anuric patients in the three UK centers and a high dropout (66%). There was a nonsignificant increase in deaths in this patient group randomized to the BI intervention; careful analysis of these 4 deaths (cancer or sepsis) and adverse outcomes did not indicate or to the but this group was from analysis from the in the groups was over the of the as by plots and There were significant patients randomized to the BI intervention or control in of the groups in of their dialysis residual kidney peritoneal blood or body composition 1). Shanghai patients to be have with one and than UK patients which was in a dialysis prescription The blood and peritoneal were in the Shanghai patients. the nonanuric residual were in UK as was the residual this was for body weight the was nonsignificant anuric patients been on for and UK anuric UK of patient groups according to weight blood ECW, extracellular peritoneal TBW, total body in a blood ECW, extracellular peritoneal TBW, total body both the UK and Shanghai nonanuric and the Shanghai BI intervention group, there were significant changes in body composition over the study and and 2). the BI intervention group, there was a reduction in TBW 95% to that was associated with a reduced target weight by clinicians 95% CI: to and weight 95% CI: to there was change in the ECW 95% CI: to or the ECW/TBW 95% CI: to the Shanghai anuric a significant in body composition in the control patients because of a in TBW (−1.76 95% CI: to −0.82), increase in ECW 95% CI: to and thus worsening of the ECW/TBW ratio 95% CI: body composition stable in the BI intervention group (Figure was associated with of the BI vector plot used to aid clinical decision-making in the intervention group, in the there was a reduction in in with reduced muscle mass and worsening in the 95% CI: a in (Figure multilevel analysis the data from both countries and study these significant and independent of of fluid volumes and of the ECW/TBW ratio worsened in both the anuric control and intervention changes on body weight weight ECW, extracellular TBW, total body is as the in in a plot showing changes in bioimpedance and their associated for the patients body and of the vector plot to The control group worsening overhydration a reduction in body the BI intervention group was BI, analysis of of fluid volumes and incorporating patients and centers intervention intervention intervention of of TBW 95% 0.06) with or more with or more in this from from 4 from from as 95% with or more with or more in this from from 4 from from ECW, extracellular TBW, total body are in the BI fluid volume is for and TBW is by in and over in because of target weight reduction in the ECW/TBW is by and and over in anuric patients. from are in with or more from in a ECW, extracellular TBW, total body is as the in BI, ECW, extracellular TBW, total body are in the BI fluid volume is for and TBW is by in and over in because of target weight reduction in the ECW/TBW is by and and over in anuric patients. from are in There were significant longitudinal changes in blood in of intervention or control groups the nonanuric patient groups over 12 there was a significant in the residual associated with in with the exception of the UK intervention group in which residual volume was of volume in the of loss in is in with increased it was not to demonstrate this The reduction in residual function in the nonanuric groups was associated with in the dialysis in patient There were changes in ultrafiltration or in of the groups with the exception of a significant increase in in the Shanghai anuric in both study in the intervention changes in outcomes and dialysis volume glucose volume blood not ultrafiltration from peritoneal in a of clinical assessments associated with interventions that were not by and on more than one intervention was or by as interventions the of the fluid fluid for the of interventions in nonanuric dialysate glucose dialysate glucose not as it is in was in the of interventions active and control and the proportion of assessments not the there were more of the glucose prescription in the but the increase in prescription in Shanghai was for the of interventions in nonanuric patients. in a BI, blood not ultrafiltration from peritoneal not as it is in China. There was in the of interventions active and control and the proportion of assessments not the there were more of the glucose prescription in the but the increase in prescription in Shanghai was of the study design was to decisions the of fluid management assessments so as to better clinicians were using the additional information from a potential and assessments assessments were in the active BI of fluid status, data were and for analysis for Shanghai with data and for the UK with the are not by randomization of intervention was by usually and but not according to randomization, with the exception of glucose prescription in the BI group a change in target weight of in direction as a for a significant it can be that the weight was more to be reduced in the intervention with the control groups in three with the primary as improve control of high blood status on clinical was to be a reduction of weight in the BI group was this on decision-making was in the BI thus the of weight reduction in this intervention group, as (Figure of interventions according to changes in target weight as in target for achieve blood for weight to fluid status based on clinical to change weight in of not target weight by change in target weight target weight increased by are for the in a not target weight by change in target weight target weight increased by are for the The of this trial was the than over 12 months of fluid volumes, especially in nonanuric patients. in which significant changes were with or without for associated with BI volumes these were in the TBW with or without parallel change in the anuric this was and to loss in as we to an increase in the ECW/TBW There was evidence that this could be in the Shanghai BI intervention group that of the BI vector associated with dialysate glucose but in blood which was controlled in both the BI intervention group, the in TBW was because of the of a target weight but change in the ECW/TBW ratio or in blood pressure. these that longitudinal BI vector additional to clinical fluid management. The in TBW with an increase in ECW/TBW in the anuric patients is in with our study of longitudinal body composition in hemodialysis patients in increased ECW/TBW associated with overhydration supported by independent measures of TBW from deuterium C. C. Smith D. et and measures of longitudinal in J Am Soc Nephrol. PubMed Scopus Google Scholar as as longitudinal studies using in K.L. Kaysen Ga Young B.S. et al.Longitudinal study of nutritional status, body composition, and physical function in hemodialysis patients.Am J Clin Nutr. 2003; 77: 842-846Crossref PubMed Scopus (145) Google Scholar, A.C. B. et composition in patients with peritoneal Dial Transplant. PubMed Scopus Google Scholar is to that our study design was to this problem by showing BI be used to detect longitudinal changes in body composition and to adjust for these than using it as a to patients to a target status. is in to et fluid study a randomized controlled study on the effects of bioimpedance analysis and on mass in peritoneal dialysis Dial Int. 2011; PubMed Scopus Google Scholar, L. G. et al.A randomized controlled trial of of fluid management in patients and study Nephrol. PubMed Scopus Google Scholar or study M. S. L. et fluid management in a randomized controlled J 2014; Full Text Full Text PDF PubMed Scopus Google Scholar, M. et of fluid management by bioimpedance on in hemodialysis patients: a randomized controlled J 2013; Full Text Full Text PDF PubMed Scopus Google Scholar and it was in to especially loss of residual kidney function. a study using the overhydration to reduce weight was associated with a significant loss of in nonanuric hemodialysis M. et of fluid management by bioimpedance on in hemodialysis patients: a randomized controlled J 2013; Full Text Full Text PDF PubMed Scopus Google Scholar problem was not in the trial in patients that did that BI improved fluid status and blood but was to months et control in peritoneal dialysis patients by bioimpedance 2011; PubMed Scopus Google Scholar Our to use vector plots of BI data than volumes is also of was by a of the lack of a for ECW volume in the in which overhydration is by as and we were to was for the and the vector plot clinicians from this it to clinicians from the BI which it can be and by the of volumes on which the study was after the study data we were to the of clinician there are for in the A. S. et of impedance vector by body mass and for J Clin Nutr. PubMed Scopus Google Scholar as as the dialysis in which both and vector length have been to be of L. Piccoli A. Lowrie E.G. et al.Vector length as a proxy for the adequacy of ultrafiltration in hemodialysis.Kidney Int. 2004; 66: 1266-1271Abstract Full Text Full Text PDF PubMed Scopus (99) Google Scholar the time of study design, these data were not available for in the overhydration the BI that we used have a in and were a has been in the study et fluid study a randomized controlled study on the effects of bioimpedance analysis and on mass in peritoneal dialysis Dial Int. 2011; PubMed Scopus Google Scholar of the of our study was to information on decision-making by clinicians in an to the use of BI clinical assessments did not to a change in target and it be that changes in weight without a change in have in for changes in body the control the to increase or target weight was the weight in the intervention groups was is that for the UK clinicians the to make weight in the intervention group so as to achieve better blood control was because of study The of intervention difficult to as these were usually of and within patients and not control and BI The of a reduction in target could be by as in a trial use with of in the of residual function in patients on peritoneal Int. Full Text Full Text PDF PubMed Scopus Google Scholar but this could not be clearly because of the that patients were on and that interventions were this did not into in status, as by ECW or blood management. However, this support observational that of residual volume not on maintaining K. S. Davenport A. volume measured by not help residual function in peritoneal dialysis patients.Kidney Int. 2014; Full Text Full Text PDF PubMed Scopus Google Scholar There is a concern that the loss of weight a reduction in muscle but the change in TBW was not associated with a significant change in the electrical and is by the weight change into the used to TBW, and thus it be because of loss in body evaluation and a better of patients to to reduce their target weight that trial The to reduce especially for the of blood was also in the Shanghai but the changes were not so the better blood measurements in the Chinese patients. The of BI in the anuric intervention group to have been a better of decision-making with a significant increase in glucose thus the worsening increase in glucose use also have increased thus loss of which was in this study has a of we to to the and to clinicians to the control BI data and it was by to the of randomization group and thus by for that the more weight reduction in the intervention group this than information from the BI measurements not to be included in the study by modality especially in the UK of to achieve in the in Shanghai, which that the Chinese anuric patients could be more However, it can be that this to the UK more so the in anuric patients. The in the UK randomization, which in have been by did not the of measured patient and the failure to sufficient anuric patients from the UK has been by a analysis incorporating both data from both analysis showed that and are the of body composition and that centers were a function of and interventions in the interventions would have been data on as this to be more than was we but these have been to be an of change in T. in total not to changes in total in Dial Int. Google Scholar The of the study were design, which routine use of longitudinal BI to fluid management in patients over 12 this was because body composition is stable in nonanuric it could be because we to maintain a stable BI vector than more to fluid status. studies are to an but careful of outcomes as residual kidney function and blood this study difficult it is to and the interventions and associated with fluid management and thus and increase in glucose prescription was the intervention associated with The study design was a controlled trial based on the L. T. B. randomized blinded end-point design for intervention PubMed Scopus Google prospective, randomized, open-label, blinded end-point in which patients were to an active in which BI measurements were available to clinicians in their of fluid status or to a control in which measurements were but of control BI data was by using vector which two-dimensional (Figure their clinicians were blinded to the primary body composition outcome for as fluid volumes were after of the study and data this design, four groups of patients were recruited using an and The four groups were nonanuric and as a volume with patients recruited from three UK and one Chinese The the for four independent patient groups nested within a trial design is that clinical decisions are to be by the of residual kidney that the for intervention are in the and are with are more a proportion of anuric and that there were in the two body weight and and were used to routine Clinicians were to patients were to be patients be on for more than months for subjects and the trial was with clinicians were to months to fluid status using standard clinical not and to that the patients from after randomization, fluid assessments BI were a of months over the following the the assessments were the use of BI in the active the The primary outcome was body composition, as from outcomes included blood residual kidney peritoneal function and ultrafiltration and dialysis of fluid status was using an (Figure that included physical and blood pressure. the BI a vector plot was to and clinicians were the study in this could be used to maintain stable fluid status Using this an increase in total fluid a change in a change in the vector that a increased and this changes in and clinical can be alongside but not measures of body The interventions used by clinicians were their and but included of weight by the patient to reduce and fluid the dialysis prescription of glucose or modality or peritoneal or prescription for of These were with the the for used to fluid status by the support and of of of available in on fluid to the for of on the nutritional on using dialysate in patients of in patients a by glucose volume available with of patients. to and fluid to on and used more there is of one is in a of patients a available in with of patients. for and on to using dialysate in patients of in of patients a not available in and on fluid and than target of and fluid the is to to be more or than the used volume that and to achieve more fluid is of for as and are peritoneal in a peritoneal were with BI body composition These resistance and a single using the in dialysis R. N. et of impedance to and survival in peritoneal dialysis patients.Kidney 2003; Full Text Full Text PDF Scopus Google Scholar, M. A. B. impedance analysis after a hemodialysis in evaluation of nutritional Nutr. Full Text Full Text PDF PubMed Scopus Google Scholar, G.M. Lowrie E.G. et with impedance analysis in hemodialysis Am Soc Nephrol. 6: PubMed Google Scholar BI measurements were using a with of the of without the of dialysate and after the patient been for Using this we have within patient on the same of of and to determine of was using the S.J. L. et in peritoneal dialysis patients and to predictors of Dial Transplant. PubMed Scopus Google Scholar was measured in using standard function was determined from and as the and to body of was determined from and as function was determined using the peritoneal test and as the ratio and ultrafiltration for was from the of the volume and the and as used was determined using the in the UK and in The these is and in Dial Transplant. PubMed Scopus Google Scholar, but data are after to the to enable The trial was so as to be to detect a longitudinal change in body composition from ECW with an a that changes would in the control group that would be in the active the of data from longitudinal studies of body composition in patients over 12 months in which the of the in ECW was a in ECW, a with 80% would patients group with randomization, and a change would patients in The was to sufficient to detect a 1-kg but to within the nonanuric and anuric the dropout was and patients The primary analysis was as after study and data the and data were to of ECW, TBW using an based on the with the BI and the change 12 months from determined by to changes in target and weight and To dropout in patient and study survival were determined from survival and the and of fluid assessments and interventions were using or as multilevel analysis with fluid volumes as the was to and centers and for the associated with BI volumes on of the data and were for active and control in anuric and nonanuric with and were using for the multilevel analysis for which we used for of the was supported by a and by a UK in with from the which also of the BI was by a The study was the UK clinical We are to for help with power and from and to fluid but for this study was from the of bioimpedance and of the or have in or The was the in in with survival plots for patient dropout by