Evaluating Intra-household Agreement on Multi-domain Family-Level Social Determinants of Health and Exploring Individual Correlates of Agreement in Two Southern California Family Studies

Prev Sci. 2026 May 12. doi: 10.1007/s11121-026-01914-2. Online ahead of print.

ABSTRACT

Accurate self-reported data on social determinants of health (SDoH) are essential for improving prevention initiatives. Beyond survey content and validation, deciding which household member should complete family-level SDoH assessments can affect data quality. Yet, few studies have explored how family members report on various SDoH, especially in the Latino/Hispanic community experiencing greater health challenges. This study not only examined intra-household agreement on family-level SDoH items but also assessed combined individual SDoH linked to agreement in Southern California, which hosts one of the largest Latino/Hispanic communities in the USA. We analyzed data from 277 adult pairs (n = 554) across Southern California who completed the National Institute on Minority Health and Health Disparities Common Data Elements questionnaire. Each respondent answered 19 household-level items across four domains: Demographics, Economics, Health and Clinical Care, and Housing. Agreement was evaluated using simple or weighted Cohen’s Kappa (range, ≤ 0 to 1). Modified Poisson regression examined associations between agreement and combined individuals’ health literacy, employment, sex at birth, age, birthplace, and ethnicity. Tests were corrected for multiple comparisons. Among respondents, 56.3% were women, participants’ mean age was 41.8 years (standard deviation = 8.6), 93.1% identified as Latino/Hispanic, and 37% reported low health literacy. Agreement across domains varied (Kappa = 0.14-0.85), with higher agreement observed for Demographics and transportation items, and poor agreement on financial adversity, healthcare, and food insecurity items. Agreement on food insecurity items varied by dyad composition: same-sex dyads demonstrated greater agreement than mixed-sex dyads, whereas dyads in which both members identified as Latino/Hispanic showed lower levels of agreement. Variations seem to exist in how household members report on family-level SDoH, particularly for sensitive areas like finances and food access. To improve accuracy, researchers and public health professionals might consider collecting data from multiple household members while encouraging joint responses and accounting for specific when designing assessments.

PMID:42118237 | DOI:10.1007/s11121-026-01914-2