Global estimates of diabetic retinopathy prevalence and progression in pregnant women with preexisting diabetes: a systematic review and meta-analysis

Authors: Widyaputri F, Rogers SL, Kandasamy R, Shub A, Symons RCA, Lim LL.

Geographical coverage: The USA, Europe (UK, Finland, Italy, Denmark), Saudi Arabia, Israel

Sector: Burden of disease

Sub-sector: Epidemiology, changes in diabetic retinopathy in pregnancy

Equity focus: Not specified

Study population: Pregnant women with pre-existing type 1 or type 2 diabetes

Review type: Systematic review

Quantitative synthesis method: Meta-analysis

Qualitative synthesis method: Not applicable

Background:

Diabetic retinopathy (DR), a leading cause of blindness, affects over 30% people with diabetes mellitus. The risk of irreversible blindness due to DR is greater among women than men, and this risk increases with pregnancy. Moreover, unlike other changes that occur during pregnancy, progression to sight-threatening disease is not reversible. Furthermore, the risk of this progression remains present for up to 12 months postpartum. Although, the standard of care for pregnant women with diabetes (1989) aimed to achieve similar pregnancy outcomes as women without diabetes (including reducing maternal complications such as diabetes related blindness), the prevalence of DR and progression rates ranges from 8% to 63% and 5% to 41.5%, respectively, in this population. This study was therefore conducted to obtain a more precise estimate of DR prevalence and its progression in this group.

Objectives:

To estimate the prevalence of DR and its progression rate in pregnant women with pre-existing type 1 or type 2 diabetes (diagnosed before pregnancy).

Main findings:

The search identified a total of 1,225 articles through database and citation searches, of which 18 observational studies (prospective or retrospective) were included in this review, with 1,464 pregnant women with type 1 diabetes and 262 pregnant women with type 2 diabetes. Fifteen studies (83%) were prospective cohort studies and three (17%) were retrospective cohort studies.

The pooled prevalence of DR in early pregnancy was 52.3% (95% CI, 41.9-62.6) this is higher than the pooled prevalence estimated in the nonpregnant diabetic population (34.6%). The pooled prevalence of DR in early pregnancy for proliferative DR (PDR) was found to be 6.1% (95% CI, 3.1-9.8).

The pooled progression rate per 100 pregnancies for new DR development was 15.0 (95% CI, 9.9-20.8), worsened non-proliferative DR was 31.0 (95% CI, 23.2-39.2), progression from non-proliferative DR to PDR was 6.3 (95% CI, 3.3-10.0), and worsened PDR was 37.0 (95% CI, 21.2-54.0).

DR progression rates per 100 pregnancies were similar between type 1 and type 2 diabetics, except for the development of new DR: in type 1 diabetics it was 15.8 (95% CI, 10.5-21.9) type 2 diabetics 9.0 (95% CI, 4.9-14.8).

In summary, the progression of DR in pregnancy was similar for women with type 1 and type 2 diabetes. However, the prevalence and progression of DR in pregnant women with diabetes remains higher than the nonpregnant population with diabetes, suggesting a need to improve DR management in pregnancy.

Methodology:

Two independent reviewers conducted the searches for relevant articles in Medline, EMBASE and Scopus from inception to June 2021. Observational studies (retrospective or prospective) that investigated pregnant women diagnosed with diabetes before pregnancy (type 1 or type 2 diabetes) with or without DR and published in the English language were included. Reference lists of the included studies were searched to identify additional relevant articles.

Two reviewers independently assessed the literature and conducted data extraction. Studies were appraised for the quality of the data included using an existing model (details shared in the supplementary content of the review). The findings were synthesised using random-effects model meta-analysis to calculate the pooled estimate of prevalence and progression rates with 95% CIs. Pooled estimates were only done using data from studies deemed high quality.

Heterogeneity was assessed with the χ² test and quantified with the I² test. Subgroup analysis was performed to identify potential causes of heterogeneity.

Applicability/external validity:

Authors acknowledge that most included studies are from HICs, so the estimated prevalence of DR in pregnant women likely underestimates the burden in the global population, as patients in low income settings have less access to medical care in pregnancy.

Geographic focus:

The included studies were from the USA, Europe and the Middle East.

Summary of quality assessment:  

Overall, there is medium confidence in the conclusions about the effects of this study. Although the authors used appropriate methods to screen studies for inclusion and extract data of those included in the review, the searches were limited to peer-reviewed articles written in the English language only.