Evaluating the impact of the health navigator model on housing status among people experiencing homelessness in four european countries

Background: People experiencing homelessness (PEH) face significant health disparities and systemic barriers to healthcare, elevating their risk for cancer and other chronic diseases. To tackle PEHs' challenges in accessing cancer preventive care, the CANCERLESS project implemented the Health Navigator Model (HNM)-a person-centered intervention that utilizes trained Health Navigators to provide tailored support and facilitate service access. Recognizing housing as a key determinant of health, this analysis assessed changes in housing status associated with participation in the HNM among CANCERLESS participants in Austria, Greece, Spain, and the UK. 

Methods: This was a secondary analysis of cross-national data collected during a single-arm interventional study. Of 652 enrolled PEH, 277 (42.5%) completed the HNM intervention follow-up and were included in the analysis. Changes in housing status from baseline to follow-up were categorized using the European Typology of Homelessness and Housing Exclusion (ETHOS) and treated as an ordered outcome. Descriptive statistics were complemented by a cumulative link mixed model with a participant random intercept to estimate the association between time (follow-up vs. baseline) and housing transitions among completers, adjusting for age, residence/legal status, and daily smoking. 

Results: Participants had a mean age of 47.4 (SD 13.8), primarily identified as male (64.1%), reported upper secondary education (33.9%), and were from Western European countries (39.7%), with varying housing situations. Among intervention completers, time (follow-up vs. baseline) was associated with higher odds of being in a higher ETHOS category (OR = 1.49, 95% CI = 1.02-2.20, p = 0.042), consistent with a modest improvement in housing status. Larger estimates were observed among migrants without legal documents (OR = 24.13, 95% CI = 6.41-90.89, p < 0.001), while daily smoking was associated with lower odds (OR = 0.33, 95% CI = 0.11-0.96, p = 0.041); other residence status categories were not statistically significant. 

Conclusions: Suggesting that tailored, navigation-based models, such as the HNM, may be linked to improved housing stability for PEH, these findings can inform piloting and context-aligned integration of the HNM within public health strategies as an alternative approach to address the complex, interconnected health and social needs of PEH. However, the lack of a comparison group and high attrition limit the results' conclusiveness, and future evaluations should aim to include assessments of housing-associated contextual factors.