Retrospective Study of Characteristics and in-Hospital Mortality in Adults with Sickle Cell Disease Admitted to the Intensive Care Units in the US: Findings from the Nationwide Inpatient Sample 2016-2020

Introduction

Patients with sickle cell disease (SCD) are prone to complications and frequent hospitalizations, sometimes requiring care in the intensive care unit (ICU) for those who are critically ill. There is a lack of data on the epidemiology of adult patients with SCD admitted to ICUs in the United States. This study aims to describe these patients' characteristics and in-hospital outcomes and identify predictors of in-hospital mortality.

Methods

The National Inpatient Sample (NIS) database was used to identify ICU hospitalizations of adults aged 18 years and above between 2016 and 2020. This was further categorized into SCD and non-SCD groups. The ICD-10 codes used as surrogates for ICU admissions were 5A1935Z (respiratory ventilation, Less than 24 consecutive hours), 5A1945Z (respiratory ventilation, 24-96 consecutive hours), 5A1955Z (respiratory ventilation, Greater than consecutive hours), 5A12012 (Performance of Cardiac Output, Single, Manual), and 5A1221J (Performance of Cardiac Output, Continuous, Automated). Patient (age, sex, race, comorbidity index, income status, year of hospitalization, and insurance) and hospital (ownership, region, and location) characteristics were compared between the SCD and non-SCD groups. Also, the frequency of complications, including heart failure, stroke, sepsis, pulmonary embolism, transfusion reactions, acute kidney injury, and need for extracorporeal membrane oxygenation (ECMO), vasopressor use, and continuous renal replacement therapy (CRRT) were compared between the two groups. The primary outcome was in-hospital mortality, and multivariate logistic regression was performed on the SCD cohort to identify predictors of in-hospital mortality.

Results

During the study period, there were 991,367 ICU admissions, of which only 1,329 (0.13%) were patients with SCD. Compared to non-SCD patients, those with SCD were younger (43.27% vs. 12.02%), females (52.90% vs. 44.18%), and had a lower comorbidity index (70.05% vs. 67.32%), all with p<0.001. The patients with SCD were more frequently in the lowest income quartile (49.36% vs. 34.59%, p<0.001), have Medicaid (31.38% vs. 17.77%, p<0.001), be treated in public hospitals (16.63% vs. 12.69%, p<0.001), and be in urban teaching hospitals (82.69% vs. 76.14%, p<0.001), predominantly in the South (55.83% vs. 41.91%, p<0.001).

Patients with SCD, compared with the non-SCD group, were less likely to have heart failure/acute coronary syndrome (36.19% vs 38.97%, p=0.043) but more likely to have stroke (12.57% vs. 10.53%, p=0.026), sepsis (47.40% vs. 40.94%, p<0.001), pulmonary embolism (6.55% vs. 3.17%, p<0.001), transfusion reactions (0.07% vs. 0.01%, p<0.001), acute kidney injury (54.10% vs. 46.73%, p<0.001), vasopressor use (12.49% vs. 10.21%, p=0.007), and higher use of continuous renal replacement therapy (20.32% vs. 10.73%, p<0.001). However, there was no significant difference in in-hospital mortality between the SCD and non-SCD groups (33.26% vs. 30.87%, p=0.062).

In the SCD cohort, independent predictors of in-hospital mortality included vaso-occlusive crisis (aOR 1.46, 95% CI 1.10-1.94, p=0.008), sepsis (aOR 1.32, 95% CI 1.02-1.72, p=0.037), pulmonary embolism (aOR 2.50, 95% CI 1.57-4.00, p=0.0001), vasopressor use (aOR 3.21, 95% CI 2.23-4.64, p<0.0001), and CRRT use (aOR 1.69, 95% CI 1.24-2.32, p=0.001) were independent predictors of in-hospital mortality. Conversely, acute chest syndrome was associated with reduced odds of in-hospital mortality (aOR 0.52, 95% CI 0.35-0.77, p=0.001). There was no association between in-hospital mortality and gender, race/ethnicity, insurance type, hospital region, location, or ownership.

Conclusion

Patients with SCD constitute a small proportion of patients admitted to ICU. However, they experienced more severe complications than non-SCD patients, though in-hospital mortality was not statistically different between the two groups. Further studies are needed to explore the unexpected favorable association of acute chest syndrome with in-hospital mortality of patients with SCD managed in the ICU setting.

No relevant conflicts of interest to declare.