Perceived barriers to Aedes aegypti larval habitat control among surveyed residents in five Panamanian districts: a cross-sectional health belief model analysis

Background Household source reduction remains central to Aedes aegypti control, but reported intention may coexist with practical and contextual barriers. We examined perceived barriers and related Health Belief Model (HBM) constructs among surveyed residents of five Panamanian districts. Methods This secondary cross-sectional analysis used a non-probability convenience sample collected from October 2024 through February 2025. One respondent per household—the person responsible for checking or eliminating larval habitats—was eligible. Of 2,115 records, 2,091 met the analytical criteria. After outcome distributions were inspected during manuscript revision, the three-item perceived-barriers score (range 3–9; higher values indicate more barriers) was designated the primary analytical outcome; the two-item self-efficacy score was exploratory because of a severe ceiling effect. Multivariable linear models used HC3 robust standard errors and district adjustment. Cumulative-logit models and influence analyses assessed robustness. Results Perceived barriers had a mean of 5.24 (SD 2.34); 42.0% were at the minimum. Self-efficacy had a mean of 5.77 (SD 0.69), with 88.4% at the maximum. Compared with Panamá district, adjusted barrier scores were lower in San Miguelito (adjusted mean difference [AMD] −0.98, 95% CI −1.28 to −0.68) and higher in Chepo (1.10, 0.80 to 1.41) and Arraiján (1.52, 1.09 to 1.96). Greater breeding-site knowledge (−0.17, −0.28 to −0.06) and care-seeking knowledge (−0.51, −0.63 to −0.38) were associated with fewer barriers. The essential-services index was positively associated with barriers (0.36, 0.21 to 0.51); exploratory component analysis suggested this association was primarily driven by sewer access (0.67, 0.45 to 0.90). Ordinal models produced the same directions for the main contextual and knowledge associations, although proportional-odds diagnostics were not satisfied for all predictors. Conclusions Within this convenience sample, district-level heterogeneity in perceived barriers was larger than most individual-level associations despite near-universal reported intention. The findings identify contextual signals for local investigation; they do not estimate population prevalence, actual household practice, or causal effects of residence or services.