Abstract
RATIONALE & OBJECTIVE: Fluid management in hemodialysis aims to remove excess fluid while avoiding symptoms of fluid depletion. This study evaluated the impact of determining euvolemic body mass by bioimpedance spectroscopy, and absolute blood volume (ABV), on clinical practice and patient-reported outcome measures (PROMs) at a bioimpedance/ABV-naïve dialysis center. STUDY DESIGN: Fourteen week quality improvement project with 3 Assessment phases separated by 2 Adjustment phases. SETTING & PARTICIPANTS: Total of 127 patients at a single dialysis center. QUALITY IMPROVEMENT ACTIVITIES: Bioimpedance-spectroscopy-derived fluid overload (FO), ABV, and PROMs were longitudinally recorded. Physicians received data from each Assessment phase to inform treatment decisions. OUTCOMES: Fluid overload, ABV, PROMs, and agreement between perceived fluid status and FO. ANALYTIC APPROACH: Generalized linear mixed-effects models analyzed changes over time and associations between FO, ABV, and PROMs. Agreement between perceived fluid status and FO was evaluated with linearly weighted Cohen's κ. RESULTS: With each Assessment phase, pre-dialysis FO, systolic and diastolic blood pressure decreased overall (-0.12 L, P = 0.006; -1.43 mm Hg, P = 0.003; -0.83 mm Hg, P < 0.001), with a stronger reduction in baseline fluid overloaded (FO relative to extracellular fluid >15%) patients (-0.25 L, P < 0.001; -1.84 mm Hg, P = 0.011; -1.18 mm Hg, P = 0.01). The difference between post-dialysis and euvolemic body mass decreased in fluid overloaded patients (-0.19 kg, P = 0.006). Odds of longer recovery time increased (1.43, P = 0.017), but no significant changes in intradialytic complications or hypotension occurred, and FO and ABV were not associated with any PROM (including recovery time) overall and separately in fluid overloaded patients. Agreement between perceived and bioimpedance-spectroscopy-derived FO was poor (κ: 0.007-0.037 for patients, 0.022-0.018 for nurses). CONCLUSIONS: During the introduction of bioimpedance- and ABV-guided fluid management at a hemodialysis center, fluid status improved without significant changes in intradialytic morbid events. The lack of agreement between perceived fluid status and bioimpedance-spectroscopy-derived FO reflects the daily clinical struggle when negotiating fluid management based on objective measures.