BIS-guided Fluid Management in HD Patients

The study was designed as a randomized controlled open label and parallel trial, including adult chronic kidney disease (CKD) patients (older than 18 years) on in-center HD for longer than three months in the Dialysis and Transplantation unit of a southern Brazilian University Hospital. The exclusion criteria were less than thrice a week HD session, lower limbs amputation, pregnancy and pacemaker, implantable cardioverter-defibrillator or orthopedic protheses use. All eligible patients or their relatives were asked to sign an informed consent form before inclusion in the trial. The enrollment was performed from September 2019 to September 2021.

The research protocol was approved by the Catholic University of Pelotas Ethics Committee.

The randomization was performed by blocs according to dialysis shifts, using random number table created in Stata 15.1 statistical software (StataCorp, College Station, TX) Participants and researchers were not blinded due to the nature of the intervention. The statistical analysis was performed by a blinded evaluator.

The main outcome was all-cause two-year survival. Secondary outcomes include overall hospitalization rate and according causes, systolic and diastolic BP change (mean of the three last post-dialysis BP measurement - mean of the three first post-dialysis BP measurement), number of anti-hypertensive drug classes used after intervention time.

At inception, eligible patients had their sex, age and dry weigh recorded from electronic clinical records before the randomization to one of two groups: 1) BIS, whose dry weight is evaluated monthly by clinical examination and biannually BIS-guided; 2) CE, whose dry weight is evaluated monthly by clinical examination alone. The follow-up time for primary and secondary outcomes was two years. Data for hospitalizations, deaths, cause of death, and anti-hypertensives in use were collected from the electronical medical registry.

Procedures and Equipment Bioimpedance spectroscopy Patients randomized to BIS group were submitted to an initial evaluation using Body Composition Monitor (BCM) (Fresenius Medical Care, Bad Homburg, Germany). The BCM provides information on extracellular (ECW), intracellular (ICW) and total body water through a range of electric frequencies from 5 to 1,000 kilohertz (kHz). Based on these measurements, the equipment calculates parameters such as overhydration (OH), derived from the difference between the extracellular water measured and that expected for an individual of the same weight and stature in euvolemia. BCM also calculates extra- to intracellular water ratio (E/I). The measurement was performed by previously trained personal before a mid-week HD session, after five minutes of rest in the supine position, using electrodes placed on ipsilateral high and lower limbs. The patient was advised to avoid coffee or any meal 30 minutes before the evaluation. The data obtained was stored in a memory card to be discharged and analyzed with use of an accompanying software (Fluid Management Tool, Fresenius Medical Care). BIS has been validated as an accurate estimator of extracellular volume against dilution techniques in HD patients.

1. Clinical Examination The clinical evaluation was performed by a nephrologist with hemodialysis expertise and includes blood pressure measurements before and after a dialysis session, lung examination searching for crackles and lower limbs (or sacral region for bedridden patients) for edema. The aim of the examination is the estimation and/or revaluation of estimated dry weight. The blood pressure was measured using a sphygmomanometer (Tycos@, Welch Allyn, EUA) by the auscultatory technique, with the patient seated in the dialysis chair and upper limb extended. The cuff was placed two to three centimeters from cubital fossa, and a stethoscope (Litmann@, 3M, EUA) placed over the brachial artery to detect Korotkoff sounds during inflation and disinflation of the sphygmomanometer cuff. Lung examination was performed using the same stethoscope placed on posterior and anterior chest during an entire inspiration and expiration periods. Lower limbs were examined by pressing the indicator toe over the ankle and pretibial area searching for pitting edema.

Data Analysis The minimal required sample size was calculated based on the expected difference in the main outcome (survival in two years) between intervention (BIS) and control (CE) groups. According to previous publications on chronic HD mortality, we intended to find a difference in survival of 20% between groups, resulting in a minimal sample of 134 patients (67 in each group), to obtain an alpha error below 5% and an 80% power.

The distribution of the variables was tested by Shapiro-Wilk. Parametric variables were described as mean and standard deviation and non-parametric by median and interquartile range. The survival analysis was performed using Cox Proportional Hazard Regression. The hospitalization density-rate, point estimates and confidence interval for the incidence-rate ratio was calculated. The number of anti-hypertensive classes (0, 1, ≥2) were compared between groups by Person's chi-square. The difference between baseline and final systolic and diastolic blood pressure was tested by student's t test. The statistical package Stata 15.1 (StataCorp, College Station, TX) was used in analysis.