Screening for Social Risk Factors: A Comparison of Two Multi‐Domain Social Risk Screening Tools

Abstract

Research ObjectiveAmidst growing interest in identifying patients' social and economic risks in the context of health care delivery, many health care organizations have developed recommendations for multidomain social risk screening. A 10‐item instrument developed by the Centers for Medicare and Medicaid Innovation (CMMI) includes six questions on material hardship (housing stability and quality, food security, transportation access, utilities security) and four on interpersonal violence. A set of measures recommended by the National Academy of Medicine (NAM) includes one question on financial strain (difficulty paying for basic needs) and four on intimate partner violence (IPV), in addition to behavioral topics like smoking, alcohol use, and stress. We field‐tested both the CMMI and NAM questions to compare (1) rates of screening positive for socioeconomic risk factors (material hardship, financial strain, personal safety); and (2) patient acceptability.Study DesignCross‐sectional survey design comparing patient responses to surveys with the CMMI versus NAM questions. Descriptive analyses were done using chi‐square with two‐sided Fisher's exact tests.Population StudiedA convenience sample of adult English and/or Spanish speaking/reading patients at three participating primary care clinics in San Francisco, Chicago, and Boston.Principal FindingsOf 457 respondents, 56.2% (127/226) screened positive for any material hardship on the CMMI tool; 51.1% (118/231) screened positive for the single comparative question on the NAM tool that measured global financial strain (P = .302). Using the recommended screening rubrics from each agency, 1.3% (3/226) screened positive on the CMMI interpersonal violence measure vs 10.0% (23/231) on the NAM IPV measure (P < .001). More patients indicated interest in assistance with material hardship on the CMMI tool compared with those indicating interest in assistance for financial strain on the NAM measure (38.7% vs 21.2%, respectively, P < .001). Differences in interest in assistance with safety did not reach statistical significance (1.8% vs 5.2%, P = .072). There were no statistically significant differences in the perceived appropriateness of screening (85.8% vs 79.3%, respectively, P = .072). Patient comfort incorporating social risk factor results into the electronic health record was also similar (65.6% vs 63.4%, P = .631). Few patients reported discomfort answering specific questions on either questionnaire: 4.9% reported discomfort answering questions about material hardship on the CMMI tool vs 5.7% reported discomfort with the NAM question about financial strain (P = .841). Similarly, 1.6% reported discomfort answering the CMMI safety measure vs 2.9% on the NAM (P = .544).ConclusionsOverall, acceptability of the CMMI and NAM recommended social risk factor measures was high. Rates of patients screening positive for socioeconomic risks were similar across the two measure sets, though more patients indicated interest in assistance when answering the CMMI items that asked about specific socioeconomic risks. There were significantly lower rates of positive screens for personal safety using the CMMI tool. More research is warranted on both the validity and comparative validity of each set of measures.Implications for Policy or PracticePatients reported high acceptability of both the CMMI and NAM social risk screening measures. Measure selection should depend on the population served, screening goals and resources available for providing assistance.Primary Funding SourceThe Commonwealth Fund.