Interventions for improving health literacy in migrants.

Background: Health literacy (HL) is a determinant of health and important for autonomous decision-making. Migrants are at high risk for limited HL. Improving HL is important for equitable promotion of migrants' health.

Objectives: To assess the effectiveness of interventions for improving HL in migrants. To assess whether female or male migrants respond differently to the identified interventions.

Search Methods: We ran electronic searches to 2 February 2022 in CENTRAL, MEDLINE, Embase, PsycInfo and CINAHL. We also searched trial registries. We used a study filter for randomised controlled trials (RCTs) (RCT classifier).

Selection Criteria: We included RCTs and cluster-RCTs addressing HL either as a concept or its components (access, understand, appraise, apply health information).

Data Collection And Analysis: We used the methodological procedures recommended by Cochrane and followed the PRISMA-E guidelines. Outcome categories were: a) HL, b) quality of life (QoL), c) knowledge, d) health outcomes, e) health behaviour, f) self-efficacy, g) health service use and h) adverse events. We conducted meta-analysis where possible, and reported the remaining results as a narrative synthesis.

Main Results: We included 28 RCTs and six cluster-RCTs (8249 participants), all conducted in high-income countries. Participants were migrants with a wide range of conditions. All interventions were adapted to culture, language and literacy. We did not find evidence that HL interventions cause harm, but only two studies assessed adverse events (e.g. anxiety). Many studies reported results for short-term assessments (less than six weeks after total programme completion), reported here. For several comparisons, there were also findings at later time points, which are presented in the review text. Compared with no HL intervention (standard care/no intervention) or an unrelated HL intervention (similar intervention but different information topic) Self-management programmes (SMP) probably improve self-efficacy slightly (standardised mean difference (SMD) 0.28, 95% confidence interval (CI) 0.06 to 0.50; 2 studies, 333 participants; moderate certainty). SMP may improve HIV-related HL (understanding (mean difference (MD) 4.25, 95% CI 1.32 to 7.18); recognition of HIV terms (MD 3.32, 95% CI 1.28 to 5.36)) (1 study, 69 participants). SMP may slightly improve health behaviours (3 studies, 514 participants), but may have little or no effect on knowledge (2 studies, 321 participants) or subjective health status (MD 0.38, 95% CI -0.13 to 0.89; 1 study, 69 participants) (low certainty). We are uncertain of the effects of SMP on QoL, health service use or adverse events due to a lack of evidence. HL skills building courses (HLSBC) may improve knowledge (MD 10.87, 95% CI 5.69 to 16.06; 2 studies, 111 participants) and any generic HL (SMD 0.48, 95% CI 0.20 to 0.75; 2 studies, 229 participants), but may have little or no effect on depression literacy (MD 0.17, 95% CI -1.28 to 1.62) or any health behaviour (2 studies, 229 participants) (low certainty). We are uncertain if HLSBC improve QoL, health outcomes, health service use, self-efficacy or adverse events, due to very low-certainty or a lack of evidence. Audio-/visual education without personal feedback (AVE) probably improves depression literacy (MD 8.62, 95% CI 7.51 to 9.73; 1 study, 202 participants) and health service use (MD -0.59, 95% CI -1.11 to -0.07; 1 study, 157 participants), but probably has little or no effect on health behaviour (risk ratio (RR) 1.07, 95% CI 0.91 to 1.25; 1 study, 135 participants) (moderate certainty). AVE may improve self-efficacy (MD 3.51, 95% CI 2.53 to 4.49; 1 study, 133 participants) and may slightly improve knowledge (MD 8.44, 95% CI -2.56 to 19.44; 2 studies, 293 participants) and intention to seek depression treatment (MD 1.8, 95% CI 0.43 to 3.17), with little or no effect on depression (SMD -0.15, 95% CI -0.40 to 0.10) (low certainty). No evidence was found for QoL and adverse events. Adapted medical instruction may improve understanding of health information (3 studies, 478 participants), with little or no effect on medication adherence (MD 0.5, 95% CI -0.1 to 1.1; 1 study, 200 participants) (low certainty). No evidence was found for QoL, health outcomes, knowledge, health service use, self-efficacy or adverse events. Compared with written information on the same topic SMP probably improves health numeracy slightly (MD 0.7, 95% CI 0.15 to 1.25) and probably improves print literacy (MD 9, 95% CI 2.9 to 15.1; 1 study, 209 participants) and self-efficacy (SMD 0.47, 95% CI 0.3 to 0.64; 4 studies, 552 participants) (moderate certainty). SMP may improve any disease-specific HL (SMD 0.67, 95% CI 0.27 to 1.07; 4 studies, 955 participants), knowledge (MD 11.45, 95% CI 4.75 to 18.15; 6 studies, 1101 participants) and some health behaviours (4 studies, 797 participants), with little or no effect on health information appraisal (MD 1.15, 95% CI -0.23 to 2.53; 1 study, 329 participants) (low certainty). We are uncertain whether SMP improves QoL, health outcomes, health service use or adverse events, due to a lack of evidence or low/very low-certainty evidence. AVE probably has little or no effect on diabetes HL (MD 2, 95% CI -0.15 to 4.15; 1 study, 240 participants), but probably improves information appraisal (MD -9.88, 95% CI -12.87 to -6.89) and application (RR 1.51, 95% CI 1.29 to 1.77) (1 study, 608 participants; moderate certainty). AVE may slightly improve knowledge (MD 8.35, 95% CI -0.32 to 17.02; low certainty). No short-term evidence was found for QoL, depression, health behaviour, self-efficacy, health service use or adverse events. AVE compared with another AVE We are uncertain whether narrative videos are superior to factual knowledge videos as the evidence is of very low certainty. Gender differences Female migrants' diabetes HL may improve slightly more than that of males, when receiving AVE (MD 5.00, 95% CI 0.62 to 9.38; 1 study, 118 participants), but we do not know whether female or male migrants benefit differently from other interventions due to very low-certainty or a lack of evidence.

Authors' Conclusions: Adequately powered studies measuring long-term effects (more than six months) of HL interventions in female and male migrants are needed, using well-validated tools and representing various healthcare systems.