Scientific teams that are comprised of different types of researchers have higher research productivity, and there is a need for evidence-based methods to improve the biomedical research workforce. Building Up a Biomedical Research Workforce (Building Up) was a multi-center, cluster-randomized, unblinded controlled trial with one intervention arm and one control arm, conducted at 25 United States academic medical centers. The authors tested the hypothesis that participants from backgrounds underrepresented in science who are randomized to the intervention will have greater numbers of peer-reviewed publications and increased Psychological Capital, compared to the control group.

The study included a 10-month intervention period and follow-up assessments occurring one, two, and three years after the intervention began. The intervention arm received a 10-month intervention with monthly meetings, near-peer mentoring, networking opportunities, and grant- and scientific-writing coursework. Participants in the control arm experienced the usual forms of mentoring, networking, and coursework that their institutions provided.

Of the 220 participants who completed the pre-intervention assessment (98% of all enrolled participants), 71% completed the post-intervention assessment at year 1, 60% at year 2, and 66% at year 3. Individuals in the intervention arm had significantly higher levels of self-efficacy, resilience, and optimism in the three years following the start of the intervention, compared to the control arm.

This finding suggests that the Building Up intervention can increase participants’ Psychological Capital.

Social determinants of health (SDOH) are an important contributor to health status and health outcomes. In this analysis, we compare SDOH measured both at the individual and population levels in patients with high comorbidity who receive primary care at Federally Qualified Health Centers in New York and Chicago and enrolled in the Tipping Points trial.

We analyzed individual- and population-level measures of SDOH in 1,488 patients with high comorbidity (Charlson Comorbidity Index ≥ 4) enrolled in Tipping Points. At the individual level, we used a standardized patient-reported questionnaire. At the population level, we employed patient addresses to calculate the Social Deprivation Index (SDI) and Area Deprivation Index. Multivariable regressions were conducted in addition to qualitative feedback from stakeholders.

Individual-level SDOH are distinct from population-level measures. Significant component predictors of population SDI are being unhoused, unable to pay for utilities, and difficulty accessing medical transportation. Qualitative findings mirrored these results. High comorbidity patients report significant SDOH challenges at the individual level. Fitting a binomial generalized linear model, the comorbidity score is significantly predicted by the composite individual SDOH index (p < 0.0001) controlling for age and race/ethnicity.

Individual- and population-level SDOH measures provide different risk assessments. The use of community-level SDI data is informative in the aggregate but should not be used to identify patients with individual unmet social needs. Health systems should implement a standardized individualized assessment of unmet SDOH needs and build strong, enduring partnerships with community-based organizations that can provide those services.