Language Access in Healthcare
EvidenceE-0016Initial AI draft

LEP patients had lower medication learning self-efficacy than English-proficient patients but equal taking self-efficacy

2026-06-053 out · 0 in

Source

Zhang (2018). Medication Related Self- efficacy among Linguistically Diverse Patients with Chronic Illnesses. Journal of Health Care for the Poor and Underserved.

Description #

In a six-language telephone survey of 509 chronically ill adults, LEP patients had significantly lower learning self-efficacy (confidence in learning about their medications) than English-proficient (EP) patients — adjusted −0.59 points (p<.001) — but there was no difference in taking self-efficacy. Self-efficacy is a modifiable antecedent of medication adherence, so the gap is specific to understanding/learning about medications, not the act of taking them.

"patients with limited English proficiency (LEP) had a lower mean learning self-efficacy subscale score (LEP: 14.5, EP: 15.4; p<.001) and no difference in the mean taking self-efficacy subscale score (LEP: 14.4, EP: 14.6; p=.40)." (Zhang, 2018)

"For the learning self-efficacy subscale, LEP patients had a mean score of 0.59 points lower than EP patients (p<.001)" (Zhang, 2018)

Methods Context #

What? #

The observable: medication-related self-efficacy on the Medication Understanding and Use Self-Efficacy (MUSE) scale — an overall score plus two subscales (taking self-efficacy, learning self-efficacy), each 4–16.

"Medication adherence related self-efficacy was assessed using the Medication Understanding and Use Self-Efficacy (MUSE) scale." (Zhang, 2018)

How? #

Cross-sectional six-language telephone survey; multiple-variable linear regression of MUSE scores on English-proficiency status, adjusting for age, gender, education, self-reported health status, and number of prescribed medications, with inverse-probability weights for survey non-response.

"We examined the relationship between English language proficiency and overall MUSE score, as well as each of the two subscale outcome scores (taking self-efficacy and learning self-efficacy) using multiple variable linear regression, adjusting for age, gender, education, self-reported health status, and number of prescribed medications." (Zhang, 2018)

Who? #

509 adults (LEP n=328, EP n=181) with diabetes, hypertension, or hyperlipidemia at Group Health Cooperative (the same integrated-system cohort as Moreno 2016); six languages (English, Cantonese, Mandarin, Korean, Vietnamese, Spanish).

"A total of 509 patients completed the survey and received a $10 gift card in the mail. The overall response rate for the survey was 35.5% (n=509; 34.0% LEP, n=328; and 37.4% EP, n=181)." (Zhang, 2018)

Other Notes #

Same survey cohort as S-0025 (Group Health, n=509) — not statistically independent of it. The authors note the IPW modeling was described in Moreno's prior publication.

Caveats #

  • Self-reported VMI and a non-validated MUSE scale in other languages weaken the self-efficacy findings Receipt of verbal medication information was self-reported (recall bias) and does not confirm what providers actually delivered. The cross-sectional design precludes causal inference, and the MUSE self-efficacy scale was validated only in English speakers — not in the non-English languages surveyed — so cross-language score comparisons are uncertain. The authors also note it is unknown whether MUSE differences map to clinically significant adherence outcomes.