Methodological quality of the review: Low confidence

Author: Priya Morjaria, Ian McCormick, Clare Gilbert

Region: China, India, USA, Oman and Nepal (Asia), Brazil, Mexico and Chile (South America), Saudi Arabia (Middle East), and South Africa and Tanzania (Africa)

Sector: Refractive error

Subsector: Compliance and predictors of spectacle wear

Equity focus: No

Study population: School children

Type of programme: School and community based

Review type: Other review

Quantitative synthesis method: Systematic review

Qualitative synthesis method: Not applicable

Background: The World Health Organization (WHO) estimates that there are 19 million children with vision impairment globally, 12 million of whom have uncorrected refractive error. The majority of children have uncomplicated REs which can be readily and cost effectively corrected with spectacles. Despite the benefits of wearing spectacles, there is some evidence that a high proportion of children in many settings do not wear them. An earlier review of school-based approaches to the correction of refractive errors in children included a section spectacle compliance, which included only five studies.

Objectives: The purpose of this review is to collate the evidence on compliance with spectacle wear, factors which predict spectacle wear and reasons for non-compliance among schoolchildren. This information will be of value to those designing and implementing school eye health programmes.

Main findings: Of the 35 included, studies reported more than one of the outcomes of interest (i.e., compliance, predictors and reasons for non-wear). There were 27 studies on compliance, 19 studies on predictors (cross-sectional studies and RCTs) and 13 studies reported reasons for non-wear; seven used qualitative methods and six used structured questionnaires, or details of the methods were not given. Using all definitions, compliance levels ranged from as low as 13.4% (wearing) in to 87.4% (wearing) in the USA. Spectacle wear in studies which defined compliance as wearing spectacles assessed by direct observation ranged from 28% to 73%. Self-reported wear ranged from 58% to 82%. Factors which influenced compliance with spectacle wear identified in this review can be categorised as biomedical, socio-demographic and other factors. Biomedical factors, which are presented first, include UCVA, degree and type of RE, improvement in VA and headaches/eyestrain. Among the quantitative studies, increasing age was associated with lower spectacle wear in four studies, 11-14, whereas in two studies, in India and the USA, younger children were less compliant. Boys were less likely to wear spectacles than girls in eight studies. Odds ratios for greater compliance in girls were reported in an observational study, a cluster RCT in China and in an observational study in the USA [OR 1.72 (95% CI 1.10-2.68) and OR 1.78 (95% CI 1.21-2.62); OR 1.8 (95% CI 1.1-3.2)], respectively. Only a few quantitative studies have investigated whether the type of RE (i.e., myopia, hyperopia or astigmatism) affect spectacle wear. In Tanzania, compliance was zero among hyperopes and astigmats, compared with 43% in myopes. An Indian study reported differences between myopes, hyperopes and emmetropes (better than -0.50D), but did not provide statistical analysis.

All three randomised controlled trials (RCT) showed that free spectacles significantly improved compliance. Parental level of education was assessed in nine studies (seven quantitative and two qualitative). Four of the quantitative studies reported no significant difference in spectacle wear by level of parental education.

Authors concluded that there was considerable variation between studies in how spectacle compliance was defined, the time interval between dispensing the spectacles and assessment, and how compliance was assessed. There is need to standardise all aspects of the assessment of compliance. Further qualitative and quantitative studies are required in a range of settings to assess the biomedical and socio-demographic factors which affect spectacle wear compliance using standard definitions.

Methodology:

The search was wide-reaching, to identify as many studies as possible which reported on the correction of refractive errors in children. Papers were reviewed for inclusion even if compliance was not an expressed purpose of the study.

Literature searches were conducted on Medline, Embase, Global Health and the Cochrane Library. (See appendices for search strategies used.) The date range was January 2000 to November 2017, and there were no language restrictions. Two reviewers independently assessed articles for potential inclusion in the review. In addition, further publications were identified from checking the citations from appropriate studies. The following information was extracted from included studies, as relevant, and entered into an Excel spreadsheet: study design; setting (country) and participants (age, gender, number and comparison groups, if relevant). Main outcomes: how compliance with spectacle wear was defined and assessed, and rates of compliance; predictors of spectacle wear with relevant statistics, and reasons for non-wear. Other outcomes were follow-up rates, use of prescribing guidelines, health education and medium of delivery, and whether students could select their preferred spectacle frames.

Applicability/external validity: Not reported.

Geographic focus: Not discussed.

Summary of quality assessment:

Overall, there is low confidence in the conclusions about the effects of this review. Authors did conduct thorough searches of the literature, and risk of bias assessment of included studies were not performed. In addition, only one reviewer conducted data extraction of studies, increasing the risk of bias.