Bilingual consent form raised Total Comprehension Score (median 4 vs 3) significantly in univariate but not multivariate models

Description #

On the broader secondary Total Comprehension Score (max 6) in the consent phase, participants given the bilingual English–Bengali consent form had a higher median score (4) than those given the English-only form (median 3). This difference reached significance in univariate regression (beta = 0.74, 95% CI 0.007–1.49, p = 0.048) but was attenuated to non-significance after covariate adjustment (multivariate beta = 0.67, 95% CI −0.101–1.434, p = 0.09) (Supplementary Fig. 2). It is therefore a mixed/borderline result: directionally favouring the bilingual form but not robust to adjustment.

"Out of a maximum Total Comprehension Score of 6, participants randomised to receive the bilingual consent form had a median score of 4 compared to 3 for those receiving the English-only consent form (Supplementary Fig. 2)." (Hibbs, 2026, p. 6)

"This difference was statistically significant in univariate regression models (beta=0.74, 95% CI=0.007–1.49, p=0.048) but not multivariate (beta=0.67, 95% CI= −0.101—1.434, p=0.09)." (Hibbs, 2026, p. 6)

Methods Context #

What? #ⓘ

The observable: the Total Comprehension Score (0–6), a composite of critical SACT-understanding items assessed in the consent phase.

"other critical points of SACT understanding within each component (Total Comprehension Score)" (Hibbs, 2026, p. 3)

How? #ⓘ

Score derived from the Consent Comprehension Tool completed immediately after the simulated interpreter-mediated consent consultation; the ordinal outcome was modelled with cumulative-link (ordinal logistic) regression, univariate and covariate-adjusted. See A-0011ArtifactA-0011Initial AI draftBilingual English-Bengali interlinear SACT consent formA newly created, professionally translated bilingual English–Bengali version of the nationally approved generic SACT (systemic anti-cancer therapy) consent form, used during interpreter-mediated informed-consent consulta….

"For the (ordinal) secondary outcomes in both phases, we examined associations using ordinal logistic regression models, specifically with cumulative link models from the R package "ordinal"." (Hibbs, 2026, p. 4)

Who? #ⓘ

Healthy Bengali-/Sylheti-speaking adults with LEP completing the simulated consent exercise (91 analysed; 45 bilingual, 46 English-only).

"Amongst 91 participants completing the simulated consent exercise" (Hibbs, 2026, p. 4)

Other Notes #

Substantive-fidelity flag for reviewers: the primary treatment-intent outcome for the bilingual form was robust to adjustment (OR 3.73), whereas this broader Total Comprehension Score lost significance after adjustment (p = 0.09). Wired as supporting the comprehension claim on the strength of direction and the univariate result, but the multivariate attenuation is recorded honestly here.

Caveats #

• Exploratory study not powered to its objectives, yielding wide confidence intervals This was an exploratory study with no pre-existing data on expected SACT-comprehension rates in this population, so no formal sample-size calculation was possible and the sample was set by available time, funding and community-partner capacity. Consequently the study was not definitively powered to its objectives and the effect-size estimates carry wide confidence intervals — the odds ratios and regression coefficients are imprecise and should be treated as hypothesis-generating rather than definitive.
• Healthy-volunteer, simulated group consent with immediate no-recall comprehension measurement The findings come from healthy volunteers rather than patients with cancer, in a simulation of only partial fidelity to real practice (participants completed consent consultations in small groups rather than individually), with comprehension and confidence measured immediately after each intervention rather than as later recall. Real cancer patients face additional psychological and cognitive demands and greater salience of treatment intent that this design cannot capture, and immediate post-test measurement may overstate durable understanding. Together these external-validity gaps limit direct extrapolation of the comprehension and confidence estimates to real interpreter-mediated oncology consent.