Effect of a standardized fluid management algorithm on acute kidney injury and mortality in pediatric patients on extracorporeal support

In vivo measurement of pediatric extracorporeal oxygenator insensible losses; a single center pilot study

Introduction

Fluid overload on Extracorporeal Membrane Oxygenation (ECMO) is associated with worse outcomes. Previous in vitro studies have attempted to quantify oxygenator-related insensible losses, as failure to account for this fluid loss may lead to inaccurate fluid balance assessment and potentially harmful clinical management, such as unnecessary exposure to diuretics, slow continuous ultrafiltration (SCUF), or continuous kidney replacement therapy (CKRT). We performed a novel in vivo study to measure insensible fluid losses in pediatric ECMO patients.

Methods

Pediatric ECMO patients were approached over eleven months in the pediatric and cardiac intensive care units. The water content of the oxygenator inflow sweep gas and exhaust gas were calculated by measuring the ambient temperature and relative humidity at frequent intervals and various sweep flow.

Results and discussion

Nine subjects were enrolled, generating 431 data points. The cohort had a median age of 11 years IQR [0.83, 13], weight of 23.2 kg IQR [6.48, 44.28], and body surface area of 0.815 m2 IQR [0.315, 1.3725]. Overall, the cohort had a median sweep of 2.5 L/min [0.9, 4], ECMO flow of 3.975 L/m2/min [0.75, 4.51], and a set ECMO temperature of 37 degrees Celsius [36.6, 37.2]. The calculated net water loss per L/min of sweep was 75.93 ml/day, regardless of oxygenator size or patient weight. There was a significant difference in median documented vs. calculated fluid balance incorporating the insensible fluid loss, irrespective of oxygenator size (pediatric oxygenator: 7.001 ml/kg/day [-12.37, 28.59] vs. -6.11 ml/kg/day [-17.44, 13.01], respectively, p = 0.005 and adult oxygenator: 14.36 ml/kg/day [1.54, 25.77] and 9.204 ml/kg/day [-1.28, 22.05], respectively, p = <0.001). We present this pilot study of measured oxygenator-associated insensible fluid losses on ECMO. Our results are consistent with prior in vitro methods and provide the basis for future studies evaluating the impact of incorporating these fluid losses into patients' daily fluid balance on patient management and outcomes.