Characteristics Associated With Fluid Responsiveness and Fluid Toxicity Demonstrated by Passive Leg Raise Maneuver in Septic Patients in a Mozambican Emergency Department

Abstract

Background. Early initiation of fluid resuscitation and appropriate antibiotics improve survival. Studies conducted in resource-limited settings suggest that over-hydration is perilous. Passive leg raise (PLR) is one tool available in resource-limited areas to guide fluid responsiveness. We piloted the use of PLR as a tool to gauge adequacy of resuscitation (fluid responsiveness) and risk of respiratory difficulties with further fluid administration (fluid toxicity) in the Emergency Department of a public hospital a large proportion of HIV/AIDS patients in Maputo, Mozambique. Methods. Convenience sample of adults May 2014-July 2015 with standard definition of sepsis. PLR was performed in non-intubated patients without ascites to assess for possible fluid responsiveness defined as increase in mean arterial pressure (MAP) >10 mmHg or decrease in heart rate (HR) of >20. Fluid toxicity from PLR was defined as oxygen saturation (O2sat) drop from >92% to <92% or a decrease of >2% in those with sitting O2sat >92%. Results. Median age of the 93 participants was 35 (range 20-80 years). At the time of study entry HIV serostatus was positive in 55 (59.1%) negative in 6 (6.5%) and was unknown in 29 (31%). Prior to PLR evaluation 48 (51.6%) had already received at least one liter of fluid. Of the 80 patients completed the entire PLR procedure 26 (32.5%) were fluid responsive, and 9 (11%) demonstrated fluid toxicity. In the bivariate analysis, a prior diagnosis of HIV was associated with lower odds of fluid responsiveness (OR 0.337; 95% CI 0.128-0.886). There was a trend toward increased odds of fluid toxicity in those with older age (OR 1.231; 95% CI 0.974- 1.557). Conclusion. With potential for fluid overload in areas without ready access to positive-pressure ventilation and hemodialysis, PLR has the potential to predict respiratory compromise with additional fluid boluses. HIV status was associated with lower odds of fluid responsiveness, but there were no baseline characteristics associated with fluid responsiveness or toxicity highlighting the utility of the PLR to help guide therapy. Further study is needed to evaluate the impact of PLR-guided fluid resuscitation on clinical outcomes in limited resource settings.