Background: Social determinants affect health, yet there are few systematic clinical strategies in primary care that leverage electronic health record (EHR) automation to facilitate screening for social needs and resource referrals. An EHR-based social determinants of health (SDOH) screening and referral model, adapted from the WE CARE model for pediatrics, was implemented in urban adult primary care. Objectives: This study aimed to: (1) understand the burden of SDOH among patients at Boston Medical Center; and (2) evaluate the feasibility of implementing a systematic clinical strategy to screen new primary care patients for SDOH, use EHR technology to add these needs to the patient’s chart through autogenerated ICD-10 codes, and print patient language-congruent referrals to available resources upon patient request. Research Design: This observational study assessed the number of patients who were screened to be positive and requested resources for social needs. In addition, we evaluated the feasibility of implementing our SDOH strategy by determining the proportion of: eligible patients screened, providers signing orders for positive patient screenings, and provider orders for resource referral guides among patients requesting resource connections. Results: In total, 1696 of 2420 (70%) eligible patients were screened. Employment (12%), food insecurity (11%), and problems affording medications (11%) were the most prevalent concerns among respondents. In total, 367 of 445 (82%) patients with ≥1 identified needs (excluding education) had the appropriate ICD-10 codes added to their visit diagnoses. In total, 325 of 376 (86%) patients who requested resources received a relevant resource referral guide. Conclusions: Implementing a systematic clinical strategy in primary care using EHR workflows was successful in identifying and providing resource information to patients with SDOH needs.

The aim of this study was to identify patients with methicillin-resistant Staphylococcus aureus (MRSA) bacteremia with low risk of infective endocarditis (IE) who might not require routine trans-esophageal echocardiography (TEE). We retrospectively evaluated 398 patients presenting with MRSA bacteremia for the presence of the following clinical criteria: intravenous drug abuse (IVDA), long-term catheter, prolonged bacteremia, intra-cardiac device, prosthetic valve, hemodialysis dependency, vertebral/nonvertebral osteomyelitis, cardio-structural abnormality. IE was diagnosed using the modified Duke criteria. Of 398 patients with MRSA bacteremia, 26.4 % of cases were community-acquired, 56.3 % were health-care-associated, and 17.3 % were hospital-acquired. Of the group, 44 patients had definite IE, 119 had possible IE, and 235 had a rejected diagnosis. Out of 398 patients, 231 were evaluated with transthoracic echocardiography (TTE) or TEE. All 44 patients with definite IE fulfilled at least one criterion (sensitivity 100 %). Finally, a receiver operator characteristic (ROC) curve was obtained to evaluate the total risk score of our proposed criteria as a predictor of the presence of IE, and this was compared to the ROC curve of a previously proposed criteria. The area under the ROC curve for our criteria was 0.710, while the area under the ROC curve for the criteria previously proposed was 0.537 (p < 0.001). The p-value for comparing those 2 areas was less than 0.001, indicating statistical significance. Patients with MRSA bacteremia without any of our proposed clinical criteria have very low risk of developing IE and may not require routine TEE.