301 Identifying Unmet Needs for Social Services Among Older Emergency Department Patients

We sought to identity questions that can be used in the ED to identify older adults with unmet social service needs. We selected questions from validated malnutrition and food insecurity screening tools as these problems are prevalent among older ED patients, can be assessed with brief questionnaires, and were hypothesized to accurately predict the need for a range of social services. We conducted a prospective observational study in an academic ED serving a racially and socioeconomically diverse population of older adults. Non-critically ill (ESI Score >2) adults aged 60 years and older were eligible. Screening questions were selected from four validated malnutrition and food insecurity screening tools and compared to self-reported need for and willingness to receive social services. Social services included home-delivered and congregate meals, SNAP, transportation, home health, home care, utility assistance, and insurance and benefits assistance. Of 127 patients, 28% screened positive for malnutrition risk per the Malnutrition Screening Tool (MST), 16% for food insecurity per Hunger Vital Sign (HVS), and 5% (6) for both. A majority of patients who were positive for both MST and HVS expressed a need for (100%) and a desire to receive (83%) social services. Among patients who were negative for both malnutrition and food insecurity screens, a minority expressed a need for (42%) and a desire to receive (31%) social services. As a predictor of self-reported need for one or more social services, MST had a sensitivity of 0.37 and a specificity of 0.84; HVS had a sensitivity of 0.25 and a specificity of 0.98. Combined, MST and HVS had a sensitivity of 0.24 and a specificity of 0.98. As a predictor of receptivity to services, MST and HVS combined had a low sensitivity (0.15) and a high specificity (0.99). Questions assessing malnutrition and food scarcity provide a subgroup of older ED patients with a need for and willingness to receive social services with high specificity but poor sensitivity. Further refinement of these screening tools may be needed to optimize accuracy of a brief ED-based screening to identify patients who may benefit from linkage to social services.