In-Person Navigation Outperforms Signposting in Social Prescribing Trial, With Key Equity Gains for Language Minorities
Key Insights
A randomized controlled trial in Ontario primary care found that comprehensive, in-person navigation (ARC) led to 41% higher access to needed health or social resources compared to signposting to 211-Ontario (search) (50.3% vs 35.8%).
After adjusting for patient characteristics, ARC patients had nearly twice the odds of accessing resources (aOR 1.82; 95% CI, 1.13–2.94), with the greatest benefit seen for social needs (search) such as transportation, housing, and employment.
Francophone patients in the ARC arm achieved French language–concordant services in 92% of cases versus 36% in the signposting arm, demonstrating that culturally responsive navigation can reduce minority language disparities.
A multisite randomized controlled trial conducted across primary care practices in Ontario, Canada, has demonstrated that comprehensive, in-person navigation services significantly outperform simple signposting in helping patients access needed health and social resources. The study, published in the Annals of Family Medicine, compared the Access to Resources in the Community (ARC) (search) model—a longitudinal, bilingual, patient-centered navigation service—against referral to 211-Ontario (search), a free telephone and online information service.
Among all 326 randomized patients, those in the ARC arm were 41% more likely to access at least one needed health or social resource compared to those signposted to 211-Ontario (search) (50.3% vs 35.8%; absolute difference = 14.5%; 95% CI, 3.9%–25.2%). After adjusting for patient characteristics, the odds of access were nearly twice as high in the ARC arm (aOR = 1.82; 95% CI, 1.13–2.94).
Study Design and Population
The trial enrolled 55 primary care practitioners (50 physicians and 5 nurse practitioners) across 12 practices in Greater Sudbury and Eastern Ottawa—two Ontario regions with a high prevalence of Francophones. A total of 326 patients with health and/or social needs (search) were randomized to either the ARC intervention or the 211-Ontario (search) control arm. The 967 needs identified on referral forms related to health (42%), healthy behavior (33%), and social conditions (26%).
Patients in the ARC arm were offered scheduling with a bilingual navigator who provided informational, instrumental, and emotional support over a three-month period. The navigator worked to understand patient motivation, identify priorities and access barriers, and co-develop personalized action plans. In contrast, the 211-Ontario (search) arm received a prepared script describing the service and were encouraged to use its inbound telephone and online information resources.
Primary and Secondary Outcomes
The primary outcome—self-reported access to at least one needed health or social resource—was significantly higher in the ARC arm across all analyses: unadjusted (RR = 1.41; 95% CI, 1.15–1.66), adjusted (aOR = 1.82; 95% CI, 1.13–2.94), and in the complete-case sensitivity analysis restricted to the 237 patients who completed the end-of-study survey (aOR = 1.82; 95% CI, 1.23–2.94). ARC patients also accessed a greater number of resources on average (mean = 0.93 vs 0.67, P = .04).
Patient experience measures strongly favored the ARC model. Among the 157 patients who completed the experience survey (89% in the ARC arm vs 41% in the 211-Ontario (search) arm, P < .001), ARC patients reported significantly higher overall satisfaction (90.4% vs 61.5%, P < .001) and were more likely to report positive experiences across all 13 survey questions addressing effectiveness, communication, and overall satisfaction.
The ARC arm also demonstrated significant improvement in patients' ability to engage in their care, while the 211-Ontario (search) arm showed a decline in this dimension over the study period (P = .01).
Equity Findings for Language Minorities
A key contribution of the study was its pre-specified equity analysis for Ontario's Francophone minority population. By applying an "active offer" approach—inviting patients to participate in the language of their choice—the ARC model achieved French language–concordant service access in 92% of cases, compared to just 36% in the signposting arm (P = .01).
"ARC, which applied an 'active offer' approach, significantly increased access to French language–concordant services," the authors note. "This is highly relevant as linguistic concordance between patient and clinician is associated with better care, good patient-clinician relations, and higher patient satisfaction."
Notably, Francophone participants' overall likelihood of accessing services did not significantly differ by arm; the ARC advantage was specifically in the language concordance of those services.
Exploratory Findings on Modality and Subgroups
Exploratory analyses revealed that access to resources increased monotonically with each additional in-person encounter with the ARC navigator: from 37% for zero in-person encounters to 100% for three or more. This dose-response relationship was not observed with telephone encounters, suggesting that in-person contact may be uniquely important for building the trust and interpersonal connections that facilitate successful navigation.
The relative benefit of comprehensive navigation was concentrated in social needs (search)—including transportation, employment, housing, financial supports, and social connectedness—where ARC participants were more than twice as likely to access a relevant resource (RR = 2.03; 95% CI, 1.23–3.37). Patients with two to six social needs (versus one) had an adjusted odds ratio of 5.75 for access in the ARC arm, suggesting the model may be especially effective for people with more complex needs.
Limitations and Context
The authors acknowledge several limitations. The primary outcome is self-reported access to resources—a proximal measure that does not establish whether access translates into reduced social need, improved health, or changes in health care utilization. The single-navigator design per city maximizes fidelity but limits generalizability. The exploratory subgroup and modality findings, while hypothesis-generating, were not powered for causal inference. Additionally, the differential completion rates of the patient experience survey between arms warrant caution in interpreting those outcomes.
Nevertheless, the study adds important evidence to a growing body of comparative effectiveness research on social care navigation. As noted in an accompanying editorial, the findings suggest that "comprehensive, in person, and sustained support with active attention to language and culture all improve outcomes"—insights with direct implications for how health systems design and target social needs (search) navigation investments.