Introduction Laser scanning in vivo confocal microscopy (IVCM) enables non-invasive, high-resolution imaging of the cornea. In recent years, there has been a vast increase in researchers using laser scanning IVCM to image and quantify corneal nerve parameters. However, a range of methodological approaches have been adopted. The primary aim of this systematic review is to critically appraise the reported method(s) of primary research studies that have used laser scanning IVCM to quantify corneal sub-basal nerve plexus (SBNP) parameters in humans, and to examine corneal nerve parameters in healthy individuals.
Methods and analysis A systematic review of primary studies that have used laser scanning IVCM to quantify SBNP parameters in humans will be conducted. Comprehensive electronic searches will be performed in Ovid MedLine, Embase and the Cochrane Library. Two reviewers will independently assess titles and abstracts, and exclude studies not meeting the inclusion criteria. For studies judged eligible or potentially eligible, full texts will be independently assessed by two reviewers to determine eligibility. A third reviewer will resolve any discrepancies in judgement. Risk of bias will be assessed using a custom tool, covering five methodological domains: participant selection, method of image capture, method of image analysis, data reporting and other sources of bias. A systematic narrative synthesis of findings will be provided. A multilevel random-effects meta-analysis will be performed for corneal nerve parameters derived from healthy participants. This review will be reported according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses statement.
Ethics and dissemination As this review considers published data, ethical approval is not required. We foresee that this synthesis will serve as a reference for future studies, and can be used to inform best practice standards for using IVCM in clinical research. A manuscript reporting the results of the review will be published and may also be presented at scientific conferences.
- corneal and external diseases
- systematic review
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 and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/
Statistics from Altmetric.com
Strengths and limitations of this study
This will be the first systematic review to consider the level of methodological rigour applied when using laser scanning in vivo confocal microscopy (IVCM) for clinical research.
This systematic review will consider all primary research studies, irrespective of the study design, that have used laser scanning IVCM to quantify corneal nerve parameters in human participants.
This systematic review protocol is reported in accordance with the Preferred Reporting Items for Systematic review and Meta-Analyses Protocols (PRISMA-P).
We have developed a purpose-specific risk of bias tool for assessing IVCM methodological quality, which we consider will be a valuable guide for researchers using this technique, to consider potential sources of bias when developing IVCM protocols.
The review will not include unpublished studies or those published in a language other than English.
In vivo confocal microscopy (IVCM) is a non-invasive imaging method for visualising the structure of the living human cornea. IVCM provides high-resolution, morphological detail of the corneal architecture and can be applied to assess corneal parameters, in particular a range of metrics relating to corneal nerve integrity (eg, density and branching characteristics). Several types of IVCM instruments are commercially available, including tandem, slit and laser scanning devices.1 Laser scanning IVCM, which uses a red wavelength diode laser source that poses no ocular safety hazard,2 is currently considered the gold standard device for clinical research. This technology provides a greater depth of focus, enhanced contrast and improved resolution compared with the alternative devices.1
While early studies using laser scanning IVCM to examine corneal health were mostly qualitative in nature (eg, for diagnosing corneal infection), the technique is now used to determine a range of quantitative clinical measures (eg, corneal nerve density). Accurate quantification of corneal nerve parameters is important clinically for monitoring the potential effect of therapeutic interventions on corneal health, detecting corneal neuropathy and acting as a surrogate biomarker for early-stage diabetic peripheral neuropathy.3 4 Although a general method for examining the cornea and analysing corneal nerve parameters using laser scanning IVCM has been described,5 there is currently no gold standard protocol for using laser scanning IVCM for corneal nerve analysis available in the literature. As a result, a range of different approaches have been adopted.3 4 6 7
Various factors, in particular the methods used for image capture and analysis, may introduce bias and thereby affect the accuracy of quantitative measures, when using IVCM to investigate corneal nerve parameters.8 For example, as corneal nerve density varies with eccentricity (ie, greater in the central vs peripheral cornea),9 consideration should be given to the region of cornea imaged. In addition, factors such as the microscope field of view, depth of corneal imaging, image quality and postcapture image enhancements may influence the visibility and/or clarity of nerves within the image field, thereby potentially impacting on quantitative measurements.8 The number of images analysed, per individual, also affects the confidence of quantitative estimates; it has been shown that at least eight images, with less than 20% image overlap between each image, should be analysed to obtain reliable estimates of corneal nerve density.10 To avoid potential performance biases in studies involving different participant groups and/or clinical intervention studies, the confocal microscope operator and outcome assessor should be masked to the participant’s group allocation. In addition, the method for quantifying the sub-basal nerve parameters should be fully described by researchers, with preference given to the use of a validated, fully automated processing method (eg, ACCMetrics11), to circumvent the potential bias induced by subjective judgement.
There has not been any previous research undertaken to consider the level of methodological rigour applied when using laser scanning IVCM for clinical research. As researchers who are experienced with performing the technique, we have developed a purpose-specific risk of bias tool covering five key methodological domains that we consider important for minimising bias when using laser scanning IVCM. The five domains are participant selection, method of image capture, method of image analysis, data reporting and other sources of bias (eg, industry funding). We foresee the use of this purpose-specific risk of bias tool as a valuable guide for researchers, to consider potential sources of bias when developing their IVCM protocols. In this respect, the present paper has the capacity to contribute to significantly improving the quality of future research in the field.
The major aim of this systematic review is to critically appraise (ie, assess the risk of bias in) the reported method(s) of primary research studies that have used laser scanning IVCM to quantify corneal sub-basal nerve parameters in human participants. We will also determine key differences in methodology between studies and identify the specific methodological domains that are least well performed and/or reported (ie, are judged as having the highest risk of bias) in the literature, as a basis for informing laser scanning IVCM methods and their robust reporting, in future clinical studies. We predict that there will be considerable variation in the image-capturing methodologies used by different investigators and between the studies, which may lead to potential biases and affect the reliability of reported data. For example, studies may have used an insufficient number of corneal images as a representative measure to quantify nerve density, potentially leading to sampling bias. Finally, a meta-analysis will be conducted on studies assessing corneal nerve fibre parameters in healthy individuals. As a result, this will help to establish a more precise estimate of corneal nerve parameters for future research to use as a reference for identifying corneal nerve pathology.
The primary objective of this systematic review is to critically appraise (ie, assess the risk of bias in) the reported method(s) of primary research studies that have used laser scanning IVCM to quantify corneal sub-basal nerve parameters in human participants.
The secondary objectives are:
to identify the methodological domains that are least well performed and/or reported (ie, are judged as having the highest risk of bias) in the included studies, as a basis for informing laser scanning IVCM methods and their robust reporting, in future clinical studies. As shown in table 1, the five main methodological domains that will be assessed are participant selection, method of image capture, method of image analysis, data reporting and other sources of bias;
to determine normative values for corneal sub-basal nerve plexus parameters by pooling the estimates from available studies.
Methods and analysis
The proposed systematic review and meta-analysis will be undertaken using the approach recommended by the Preferred Reporting Items for Systematic Reviews and Meta Analyses (PRISMA) statement.12
All studies published in English will be included, from the date of database inception until 17 May 2017. In cases where multiple publications of the same data exist, the study reporting on the largest number of human participants will be included.
Also, studies will be selected according to the following criteria:
We will include all primary, empirical research studies that have used a laser scanning confocal microscope to perform corneal confocal microscopy on at least one human participant, where corneal sub-basal nerve plexus parameters were quantified. We will include studies from across the spectrum of clinical research questions defined by the National Health and Medical Research Council13 ,13 (eg, intervention, diagnostic test accuracy, aetiology, prognosis and screening intervention) and study designs (eg, randomised controlled trial (RCT), pseudo-RCT, non-RCT, cohort, case-control, cross-sectional, interrupted time series, case series, case study), to enable the comparison of methodological quality across study types.
Studies reporting only on aspects of corneal architecture other than sub-basal nerve parameters (eg, epithelial thickness, endothelial cell count/morphology, corneal haze, and so on) will be excluded. We will exclude review papers (including systematic reviews), conference abstracts and studies reporting methods for analysing laser scanning IVCM images, where human participants were not recruited. We will also exclude studies that have used alternative types of confocal microscopes for image capture (eg, tandem scanning and slit scanning), as the type of confocal microscope affects the quantitation of corneal sub-basal nerve parameters.1 14
We will include all studies that report corneal sub-basal nerve plexus findings for at least one human participant. There will be no restriction on participant health status for the systematic review (although restrictions will apply for the meta-analysis, which will only include data from healthy adults); thus, included studies may involve healthy individuals, as well as those with ocular and/or systemic conditions.
A comprehensive search to identify all relevant studies will be undertaken in the following electronic databases: Ovid MEDLINE(R) (Epub Ahead of Print, In-Process & Other Non-Indexed Citations, Ovid MEDLINE(R) Daily and Ovid MEDLINE(R) 1946 to May 2017), Ovid EMBASE (Embase Classic+Embase, 1947 to May 2017) and the Cochrane Library.
To ensure literature saturation, we will scan the reference lists of included studies, or relevant reviews, identified by our search. We will also search the authors’ personal bibliographic reference files to ensure that all relevant studies are captured. We will also circulate a list of the included articles to our review team.
The search strategies are provided in online supplementary material.
Supplementary file 1
The systematic search will be carried out by the review team, using the previously defined search strategies, and guided by Items 9 and 10 of the Preferred Reporting Items for Systematic review and Meta-Analysis Protocols (PRISMA-P) statement.15 After performing the search strategies separately in each electronic database, the researchers will import the results from each search into an EndNote library. As the same article may be located in more than one database, duplicate entries will be identified and removed.
We will use Covidence systematic review software,16 an online program that facilitates collaboration between reviewers for systematic reviews, for the study screening process. Two reviewers (MEHDS and ACZ) will independently assess the titles and abstracts of all unique studies, identified from the electronic search strategies, and exclude those that do not meet the inclusion criteria. For studies judged to be eligible or potentially eligible for inclusion, the full-text articles will be sourced and independently assessed by the two reviewers, to clarify their eligibility for inclusion. Any discrepancies in classification that arise during this process will be resolved by consensus between the two reviewers and a third reviewer. For studies that progress to the full-text screening stage, we will record the reason that studies were excluded.
If there are cases where it is unclear whether the inclusion criterion are met, we will attempt to contact the study corresponding author for clarification; if no response is received within four weeks of the request, or the requested information is not provided, the information within the full-text article will be used to decide on the eligibility of the study.
A diagram will be created to report the flow of studies through the systematic review.
Relevant data, from eligible studies, will be independently extracted by two reviewers in Covidence, using a standardised data extraction form. Extracted data will be summarised in tables. If any data extraction discrepancies arise, these will be resolved by discussion and consensus among the review team.
Extracted data from each included study will include
article details: year of publication, journal of publication;
study details: type of research question (ie, intervention, diagnostic-test accuracy, aetiology, prognosis, screening intervention), setting, location, study design (eg, RCT, pseudo-RCT, non-RCT, cohort, case-control, cross-sectional, interrupted time series, case series, case study), number of participants, health status of the participant population(s) (eg, healthy, diabetes, and so on), participant characteristics (age, gender), population eligibility criteria (inclusion and exclusion criteria);
methodological details: unit of analysis (one eye (right or left), both eyes or average of both eyes, as applicable), corneal sub-basal nerve parameters assessed (see ‘Outcomes’ section for further details), IVCM image capture field of view (ie, 300 or 400 µm), IVCM mode of image capture (eg, volume, sequence or section scan), whether a representative IVCM sub-basal nerve plexus image is provided in the paper (dichotomous);
other details: source of funding statement (dichotomous: present or absent), actual source of funding (eg, industry or Government funding), conflict of interest statement (dichotomous: present or absent), conflict of interest type (eg, employee of company conducting study);
quantitative measures: data (ie, mean (SD) or median (IQR)) for the following four key central, corneal sub-basal nerve plexus parameters: CNFL, CNFD, CNBD and CTBD, as defined in the ‘Outcomes’ section. Where data are provided for both eyes, we will also extract the correlation coefficient. If longitudinal data are reported, we will use baseline data in our analyses. As all of the sub-basal nerve plexus parameters are continuous outcomes, we will extract data on the means and SD for each parameter, or similar measures of central tendency and variability.
The primary outcome will be the methodological quality of included research studies that have used laser scanning IVCM to quantify corneal sub-basal nerve parameters in human participants.
The secondary outcomes are as follows:
Identification of the methodological domains that are least well performed and/or reported (ie, are judged as having the highest risk of bias) in the included studies, as a basis for informing laser scanning IVCM methods and their robust reporting, in future clinical studies.
Meta-analysis of mean normative values (ie, from healthy individuals) for corneal sub-basal nerve plexus parameters, quantified from the central cornea (as defined by the study authors), and using these definitions for the analysis.
Corneal nerve fibre length (CNFL) defined as the total length of all nerve fibres in the image capture frame (mm/mm2).17 18 If an alternative definition is used, such as limiting the quantification of fibres to those of a certain minimum length, these data will be excluded.
Corneal nerve fibre density (CNFD) defined as the total number of main fibres divided by the area of the image frame (fibres/mm2)18.
Corneal nerve branch density (CNBD) defined as the total number of main nerve branches, being branches that stem from a nerve fibre, divided by the area of the image frame (branches on main fibre/mm2)18.
Corneal nerve total branch density (CTBD) defined as the total number of branches within the area of the image frame (total branches/mm2).
Assessment of risk of bias in included studies
To facilitate the assessment of methodological quality in each of the included studies, as per the objective of this review, we developed a purpose-specific, 18-item risk of bias tool (table 1) to assess internal validity, encompassing five main domains:
participant selection (including selection bias)
method of image capture (including performance bias and sampling bias)
method of image analysis (including performance bias)
data reporting (selective reporting of outcomes and attrition bias)
other sources of bias (funding source and conflicts of interest).
The risk of bias tool was developed by the review team (MEHDS, ACZ, HRC, LED), who possess expertise in using IVCM for corneal nerve analyses, for this review and was framed using the Cochrane Assessing Risk of Bias in included studies (Chapter 8 in the Cochrane Handbook of Systematic Reviews of Interventions).19
Two reviewers will judge the risk of each type of bias (18 items in total) in each included study as either (1) low risk, (2) unclear risk (due to either lack of information or uncertainty about the potential for bias) or (3) high risk. Review authors will resolve any disagreements in bias assessment by consensus with a third reviewer. Reviewers will not be masked to the journal of publication or the study author name when undertaking the risk of bias assessment.
Wherever possible, we will justify each risk of bias assessment with direct quotations from the study. If there are cases where further information is considered necessary to determine the risk of bias in a particular domain, we will attempt to contact the study corresponding author for this information. If no response is received within four weeks of the request, or the requested information is not provided, the information within the full-text article will be used to inform the risk of bias assessment.
For outcomes related to methodological quality, a systematic narrative synthesis will be provided, with relevant information summarised in text and tables.
If there are at least three relevant studies, we will undertake meta-analyses of the quantitative data for the specified corneal sub-basal nerve plexus parameters (ie, mean (SD)). Data from male and female participants will be pooled, as studies have shown that gender has no significant effect on corneal sub-basal nerve plexus parameters.20 21
We will convert non-parametric data to means (SD), using an established approach.22 We will fit a multilevel random-effects model to pool the estimates. The multilevel model will take into account the correlation between estimates from the same study that are presented separately for each sex and/or estimates presented separately for the left and right eyes. Next, we will fit a meta-regression model to assess how much of the between-study variation is explained by the following characteristics: (1) participant age (as this factor is potentially important relation to sub-basal nerve plexus parameters),23 and (2) study design (eg, RCT, cohort (including pseudo-RCT and non-RCT) and other (including cross-sectional, case series/study)).
Statistical analyses will be carried out using the metafor package in R.24 25
As there are no limitations on the potential study designs eligible for inclusion in this review, we expect that we will not be able to compare the outcomes reported in published reports with study protocols, unless the included study is a RCT, to assess for selective outcome reporting or selective analysis reporting. Furthermore, as our meta-analysis is being undertaken to determine values for normative parameters (ie, corneal sub-basal nerve plexus parameters), rather than the effect of an intervention, we do not expect meta-biases (such as publication bias, delayed publication, and so on) to be a significant factor for this analysis.
Provided there are a sufficient number of studies included in the review, sensitivity analyses will be performed for the CNFL outcome, to assess for the effect of excluding studies that (1) were appraised as having a high risk of bias in the domains of image selection—number and sampling, or method for quantifying sub-basal nerve parameters; (2) included contact lens wearers (ie, contact lens wear was not listed as an eligibility exclusion criterion); (3) were lower order levels of evidence (eg, case reports, case series, interrupted time series); and (4) are from the same corresponding/senior author, in the event that at least 50% of the included papers are from the research laboratory of the same corresponding author.
Confidence in cumulative evidence
If appropriate, we will present a ‘Summary of Findings’ table for the quantitative outcomes. In this case, the quality and strength of the body of evidence will be assessed using an approach based on the Grading of Recommendations Assessment, Development and Evaluation.26
In recent years, an increasing number of research studies have adopted non-invasive, laser scanning IVCM to quantify corneal sub-basal nerve plexus parameters. However, there has not been any research to formally consider the quality of the methods used in these investigations. This systematic review will provide insight into the quality of the methods reported in clinical studies using laser scanning IVCM to quantify corneal nerve parameters. The review will also identify specific methodological domains that are least well performed and/or reported (ie, are judged as having the highest risk of bias) in the literature, as a basis for informing laser scanning IVCM methods and their robust reporting, in future clinical studies. Furthermore, by researchers considering the elements of the purpose-specific risk of bias tool as a guide when developing their IVCM protocols, this review has the capacity to significantly improve the quality of future research in the field. By undertaking a meta-analysis, we will also determine mean normative values (ie, from healthy individuals) for central corneal sub-basal nerve plexus parameters. These data will be of significant value for future studies, as reference normative values, building on a previous pooled analysis of data derived from multiple laser scanning IVCM testing centres.27
We acknowledge the assistance of Tania Celeste, Biosciences Librarian at the University of Melbourne, who assisted with developing the search strategies for this review.
Contributors MEHDS and LED drafted the initial protocol, with input, discussion and editing from ACZ and HRC. AK contributed to the ‘Data synthesis’, ‘Meta-bias(es)’, and ‘Sensitivity analyses’ sections and the protocol components for undertaking the meta-analysis. All authors provided final approval of the paper.
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.
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.