Improving translation quality does not improve LEP patient comprehension, while changing the delivery format does

Narrative synthesis #

On the only comprehension outcome measured in the cluster, improving translation quality did not help: LEP adults understood treatment intent equally poorly whether a SACT booklet was machine- or professionally translated (multivariate OR 0.99), with only 15.7% grasping intent either way. Changing the delivery format did move comprehension — a bilingual (interlinear) consent form raised understanding of non-curative intent from 34.8% to 60.0% (OR 3.73) — though the Total Comprehension Score effect held in univariate but not multivariate models, an internal qualifier worth keeping.

This is one trial plus corroborating qualitative data, not a body of evidence: five of the seven EVDs are Hibbs, with two provider-interview EVDs from Allar. State it as such. Merge candidates (proposal only — do not touch): C-0131ClaimC-0131Initial AI draftProfessional translation of written patient information yields better LEP comprehension than unsupervised machine translationProfessional (human) translation of written patient information produces better comprehension of key treatment information among LEP patients than unsupervised machine translation., C-0132ClaimC-0132Initial AI draftTranslated written patient information booklets alone are insufficient for most LEP patients to comprehend key cancer treatment informationProviding translated written patient information booklets alone — even accessibility-designed, professionally translated ones — is insufficient for most LEP patients to comprehend key cancer treatment information, implic…, C-0134ClaimC-0134Initial AI draftBilingual interlinear consent forms improve LEP patient comprehension of treatment intent during interpreted consent consultationsProviding a bilingual (interlinear) consent form alongside verbal interpretation improves LEP patients' comprehension of treatment intent during interpreted informed-consent consultations, compared with a conventional En…, C-0135ClaimC-0135Initial AI draftBilingual interlinear consent forms increase LEP patient self-rated confidence and understanding during interpreted consentProviding a bilingual (interlinear) consent form increases LEP patients' self-rated confidence in their understanding during interpreted informed-consent consultations, compared with a conventional English-only consent f…, C-0190ClaimC-0190Initial AI draftWord-for-word translation of PROMs without cultural and linguistic validation is not sufficient for valid measurement in LEP populationsSimple word-for-word translation of a patient-reported outcome instrument, without cultural and linguistic validation of its constructs in the target population, is not sufficient for valid measurement in LEP populations….