Study design
This systematic review will include retrospective and prospective case series’, cross-sectional studies, retrospective and prospective cohorts, and randomized clinical trials. It will be conducted according to the PRISMA-P (Preferred Reporting Items for Systematic Review and Meta-Analysis Protocol) guidelines.14,15 This systematic review has also been submitted for registration on PROSPERO.16
Inclusion Criteria
The proposed inclusion criteria for this systematic review are as follows:
Population: All studies, with participants diagnosed with microphthalmos or nanophthalmos in one or both eyes, will be included if they have: i. more than 4 cases, and ii. defined microphthalmos/nanophthalmos as an axial length of < 21mm or a high lens/eye volume ratio. Nanophthalmos may have an additional diagnostic criterion of posterior wall thickness greater than 1.7mm.
Condition: The prevalence of the following complications will be assessed: High hyperopia (Spherical equivalent > 3D),17 angle closure glaucoma, uveal effusion syndrome, retinal detachment, and chorioretinal folds.1,2 Spontaneous ocular complications are defined as those ocular complications that are diagnosed prior to any surgical procedure performed. If a surgical procedure occurred prior to the diagnosis, the study would fall under post-surgical complications.
Context: The studies will probably be hospital-based cohorts or case series. If population-based studies are found these will be included and analysed separately.
Since this is a prevalence study, no comparison group is necessary. However, there may be randomised studies using an intervention and comparing post-surgical complications. These complications will then be analysed according to the intervention and the comparison arms used. The comparison arm may an alternative intervention or placebo.
Exclusion criteria
Studies that will be excluded are: i. those that have not adequately defined the criteria for the diagnosis of nanophthalmos or microphthalmos, ii. studies that have less than five participants, iii. studies with criteria not defined above, and deemed unsuitable, and iv. studies in languages other than English with no published translation.
Databases and information sources
Four databases will be searched for relevant studies: PubMed, EMBASE, Web of Science and Scopus. Registered clinical trials that are published will also be sought on ClinicalTrials.gov. Reviews will be checked on the Cochrane database. The reference lists of these reviews and any reviews found on the other databases will be checked for additional articles. Where necessary, authors will be contacted for clarification by one reviewer (NA).
Search Strategy
The search strategy, using the search terms in Appendix A below, has been developed using the CoCoPop method above.18 All the information and data will be collated and entered onto Microsoft Excel (Microsoft Corporation). The data collected will be assessed on the first 5 articles and amended if required.
Study selection
The study selection process will be conducted using the PRISMA guidelines and flow diagram.14 After exclusion of duplicate studies using the Zotero™ citation manager, and after being rechecked manually, all titles and abstracts will be reviewed independently by two reviewers (NA and SI) according to the prespecified inclusion and exclusion criteria (Appendix B). Any disagreements will be resolved by a third reviewer (HDA). Systematic and narrative reviews, animal studies, editorials and letters will be excluded. The reference lists of review papers will be screened for studies that meet inclusion criteria. The full texts of eligible studies / papers will be examined for inclusion into the systematic review and meta-analysis (Appendix B). When data from the same cohort are reported in separate manuscripts, the study reporting the largest sample fulfilling our eligibility criteria will be selected. If there are doubts regarding these datasets, the corresponding authors will be contacted for clarification.
Studies that are included, and decisions made, will be recorded on Microsoft Excel™ (Microsoft Corporation).
Risk of bias assessment/Quality assessment
Studies assessing the prevalence of spontaneous complications and post-operative complications in cohorts and case series’ will be assessed using a modified version of the Joanna Briggs Institute (JBI) critical appraisal tool.18,19 If there are randomized clinical trials, they will be assessed for quality using the JBI assessment tool for randomised clinical trials20 and the overall evidence will be assessed using the GRADE approach.21
Studies comparing the effect of surgical interventions on the prevalence of post-operative complications will be assessed for publication bias using a funnel plot.
Data extraction
Data will be extracted by one reviewer (NA), checked by another reviewer (SI), and populated onto Microsoft Excel™ (Microsoft Corporation). Disagreements will be resolved by a third reviewer (HDA). The following data will be extracted: Year of study, study design (i.e., Cohort, cross-sectional etc.), total number of participants, axial length, proportion of males to females, mean/median age at presentation, number of participants with one or more complications (High hyperopia [Spherical equivalent > 3D], angle closure glaucoma, uveal effusion syndrome, retinal detachment, and chorioretinal folds), posterior wall thickness and lens/eye volume ratio.
Data Analysis
Data will be analysed in Stata 16.1 (STATACorp LLC, College station, Texas). Due to the difference in definitions and diagnosis of microphthalmos and nanophthalmos a random effects models will be used throughout. The presence of spontaneous complications will be analysed using the “metaprop” command in Stata. The Freeman-Tukey double arcsine transformation will be performed to normalise outcomes before pooling the prevalence. Study specific 95% confidence intervals will be generated using the exact method. The I2 statistic will be used to check for overall, intergroup, and intragroup heterogeneity. Forest plots will then be generated from the data.
For post-operative complications where a comparison group is used the effect size will be compared using the “meta esize” command. A random effects model will be used to present the overall effect size as an odd ratio with 95% confidence intervals. Forest plots will then be generated from the data. If possible, Egger’s test and funnel plots will be used to test for publication bias.
Subgroups
Subgroups will be analysed according to axial length to see if the proportion of complications increases/decreases with decreasing axial length in both the spontaneous complication group and post-surgical complication group. Subgroups will also be analysed according to subtypes of microphthalmos (anterior, posterior, simple and complex) if the data permits this. If randomised trials are found using interventions to decrease the post-operative complications these will be meta-analysed using odds ratios to assess the odds of getting those complications post-operatively.