Fluid replacement therapy is often used as a primary treatment modality in vaso-occlusive crises for sickle cell disease. However, aggressive intravenous fluid administration can lead to complications, including pulmonary vascular congestion and acute chest syndrome. Data is limited on the safety, efficacy, and variability of fluid replacement therapy in this context. This study describes the extent of and outcomes associated with receiving fluid replacement therapy for vaso-occlusive episodes at a single institution.
We performed a retrospective analysis on 157 unique patient encounters from 49 sickle cell patients hospitalized with a vaso-occlusive episode at the University of California Los Angeles between 2013 and 2017. Fluid administration measurements were derived from documented intakes (both oral and intravenous fluids) in the medical record. The association between fluid administration and outcomes were analyzed using chi-squared and Fisher's exact tests followed by univariate and multivariate logistic regression.
The mean age of the patient cohort at the time of hospital admission was 36.0 years (Std 7.9). Most encounters were from patients with hemoglobin SS disease (147 encounters, 93.6%). Twenty-two of the encounters (14.0%) were from patients on hemodialysis prior to admission. The majority of admissions were to an observation unit (99 encounters, 63.1%), whereas 53 admissions (33.8%) were to an inpatient service and 5 encounters (3.2%) were solely in the emergency room. The median length of hospital stay was 4 days (IQR 2-7). The mean total amount of intravenous fluid administered during the hospitalization was 7.4 L (Std 9.6). The mean total amount of fluid intake including intravenous fluids, blood transfusions, and oral fluids was 6.5 L (Std 10.0) (Table 1).
The most common clinical outcome that occurred during the hospitalizations was a new oxygen requirement in 28 encounters (17.8%). Other clinical outcomes noted were acute chest syndrome (12 encounters, 7.6%), an aspiration event (2 encounters, 1.3%), other type of hospital-acquired infection (2 encounters, 1.3%, which included Clostridium difficile colitis and Staphylococcus epidermidis bacteremia), acute kidney injury (3 encounters, 1.9%), and intensive care unit transfer (3 encounters, 1.9%). There was a significant association between receiving more than 3 L of intravenous fluid and the development of any of the adverse events listed (p = 0.029) but no association between fluid administration and each individual adverse event (Table 2).
Logistic regression analysis confirmed that patients with higher fluid intake were more likely to develop any adverse event (Table 3, Table 4). In multivariable analysis, each component including oral intake during hospitalization (p = 0.041, OR 1.065, 95% CI 1.003-1.132), intravenous fluid administered in the first 24 hours (p = 0.001, OR 1.899, 95% CI 1.319-2.733), total amount of intravenous fluid administered (p = 0.005, OR 1.081, 95% CI 1.023-1.141), and total amount of fluid intake (p = 0.040, OR 1.065, 95% CI 1.003-1.132) all revealed a statistically significant association between higher fluid administration and the development of any adverse event. Other factors found to be significantly associated with any adverse event were dialysis dependence prior to admission (p = 0.000, OR 8.686, 95% CI 2.881-26.190) and admission to inpatient service versus emergency room or observation unit (p = 0.018, OR 2.758, 95% CI 1.186-6.416).
There was a statistically significant association between higher fluid intake (both oral and intravenous) and the development of any adverse event during hospitalization for sickle cell vaso-occlusive crisis including a new oxygen requirement, acute chest syndrome, aspiration event, other hospital-acquired infection, acute kidney injury, and intensive care unit transfer. While fluid administration may theoretically slow the sickling process, our data suggests that excessive fluid administration during a vaso-occlusive episode may be harmful. Further study is necessary to further elucidate the relationships between exogenous fluids, vaso-occlusion, and adverse events in sickle cell patients.
*Equal contribution from Daria Gaut and Jennifer Jones for this work.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.