Background: The widely published socioeconomic effects of the COVID-19 pandemic motivated us to address social determinants of health on our inpatient pediatric unit. Screening for food insecurity (FI) was recommended by the American Academy of Pediatrics (AAP) for all medical care settings in 2015. Our aim was to implement universal screening and intervention for FI on our unit using the validated AAP Hunger Vital Signs (HVS) questionnaire.
Methods: Our pediatric unit is a 30-bed open medical/surgical floor with ~4000 admissions annually serving children from the Lehigh Valley metropolitan area and surrounding rural counties. A fully integrated EHMR (Epic™) and access to a charitable fund were existing assets. A multidisciplinary team approach and several PDSA-cycles were employed. We evaluated screening rates with daily compliance checks and periodic run-charts, and used retrospective chart review to compare demographics of food insecure families with published data. In cycle 1, HVS questions were asked by the nurse after a relationship with the patient and family was established, and responses were documented with an EMHR “smart phrase.” The pediatric hospitalists were then charged with assessing acuity of need and intervening with applicable referrals to food pantries, social work, or charity funds. In cycle 2, screening was moved to the time of admission due to observed poor compliance later in the hospital course. In cycle 3, we added a third question to the HVS to streamline assessment of acuity of need and interventions. In cycle 4, HVS were embedded in the EHMR nursing intake flowsheet.
Results: Over 14 weeks, 506/1041 (49%) patients were screened for FI with a positive rate of 8.6%. The screening rate was adversely affected by periods of nursing staff shortage and improved with EHMR upgrade. Retrospective analysis of demographics of food insecure families shows overrepresentation of Hispanic/Latino ethnicity (53% vs. 28% in all pediatric visits), predominance of urban (74%) over suburban (14%) and rural (12%) residence, and public insurance status (81%). Food insecure patients have a wide range of BMI and hospital diagnoses.
Conclusions: The screening rate of 60% at the beginning of the project reflects the high level of nurse engagement and perceived relevance of FI screening. The initial process was vulnerable to nurses’ workload and the screening rate only improved in PDSA-cycle 4. While the goal of universal screening was not met, our project demonstrates that FI screening is feasible in the inpatient setting, and positivity rate and demographics are in line with published data. Our analysis confirms that FI is an “invisible” social determinant of health not tied to BMI, diagnosis, or insurance status, and thus universal screening is necessary.
