In an article being early released this week in Pediatrics, Dr. Michael Luke (Children’s Hospital of Philadelphia) and colleagues from Philadelphia, Denver, and Cincinnati conducted a large cross-sectional study to examine any association between neighborhood opportunity as measured by the Childhood Opportunity Index (COI) and patient-level Health Related Social Needs (HRSNs) as assessed by an adapted version of the WE CARE screening tool in the inpatient hospital setting (10.1542/peds.2024-069735). They also aimed to assess the predictive value of the COI for identifying HRSNs.
The COI is a 28-item census-tract level measure of neighborhood health, with the main domains of (1) educational opportunities (e.g. preschool availability), (2) health and environment (e.g. green space access), and (3) social and economic indicators. The adapted WE CARE Screener examines 6 domains, including food insecurity, difficulty with transportation to medical appointments, difficulty paying for utilities, housing concerns, intimate partner violence (IPV), and caregiver mental health concerns.
This study included all 1,096 inpatients screened from July 2022 to August 2023 as part of an HRSN identification and referral program at the lead author’s home institution. Patients’ home address information was used to calculate COI. Most families (57%) lived in neighborhoods with low or very low COI index, and almost half (46%) had at least one HRSN.
The authors found that:
- Lower COI was associated with higher risk of reporting any HRSN or >2 HRSNs.
- Lower COI was associated with higher risk of reporting needs for food, housing, transportation, and utility help, but not IPV or mental health concerns (adjusted analyses).
- COI had low positive predictive value for reporting any HRSNs (maximum of 56% for any need).
The authors concluded that, though it remains a valuable tool for prioritizing neighborhood-level interventions, COI is a poor predictor of household-level HRSNs and cannot replace or guide patient-level screening for HRSNs. See their excellent video abstract for a summary of the work.
In an accompanying commentary, Drs. Lucy Marcil (Boston Medical Center), Deana Around Him (Child Trends, Rockville MD), and Emily Miller (Brown University) discuss the advantages and disadvantages of differing screening approaches (for example, universal versus targeted screening), and this provides a useful perspective (10.1542/peds.2025-070604).
They point out that mental health is one of 6 domains in the study’s screening tool, yet it is actually a highly prevalent medical problem, meriting focused screening that respects associated social stigma. Another screener domain is IPV, which is likely not best described as a “social need.” Screening of pregnant and postpartum people and women of reproductive age for IPV is recommended, yet requires an arguably even more careful approach than mental health screening. The commentary speaks frankly to the critical importance of conducting “systematic, culturally responsive, non-stigmatizing, and pragmatic” screening for HRSN.
In summary, Dr. Luke and colleagues have done valuable work differentiating population-level risk from individual patient and family social needs. Their work is an excellent springboard for future work that can identify best screening practices, optimally individualized by setting and families served.
