A Comparison of Inpatient to Outpatient Screening for Social Drivers of Health (SDOH) in a Safety-Net Hospital

Context: Health systems are working to identify and address the social drivers of health (SDOH). Little is known about screening inpatient adults for SDOH and how SDOH risks and population characteristics compare between outpatient vs inpatient settings.

Objective: Compare SDOH risks between inpatient and outpatient populations in the same health system, and examine differences by age, race, ethnicity, gender, average income and insurance.

Study Design and Analysis: Cross-sectional study

Setting: A safety-net public hospital in Cleveland, Ohio

Population Studied: Adults admitted to two inpatients units during an SDOH screening pilot Feb– April 2023 compared to previously published data from outpatient SDOH screening done Feb–March 2021

Intervention/Instrument: The SDOH screening tool included 11 (inpatient) and 9 (outpatient) domains using the same validated questionnaire items. 5 domains were directly comparable.

Outcome Measures: The primary outcomes were proportion of SDOH risk from the two groups. Demographics including age, gender, race, ethnicity, insurance status and average income were extracted from the EMR (inpatients) and previously published data (outpatients). Chi-squared test was used to determine differences.

Results: 796 inpatients and 5741 outpatients were eligible for screening with 28% (44% self-screen, 56% staff assist) and 60% (100% staff assist) of patients successfully screened in each setting. There were no differences in the proportion of screened patients based on age, gender and race (p=0.05). Differences existed between groups on insurance type and income level, with inpatients representing a higher proportion of Medicaid and lower income patients (p=0.001). For all 5 domains, (financial strain (34% v 6%), food insecurity (27% v 17%), intimate partner violence (6% vs 1%), transportation (17% to 5%) and lack of internet access (17% to 4%)), inpatients had a higher proportion at risk (p=0.001).

Conclusions: Screening for SDOH in different settings has implications for patients and health systems. Inpatients may disproportionately represent those with higher SDOH risk and outpatient data may not adequately represent population demographics or risk. Screening in the inpatient setting has unique challenges and some screening modalities may be more successful than others. To achieve universal screening, it is important to design SDOH programs that target at-risk populations, using the most successful settings and modalities.