Background: Care of hospitalized patients with Sickle Cell Disease (SCD) presents unique challenges due to variability in care, the interplay of acute and chronic pain, and the complex social determinants of health among patients with SCD. Despite comprising only a small percentage of admissions at our academic medical center, admissions for acute SCD-related illnesses have a longer length of stay and greater readmission rates leading to disproportionate resource utilization and dissatisfaction for patients and providers.
Purpose: The goal of this quality improvement intervention was to standardize care for patients hospitalized with acute SCD-related illness by emphasizing evidence-based care and creating a predictable approach for providers and patients alike. The tenants of this approach were three-fold: early interdisciplinary involvement, early and aggressive pain management, and deliberate post-discharge planning. Desired primary outcomes included reduced length of stay, reduced emergency department visits, and reduced readmissions.
Description: We developed a clinical algorithm based on the American Society of Hematology’s 2020 guidelines for SCD with input from institutional hospitalists, hematologists, pain specialists, pharmacists, and social workers. The algorithm was integrated into the electronic medical record to provide clinical decision support and improve workflow through embedded orders. We then performed an overall one-year pre-post comparison of targeted outcomes. Using a validated set of ICD-10 codes for acute SCD-related illness, we identified 194 encounters pre-intervention and 176 post-intervention. Length of stay was not reduced post-intervention (6.88 days vs. 7.45 days; p=0.0848). While 7-day readmission rates were reduced in the post-intervention group (37.11% vs. 21.02%; p=0.0002), there was no change in 30-day readmission rates (62.89% vs. 64.20%; p=0.6294). There was a slight decrease in 30-day emergency department returns without readmission (34.09% vs. 20.21%). Encounters during the post-intervention period were more likely to have multidisciplinary input as evidenced by increased hematology consults (23.20% vs. 38.64%) and social work consults (6.70% vs. 22.73%). A greater proportion of admissions in the post-intervention group had an outpatient hematology appointment scheduled within 30-days of discharge (45.88% vs. 53.98%). Perhaps most impressive was a greater than 50% decrease in the rate of discharges against medical advice (AMA) (10.82% vs. 5.11%; p=0.0382). Mortality and intensive care unit (ICU) transfers were rare in both groups (1 death in pre and 1 ICU transfer in post).
Conclusions: This intervention demonstrated success in increasing interdisciplinary involvement, improving post-discharge follow-up, reducing 7-day readmissions, and decreasing AMA discharges during hospitalization for acute SCD-related illness. Limitations to this evaluation are that data was only available for our institution so we could not see if patients followed up or were readmitted elsewhere, and algorithm use was only 25% in the post-intervention group. Future directions include improving algorithm adoption within our own institution and expanding to other hospitals within our healthcare system. To achieve a sustained reduction in readmission, we can also look to integrate enhanced post-discharge support and multidisciplinary involvement in transitions of care outside the hospital.