Area of Residence As a Screening Tool for Sickle Cell Disease Outcomes in Chicago Pediatric Patients

Background: Sickle cell disease (SCD) is a chronic condition that, in the United States, largely affects a socioeconomically vulnerable group of patients. We have previously shown that our patient population is three times more likely to be food insecure (FI) as compared to the national average. Furthermore, we have correlated proxy variables for social determinants of health (SDOH), like FI, to worsened patient quality of life (QOL), increased healthcare utilization (HCU), poorer clinical disease outcomes (CDO), and decreased uptake of standard of care (USC). We seek to identify additional predictors of which patients may be more vulnerable to suffer from their disease status (due to SDOH) in hopes of intervening upstream and improving outcomes.

Area of residence (AOR) has been identified as a modifiable factor impacting morbidity in other chronic diseases. We are especially interested in investigating this relationship in our Chicago SCD patients as numerous studies have demonstrated associations between health outcomes and the Chicago "neighborhood” one lives in. Often these studies use the census measured household hardship index (HHI) - a composite of six socioeconomic markers like income and household crowding. We hypothesize that AOR with higher HHI is associated with decreased QOL, increased HCU, worse CDO, and decreased USC.

Methods: Subjects with SCD ages 2- 24 years were recruited during routine SCD visits from June 2015- June 2019. We performed a baseline survey using the PedsQLTM Sickle module, which were scored and transformed via established methods (range 0-100; ≤60 indicating low QOL). Chart review captured HCU variables (number of ER visits, inpatient admissions); CDO variables (annual rate of vaso-occlusive events (VOE), acute chest syndrome (ACS), additional complications); and USC variables (compliance with hydroxyurea and influenza vaccination). AOR was assigned utilizing patient addresses, mapping them to one of 77 "Chicago neighborhoods", and then extracting the Chicago census’ data on HHI for that neighborhood. Pearson correlation coefficients and linear or logistic regressions, were performed to examine associations amongst these variables.

Results: There were 115 surveyed patients during the study period, 56% were female and 69% genotype HgbSS. Only 67 patients lived in Chicago city limits and were included in analysis regarding AOR related HHI. The median HHI was 51.4, while 82% of patients had a HHI between 40 and 60. Six patients (9%) had a score above 60 and six (9%) had a score below 40. No statistically significant difference in HHI was observed between patients with different Hgb genotype, gender, or adult and young adolescent (AYA) status. Average QOL was 72 and 24% of patients reported experiencing low QOL, however there was no statistically significant relationship between HHI and total QOL score. "Pain management” was the only individual QOL domain negatively associated with hardship index (r=-0.35, p=0.005). Average total number of ER visits and admissions was 6.9, ACS rate was 0.1 per year, and VOE rate was 0.4 per year. Compliance with flu vaccination and hydroxyurea was 61% and 40% respectively. No significant association of HHI with HCU, CDO, or USC was seen.

Conclusion: This study supports our previous observations that our SCD patient population suffers from low quality of life, poor clinical disease outcomes, high healthcare utilization, and low uptake of standard of care. However, while we have shown markers of SDOH like FI and AYA status to be correlated with these variables, such association was generally not observed with AOR HHI. This held true even when doing sub analyses with the independent components of the HHI. This is surprising given the establishment in the literature of this index as a marker for healthcare outcomes in other diseases. Perhaps our biggest limitation was the small sample size of patients who lived in Chicago and for whom data on HHI was available. Not only did this affect the power of our study, but it is possible that the patients excluded may have had characteristics different from those living within Chicago city limits. Future study could be geared at investigating differences between these two groups. We also excited to design interventional studies based on the SDOH we have shown to correlate with poor patient outcomes to improve care for our SCD patients.

No relevant conflicts of interest to declare.