|Year : 2022 | Volume
| Issue : 1 | Page : 79-80
Department of Ophthalmology, Kozhikode Medical College, Kozhikode, Kerala, India
|Date of Submission||20-Jan-2022|
|Date of Decision||24-Jan-2022|
|Date of Acceptance||26-Jan-2022|
|Date of Web Publication||21-Apr-2022|
Dr. Aishwarya Sadanand
Department of Ophthalmology, Kozhikode Medical College, Kozhikode, Kerala
Source of Support: None, Conflict of Interest: None
|How to cite this article:|
Sadanand A. Journal Review. Kerala J Ophthalmol 2022;34:79-80
| Cataract Surgery in Patients With Corneal Opacities|| |
Citation: Ho YJ, Sun CC, Chen HC. Cataract surgery in patients with corneal opacities. BMC Ophthalmol 2018 Apr 23;18(1):106. DOI: 10.1186/s12886-018-0765-7. PMID: 29685130; PMCID: PMC5913879.
This study concluded that phacoemulsification (PE) and intraocular lens (IOL) implantation in selected cases of coexisting cataracts and corneal opacities are safe. The authors concluded that PE and IOL could provide suboptimal but long-term vision in cases where penetrating keratoplasty is not possible or cases at high risk of graft failure. They concluded that anterior-segment optical coherence tomography (ASOCT) is a simple tool to predict visual outcomes after cataract surgery in corneas with opacity.
The study title is not specific enough and does not provide us with information on the scope or extent of the study. The authors state that the primary aim was to report the visual outcomes and complications in patients undergoing cataract surgery in eyes with coexisting corneal opacities, while the secondary aim was to evaluate whether ASOCT can be used as an alternative objective modality to predict visual prognosis. The authors hypothesized that cataract surgery (PE) with IOL, as an isolated procedure, can provide timely visual rehabilitation in eyes with corneal opacities. They also state that this is the first report to evaluate and compare nontransparent cornea with densitometry and new OCT grading methods. The authors establish that the use of ASOCT assessing eyes with corneal opacity and cataracts is the gap in the literature that they aim to address.
An appropriate study design is essential to answer the research question. The authors mention that they used a retrospective design and a descriptive study. The authors, however, do not provide additional details on the design and state they did a review of the medical case records. The term descriptive is a broad term that can cover several study designs, while the term retrospective is an element of time that can factor into the study design. However, both terms, retrospective and descriptive, do not provide sufficient detail regarding the study design used. Retrospective designs may be simpler to do, less time-consuming, and less costly, especially when cases are fewer but carry a higher risk of confounding and bias compared to a prospective study. Lack of adequate or complete documentation and lack of standardization of measures are other limitations of a retrospective study. However, pragmatically, we must also consider the distribution of the condition of interest and the caseload, and a retrospective may be more feasible than a prospective study where there are constraints of time and caseload.
In this study, the group of interest was people with coexisting cataracts and corneal opacity where Penetrating Kerato Plasty (PKP) cannot be done. This group can be further divided into a group of patients where cataracts can be done and a group where cataract surgery cannot be done. When we compare two groups, to be sure that whatever result we got did not occur by chance, the two groups must have a reasonable number of the study population. The authors mention that they have studied the records of 23 eyes in 19 patients based on the inclusion and exclusion criteria. Each participant underwent a detailed assessment that included collection of demographic details, visual acuity assessments, Lens Opacities Classification System III (LOCS III) grading of the lens, fundus examinations and B scan ultrasonography, grading of the corneal haze by two observers using the corneal haze in accordance with modified VISX protocol, manual demarcation of the boundary of corneal opacification and measurement of size using Image J software (National Institutes of Health, Maryland), estimation of the percentage of opacity occupying central corneal region, assessment of the density of corneal opacity using ASOCT, and measurement of the backscatter of the cornea by cornea densitometry using rotating Scheimpflug camera. The authors do not mention the number of observers who performed the various assessments, the standardization procedures and agreement between observers, and how conflicts in measurements were resolved. These assessments are not usually performed as part of a routine clinical examination, and hence, there is the possibility of a selection bias based on who received these assessments.
The authors mention that all participants underwent a PE with IOL under topical anesthesia with the same techniques but do not mention the number of operating surgeons. Postoperatively, patients were examined on postoperative days 1, 3, and 7, then 1 month, and every 3–4 months thereafter. A complete ophthalmic examination including Uncorrected visual acuity (UCVA), Best corrected visual acuity (BCVA), keratometry, tonometry, slit-lamp examination, endothelial cell density, central corneal thickness, and fundus examination was performed at each follow-up visit.
There are several aspects to consider regarding the methods. These include the sample size, the selection of the study subjects, the assessment of the variables of interest and lack of information on standardization, agreement, and completeness of data, and the possible variability in outcomes based on the skills of the operating surgeon. The authors mention a retrospective, case record-based review as a design but ideally should have considered a pre–postinterventional study design. A pre–postinterventional study design is more suitable as the authors look at changes postintervention in a series of patients.
The authors have assessed the normality of data distribution and chosen Spearman's correlation analysis to report on associations and linear regression to report on the fit of the data. There are several limitations in the presentation of the data. The mean age is provided in the demographic details, but the range of ages or median age is not provided. The authors have included trauma with traumatic cataracts and herpes as causes for corneal opacification. Recurrence of herpes can impact visual outcomes and must be considered as poor visual outcomes can be attributed to an underlying cause of the corneal opacity rather than lens removal. The authors present that nearly 30%–40% had opacities that did not involve the visual axis. The percentage area of opacity involving the central ellipse, density of corneal opacity, and severity was not available for all cases. In [Table 2] of the article, the preoperative endothelial cell density is not available for 16 of 23 eyes, and 14 of 23 eyes did not have preoperative corneal thickness assessments. The authors have included two eyes with PKP in the study even though the Inclusion Criteria was to include eyes where it was not possible to do PKP. The authors also included eyes amblyopia and pale disc that affects the final visual outcome.
These results show several limitations. The number of cases is few in the first instance (23 eyes of 19 patients), and a significant proportion of these 23 eyes had incomplete preoperative data. The authors mention the results of Spearman's correlation analysis to report on the association of corneal opacity with visual outcomes.
The manuscript has several limitations that preclude the translation of these results to clinical practice. (1) There is an inadequate description of the study design to assess if that is appropriate to answer the study question. (2) The sample size is inadequate for an association study, and they should have described this as a case series instead. A case series is useful to generate a hypothesis but does not provide conclusive evidence. (3) The small sample size and lack of complete data is a major limitation to explore for associations. They should have instead used the case series approach to generate a hypothesis and design a prospective study. (4) The statistical tests are not appropriate. A correlation analysis is not ideal to measure associations, and a regression model is preferred in such instances. However, it is preferable to defer formal statistical testing of hypothesis in this study considering the small sample size and the incomplete data and the higher chance for erroneous results. (5) Figure C does not show a monotonic relationship, so a correlation analysis is not appropriate.
My interpretation of this article is that it addresses a clinically relevant problem. However, the results of this study cannot be translated to clinical practice due to the limitations in the methodology used to answer the question. The conclusions of the authors are not supported by their data or research methods. We need prospective studies with adequate sample size and appropriate sample size to obtain valid and reliable results. This may have to be a multicenter study considering the limitations of obtaining an adequate sample size from a single center. The study comes under the scope of an interventional study, and a randomized clinical trial is gold standard. Alternate designs such as a pre–postinterventional design and a cohort study design may be considered as a pragmatic alternate.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.