Improving Completion Rates of Transcranial Doppler Ultrasounds in Children with Sickle Cell Disease Using Quality Improvement Efforts: In-Clinic Vs. Population-Based Assessments

Introduction:

People living with sickle cell disease (SCD) are at risk for stroke due to progressive cerebral vasculopathy. The Stroke Prevention Trial in Sickle Cell Anemia (STOP) study showed that transfusion therapy in patients with abnormal cerebral blood flow velocities measured by transcranial doppler (TCD) ultrasound significantly reduced the risk for stroke. Based on the STOP and STOP II studies, the National Heart, Lung and Blood Institute (NHLBI) and American Society of Hematology (ASH) recommend annual TCD screenings for children with HbSS and HbSb0 genotypes from age 2-16. Despite this recommendation, studies show that fewer than half of eligible children with SCD complete annual TCD screening. Recently, Cabana and Treadwell et al. (2020) found high TCD referral and completion rates in a multi-site quality improvement (QI) initiative (85% baseline) using chart review. Another feasible approach to tracking guideline adherence uses administrative claims data which are derived from diagnostic and billing codes from statewide claims. Claims data can be used to assess population estimates for a clinic or a state, for all eligible patients, including those who may not routinely access care. In this study, we use administrative claims to assess TCD completion rates in the same clinics participating in the aforementioned QI initiative. We hypothesized that, population level rates would be lower than those assessed via chart review and that QI strategies may not lead to sustained TCD completion rates.

Methods:

Between August 2017 to August 2018, a QI initiative within the Pacific Sickle Cell Regional Collaborative (PSCRC) was conducted to improve referral and completion rates of TCD screenings. Site leads participated in a monthly QI learning collaborative, implementing and reporting on Plan-Do-Study-Act (PDSA) cycles, with bimonthly chart review data collection. Medicaid administrative claims from the four states with participating clinics, from 2017, 2018 (to assess baseline and post-QI initiative performance) and 2019 (to assess sustainability), were used to assess rates of TCD completion in the eligible pediatric population, using the specifications of a previously validated quality measure by Reeves et al (2019). Annual TCD completion rates and changes in completion rates over time were assessed for each site and state.

Results:

Five sites from four states in the PSCRC were included in the analysis. There was large variability in the number of eligible patients in each clinic (13-75) and state (23-588). Based on administrative claims, TCD clinic-level completion rates at baseline ranged from 41.7% to 69.2% at individual clinics. After 12 months of QI participation, TCD completion rates improved at all

sites (range 4.6% to 29.2%). The site with the largest change improved TCD completion rate from 41.7% to 70.8% (n=24). All but one site had a decrease in TCD completion rate after completing the QI initiative and in 2019, TCD completion rates were within 10% of baseline completion rates at all sites (range -8.2% to 8.3%). At the statewide level, one state had a sustained improvement in TCD completion (improvement from baseline: 8.8%). In three of the four states providing data, TCD completion rates decreased from baseline (range -0.7% to -12.6%).

Discussion:

In a regional collaborative, we found improvements in TCD completion in the setting of a QI initiative focused on TCD, which were not sustained in the year after. This suggests the need for a systems-level approach to improvement, leading to feasible sustainability when no longer the focus of a collaborative. In addition, our data show that, when using administrative claims, rates of TCD completion are lower than rates when using chart review data (41.7% to 69.2% vs. 85% by chart review noted in prior publication). Thus, while site-specific medical record review provides insight into the quality of care for patients seen in the clinic, administrative claims data allow for a global understanding of the quality of care for the clinic population at risk, including those who do not attend clinic regularly. This suggests additional potential focus for quality improvement initiatives, such as systems to optimize outreach to patients who may not routinely access care. This type of outreach may best be done by health plans, potentially in partnership with sickle cell specialists, and would be an important tool for improving health for children with SCD.