User satisfaction with translation technology is uniformly high and does not track its measured accuracy or usability

Narrative synthesis #

Self-reported satisfaction with translation technology is uniformly high — 96.6% of LEP patients, 83–87% of nurses, means of 4.7–4.9/5 for S-MINDS, 75% of staff rating apps useful — and does not track measured accuracy or usability. The internal contradiction is the evidence: the same Hwang staff who rated the apps useful and would use them again also reported that Google Translate was slow, needed repeated attempts, and was degraded by dialect and accent — satisfaction and measured performance point opposite ways within one study.

Type check for the maintainer: this may be better filed as a methodological caveat (CVT) about the literature than as a claim about the world; it is authored as a CLM as drafted, but the question is flagged. Merge candidates (proposal only — do not touch): C-0230ClaimC-0230Initial AI draftClinicians report overall satisfaction and ease of use with point-of-care interpretation applicationsClinicians report overall satisfaction with point-of-care interpretation applications, citing ease of use, more streamlined communication with patients, and decreased frustration when communicating with limited-language-…, C-0194ClaimC-0194Initial AI draftPatients and nurses report high satisfaction with machine translation software for postoperative symptom assessmentBoth patients and nurses report high satisfaction with neural machine translation software used to assess postoperative symptoms in limited-English-proficiency patients., C-0227ClaimC-0227Initial AI draftLEP patients report high satisfaction with speech translation devices for spoken medication counselingLimited-English-proficiency patients report high satisfaction with speech translation devices used to deliver spoken medication-counseling recommendations, finding the translated audio easy to understand and the device e….