Article Text

Effect of mode of delivery on incidence of retinopathy of prematurity: a systematic review and meta-analysis
  1. Vera Sumual1,2,
  2. Reynardi Larope Sutanto3,
  3. Angelina Patricia Chandra3
  1. 1Department of Ophthalmology, Prof. R. D. Kandou General Hospital, Manado, Indonesia
  2. 2Department of Ophthalmology, Faculty of Medicine, Sam Ratulangi University, Manado, Indonesia
  3. 3Faculty of Medicine, Universitas Indonesia, Jakarta, Indonesia
  1. Correspondence to Dr Vera Sumual; verasumual{at}yahoo.com

Abstract

Background Retinopathy of prematurity (ROP) is a leading cause of childhood blindness worldwide. Prompt diagnosis and treatment are crucial in ROP management. Thus, the identification of prominent risk factors could facilitate immediate action. Among various risk factors, the effects of mode of delivery on ROP remain unclear. Therefore, this study aims to assess the association between different modes of delivery on ROP incidence.

Methods and analysis Comprehensive literature search was conducted on PubMed, ProQuest, EBSCOHost and Cochrane databases, to evaluate the association of mode of delivery—vaginal delivery or caesarean section (c-section)—and the incidence of ROP from inception to December 2023. Random-effects meta-analysis was performed to estimate the pooled OR along with their 95% CIs.

Results This review included 5 cohort studies involving 2048 babies. A higher incidence of ROP was observed in infants born through vaginal delivery compared with caesarean section. Meta-analysis showed that C-section decreased the unadjusted odds of having ROP infants by 46% with low heterogeneity (OR 0.54 (95% CI 0.40 to 0.73); I2=40.73%). However, pooled adjusted effects were statistically insignificant with moderate heterogeneity (adjusted OR 0.59 (95% CI 0.28 to 1.23); I2=70.51%), possibly stemming from multiple variations in the controlled variables of each study.

Conclusion Despite varying statistical significance, our findings underscore the crucial need to comprehend the influence of delivery mode on neonatal ophthalmic outcomes. Due to a limited number of existing studies, further research is needed to confirm the association.

PROSPERO registration number CRD42023486278.

  • Retina
  • Epidemiology
  • Neovascularisation
  • Public health
  • Vision

Data availability statement

Data sharing not applicable as no datasets generated and/or analysed for this study.

http://creativecommons.org/licenses/by-nc/4.0/

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

WHAT IS ALREADY KNOWN ON THIS TOPIC

  • Retinopathy of prematurity (ROP) is a significant cause of preventable blindness in premature infants, especially in low-income to middle-income countries where diagnostic challenges hinder prompt treatment.

  • Mode of delivery is a possible important risk factor of ROP, especially considering the rise of caesarean section (C-section) use in preterm labour.

WHAT THIS STUDY ADDS

  • This systematic review and meta-analysis summarised the current state of evidence on the association between mode of delivery and ROP incidence, acknowledging potential protective effect of C-section which might help in identifying preterm neonates susceptible to ROP.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

  • The observed association highlights the critical importance of understanding the impact of delivery mode on neonatal ophthalmic outcomes, which could lend support towards increased use of C-section in preterm labour.

Introduction

Retinopathy of prematurity (ROP), a proliferative vitreoretinopathy affecting premature infants, is a leading cause of avoidable blindness in many parts of the world. In 2010, an estimated 184 700 neonates around the world developed ROP of varying stages, of which more than 30 000 became visually impaired or blind.1 Although incidence has been continually decreasing since the advent of various ROP treatments in the 1990s, ROP continues to affect low-income to middle-income countries, especially in Asia, where it affects around 21% of all preterm infants.2

In cases of severe ROP, neonates must undergo ablative surgery within 72 hours to prevent vision loss. Therefore, prompt diagnosis and treatment are crucial in ROP management.3 Thus, identification of key risk factors is important in alleviating the burden of late ROP diagnosis.4 Unfortunately, associations between ROP and many of its hypothesised risk factors remain to be investigated. Among them, mode of delivery is a possible important risk factor of ROP, especially considering the rise of caesarean section (C-section) as the main mode of delivery in preterm neonates. However, studies have shown varying results between the modes of delivery, namely C-section and vaginal delivery.5–7 To the best of our knowledge, there has been no systematic review of the effect of mode of delivery on ROP incidence. Thus, this prognostic systematic review and meta-analysis aims to review the association of the mode of delivery on the incidence of ROP.

Materials and methods

This review has been previously registered in the PROSPERO database (CRD42023486278). The study is reported in accordance with the Preferred Reporting Items for Systematic Review and Meta-Analysis statement.8

Search strategy

We followed a population, index and comparator prognostic factor, outcome, timing and setting framework as shown in online supplemental table S1. Literature searches were systematically performed by screening for eligible studies published from inception to 6 December 2023 through PubMed, ProQuest, EBSCOHost and Cochrane Controlled Register of Trials. Keywords used for literature searching were (retinopathy of prematurity[Meshi Terms]) AND (((mode of delivery) OR (((caesarean section) OR (caesarean birth)) OR (c-section))) OR (((vaginal delivery) OR (vaginal birth)) OR (normal delivery))) with no language restrictions. Screening and searching of relevant studies were conducted by two independent investigators (RLS and APC), and any discrepancies were resolved by a third investigator (VS) in a blinded fashion, meaning that the third investigator made decisions without knowledge of which initial investigator (RLS or APC) made the inclusion or exclusion decisions. Any titles or abstracts deemed potentially eligible for inclusion by either investigator were retrieved for full-text assessments.

Eligibility criteria

Studies would be included in this review if they meet the following inclusion criteria: (1) study design: cohort and cross-sectional studies; (2) language: English; (3) comparison and intervention: mode of delivery by vaginal delivery and C-section and (4) outcome parameter: incidence of ROP. In contrast, the exclusion criteria were (1) no available data to be extracted and (2) irretrievable full-text articles.

Data extraction and quality assessment

The following data were extracted from each included study: author and publication year, study design, study location, gestational age, mode of delivery, incidence of ROP and outcomes related to ROP incidence. Baseline study characteristics are shown in table 1. The included studies were further assessed for methodological quality using the Newcastle Ottawa Scale (NOS).9

Table 1

Baseline study characteristics

The NOS was used to assess the perceived risk of bias in cohort studies. Data extraction and bias assessments were conducted by two independent authors (RLS and APC) and any discrepancies were resolved by a third independent author (VS).

Statistical analysis

Meta-analysis was performed using STATA V.18.0 (StataCorp). Both adjusted and unadjusted estimates were pooled in the meta-analysis. OR and 95% CI were used as the pooled effect estimates. Relative risks were converted using the formula outlined by the Cochrane Handbook V.6.4.10 As clinical heterogeneity was anticipated, effect sizes were pooled using a DerSimonian-Laird inverse variance, random-effects model.11 Heterogeneity between studies was assessed with the I2 value and classified as negligible (0%–25%), low (25%–50%), moderate (50%–75%) or high (>75%).

Subgroup analyses were performed to address potential sources of heterogeneity by analysing studies which were specifically conducted on extremely low birth weight (ELBW) infants. A leave-one-out sensitivity analysis was also performed to single out the cause of heterogeneity and assess statistical robustness. When the number of studies was adequate (n≥10),10 potential publication bias would be investigated visually by contour-enhanced funnel plot12 and quantitatively by Egger’s test.13

Results

Search results and study selection

Initial search retrieved 388 articles, of which 103 were deduplicated and 97 were excluded following title and abstract screening, resulting in the full-text assessments of 6 studies. After a thorough assessment of eligibility, one study6 was excluded due to the unavailability of extractable data in the article. Therefore, a total of five studies were included in this systematic review and meta-analysis. Details of the literature search process are illustrated in figure 1.

Figure 1

Literature searching and screening process.

Study characteristics and risk of bias assessment

This review included five observational studies which were published from 2007 to 2023. All studies were conducted in a single-centre tertiary referral hospital. Two studies conducted by Manzoni et al and Zhang et al investigated the association of ROP with mode of delivery in ELBW babies. A total of 2048 babies (1030 male, 1018 female) with 1237 babies classified as ELBW were included in this review. The gestational age of the babies varies from <32 weeks to >32 weeks. There were a total of 1043 babies born through vaginal delivery and 1005 babies born through C-section. Risk of bias assessment showed a low risk of bias with satisfactory studies (online supplemental table S2). Except for the study by Abdel et al, every included study provides data on adjusted OR, although each with different controlled variables (online supplemental table S3). Table 1 shows the characteristics of the included studies. Due to an inadequate number of included studies (n≤10), publication bias would not be analysed.

Study outcomes

The incidence of ROP in C-section (n=287) was lower compared with babies born through vaginal delivery (n=469). The incidence of ROP was also found to be lower in C-section (n=223) compared with vaginal delivery (n=380) in ELBW babies. We discovered that C-section decreased the odds of having ROP infants by 46% with low heterogeneity (pooled unadjusted OR 0.54 (95% CI 0.40 to 0.73); I2=40.73%; figure 2). Subgroup analysis to test the effect of ELBW-only inclusion also yielded statistically significant pooled effects, with lower heterogeneity (online supplemental figure S1). Furthermore, our findings remained robust following sensitivity analysis (online supplemental figure S2).

Figure 2

Pooled unadjusted effects of mode of delivery on ROP incidence. ROP, retinopathy of prematurity.

Different results were found when performing the analysis on adjusted effects. Our findings did not yield significant results on pooled adjusted effects (pooled adjusted OR 0.59 (95% CI 0.28 to 1.23); I2=70.51%; figure 3). A higher heterogeneity was also observed, possibly stemming from multiple variations in the controlled variables across individual studies. Further subgroup and sensitivity analysis yielded statistically insignificant results on pooled adjusted effects (online supplemental figure S3 and S4).

Figure 3

Pooled adjusted effects of mode of delivery on ROP incidence. ROP, retinopathy of prematurity.

Discussion

Our review includes five studies covering Europe, Africa and Asia with a total of 2048 babies. The study characteristics contain variations with two studies focusing on ELBW babies with birth weight <1000 g. The cumulative incidence of ROP in all studies was 756 babies. A higher incidence of ROP is observed in babies born through vaginal delivery (n=469) compared with caesarean section (n=287). Although all studies showed a higher incidence of ROP in vaginal delivery, statistical analysis showed inconsistent results among pooled unadjusted and adjusted effect estimates.5 14–17

Three studies investigated the association of mode of delivery with incidence of ROP in preterm infants.14–16 The cumulative incidence of ROP is higher in vaginal delivery (n=89, 23.5%) compared with caesarean section (n=64, 16.1%) with inconsistent statistical significance. Study conducted in Egypt by Abdel et al showed no significant association of vaginal delivery (n=17) and caesarean section (n=16) with the incidence of ROP in univariate analysis.15 Another study by Sherief et al in Ethiopia also showed no significant association of vaginal delivery (unadjusted OR 0.96 (95% CI 0.52 to 1.75) p=0.897; adjusted OR 1.19 (95% CI 0.47 to 3.03) p=0.715) with incidence of ROP.16 Meanwhile, a study Braimah et al in Ghana showed a reduced risk of ROP with babies delivered through caesarean section (adjusted OR 0.35 (95% CI 0.14 to0.91) p=0.032). However, the study only investigated the incidence of ROP in babies born with gestational age >27 weeks due to the death of all babies born below 27 weeks of gestational age before the due date of the eye examination. Therefore, this significant result may be confounded with the sample population of preterm infants aged above 27 weeks of gestation. Nevertheless, due to the relatively large sample size, the results by Braimah et al were given more weight in the pooled analysis; thus giving a significant result (figure 2).14

Two studies investigated the association of infants born with ELBW and the incidence of ROP in vaginal delivery (n=380) compared with caesarean section (n=219). Although the result showed a higher incidence of ROP in vaginal delivery, statistical analysis showed inconsistent significance between the two studies.5 17 The study conducted by Zhang et al showed a statistically significant higher incidence of ROP in ELBW infants with vaginal delivery (40.8%) compared with caesarean section (21.4%) in univariate analysis (p<0.001). However, the result was not significant in multivariate analysis (adjusted OR 0.95 (95% CI 0.76 to 1.32) p=0.755). A possible explanation for the significant finding in univariate analysis was due to the lower mean gestational age in vaginal delivery (26.5±1.8 weeks) compared with caesarean section (28.9±2.1 weeks).17 The difference in gestational age is a significant predictor of ROP as gestational age ≤28 weeks is associated with significantly higher incidence of ROP.18 Thus, the significant result founded by Zhang et al was confounded by the mean gestational age in each group.17 On the other hand, the study conducted by Manzoni et al showed a significant result in both univariate and multivariate analysis regarding the association of mode of delivery with the incidence of ROP. A multivariate analysis was also conducted using logistic regression controlling all significant associated factors of ROP, including mean gestational age (adjusted OR 0.22 (95% CI 0.089 to 0.922) p=0.04).5

These findings are reflected by the results from pooled analyses of adjusted and unadjusted OR. Although found to yield significant protective effects on ROP in unadjusted pooled analysis (pooled unadjusted OR 0.54 (95% CI 0.40 to 0.73); I2=40.73%; figure 2), different results were found in the analysis on adjusted effects (pooled adjusted OR 0.59 (95% CI 0.28 to 1.23); I2=70.51%; figure 3). The discordance of findings possibly stems from multiple variations in the controlled variables across each study which could lead to heterogeneity. Only one variable, namely birth weight, was taken into account in multivariate analyses of all four studies. This is further supported by the significant difference in I2 values between the unadjusted pooled analysis (I2=40.73% indicating low heterogeneity) and the adjusted pooled analysis (I2=70.51% indicating moderate heterogeneity).

Previous reports have indicated poorer short-term outcomes when vaginal delivery is used for delivering ELBW neonates.19 20 Before the 2007 study by Manzoni et al, there were no published data regarding the incidence of ROP via different modes of delivery. Despite the relative lack of more conclusive data up to now, several hypotheses have been proposed to account for the possible association between the two events. It has been theorised that ELBW infants born through vaginal delivery could experience harmful pressure on the cerebral vessels of the fetus, potentially leading to ischaemia and perinatal hypoxia, both of which are associated with ROP occurrence.18 Additionally, preterm vaginal delivery is often linked to vaginal infection, which may potentially affect the neonate during the labour process.4 Lastly, preterm labour via vaginal delivery is often an unexpected and undesired process, thus limiting the possibility of preventive interventions for both the mother and the fetus.5 Each of these hypotheses alone might not be able to significantly influence the neonatal outcome. Therefore, the seemingly protective effect of the C-section is likely to be related to a number of reasons that must be considered altogether.

Strengths and limitations

This study has several limitations. Our findings yielded conflicting results between pooled adjusted and unadjusted OR. Although generally more accurate, pooled adjusted OR in our study should be cautiously interpreted due to large variations of controlled variables. Moreover, the relatively small number of included studies should also be noted. For example, we did not account for different stages of ROP or subtypes of labour (eg, vacuum-assisted vaginal delivery) due to limited published data on this subject, thus limiting the scope of our subgroup analyses. As we only include studies written in English, it is possible that some publications were not assessed due to language constraints. Nonetheless, the only excluded study during literature searches was an English-language article which lacked extractable data, suggesting that any language bias may be negligible. Despite the limitations, we think that our findings deserve consideration and should be carefully verified through further studies.

Conclusion

C-section demonstrated a lower incidence of ROP compared with vaginal delivery. While this observation is supported by pooled unadjusted effect sizes, it contrasts with the analysis based on pooled adjusted effect sizes. Nevertheless, the observed association highlights the critical importance of understanding the impact of delivery mode on neonatal ophthalmic outcomes. Given the limited number of published studies on this subject, additional research is essential to validate and build on our findings.

Data availability statement

Data sharing not applicable as no datasets generated and/or analysed for this study.

Ethics statements

Patient consent for publication

References

Supplementary material

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

Footnotes

  • X @reynardisutanto

  • Contributors VS acted as the guarantor in this study. The conceptualisation, data analysis and manuscript writing were primarily undertaken by VS, RLS and APC.

  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests None declared.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.