Screening of diabetic retinopathy using teleophthalmology to complement human resources for eye health: a systematic review and meta-analysis

Authors: Zikhali T, Kalinda C, Xulu-Kasaba, ZN.

Geographical coverage: Spain, the USA, UAE, India

Sector: Service delivery

Sub-sector: Diabetic retinopathy screening, teleophthalmology, case detection

Equity focus: Not specified

Study population: Diabetic patients with ocular complications or diabetic retinopathy

Review type: Effectiveness review

Quantitative synthesis method: Meta-analysis and narrative synthesis

Qualitative synthesis method: Not applicable

Background:

Regular screening of the retina is a successful strategy for averting blindness caused by diabetic retinopathy (DR). However, the dependence on ophthalmologists for DR screening poses a challenge as the individuals with ocular conditions face difficulties in accessing this screening in resource-limited settings.

Main findings:

The searches identified 146,533 articles, of which nine on diabetic retinopathy and teleophthalmology were included in this review. All studies included were cross-sectional in design, eight out of nine studies were from high income countries (HICs) and one was from a low income country (LIC), India. These studies included 322,725 participants;97,042 of these were identified by DR screening as true positives. In most of the included studies, DR screening was conducted at primary health care level, with only one study being carried out at a private multi-speciality hospital. The commonly used fundus camera in the included studies was the non-mydriatic fundus camera, whereas only one study used a mydriatic agent.

The estimated pooled prevalence of DR in the selected studies was 29 % (95% CI: 10-64%), lower than the previously reported global prevalence of 34.6% (2012). Authors stated that the use of DR screening in HICs was evidence that DR screening was effective in reducing the burden of DR in these countries where it is in decline. The lack of such progress in LICs is linked to the lack of adequate DR screening.

This review summarised how non-ophthalmic health workers and nurses given relevant training in fundus photography can produce sufficient quality gradable pictures that can be interpreted by optometrists and ophthalmologists, lowering the overall workload of the latter, often scarce, workforce thus expanding the reach of treatment and screening for DR in LICs. The use of teleophthalmology, in which images can be graded and reviewed remotely would expand the reach of DR screening in LICs.

The authors recommended using teleophthalmology for DR screening at a primary health care level to allow for early intervention and reduce preventable blindness. In addition, the authors proposed studying the impact of teleophthalmology in LICs. Furthermore, the use of a standardised screening method/protocol would help to compare the findings from different studies.

Methodology:

A literature search was conducted using PubMed, Google Scholar and EBSCO Host databases (Health source: Nursing/Academic Edition), CINAHL and Academic Search Complete to identify studies on human subjects, published from 1 January 2014 to 8 October 2019, and reporting both ocular manifestations of diabetes (DR) and teleophthalmology. The abstract and full-text screening was conducted by two independent reviewers. The Mixed Method Appraisal Tool (MMAT) was used to study quality, and this was applied by two independent reviewers.

Data was extracted from studies using a standardised charting tool, and the results were synthesised using a random-effect model meta-analysis to estimate the pooled prevalence estimates and 95% CI for DR. Heterogeneity was assessed using the I² test and Q statistics, while publication bias was assessed using the funnel and Doi plots.

Applicability/external validity:

The authors discussed the relevance of the topic to low income contexts where health workers and resources are scarce and teleophthalmology may enable much wider DR screening reach if images can be assessed and interpreted remotely.

Geographic focus:

There was no intentional geographic focus; a global search yielded that eight out of nine studies were from HICs and one was from a LIC (India). The thematic focus was the use of teleophthalmology for the early diagnoses and management of DR, with a discussion of relevance to resource-poor settings, including low income countries.

Summary of quality assessment:

In general, the conclusions of this review were met with low confidence. The research questions posed by the authors of the review were not substantiated by the conducted data analysis or narrative synthesis. As a result, it is impossible to establish a connection between the findings and the application of teleophthalmology or the research questions. Moreover, significant methodological limitations were identified, such as the absence of detailed information on the data extraction process from the included studies, the restricted literature searches, and the lack of information on the quality evaluation of the studies included in the review.