Impact of Ophthalmic Viscosurgical Devices in Cataract Surgery

Background Ophthalmic viscoelastic devices (OVDs) used during small-incision cataract surgery have numerous advantages. However, OVDs have longer retention time in an eye after surgery resulting in intraocular pressure (IOP) spikes. The purpose of this study is to analyze and quantify the effect of various OVDs on both IOP and best corrected visual acuity (BCVA) by systematically reviewing the literature and performing meta-analysis. Methods Numerous databases from January 1, 1985, to present were systematically searched. Thirty-six (3893 subjects) of 3313 studies identified were included for analysis. Standardized mean difference (SMD) was computed, and meta-analysis was performed. Results A total of 3313 records were retrieved including 1114 from database search and 2199 from grey literature search. Significant increase in postoperative IOP in 1-day follow-up with Healon (SMD = 0.37, CI: [0.07, 0.67]), Viscoat (SMD = 0.29, CI: [0.13, 0.45]), Provisc (SMD = 0.46, CI: [0.17, 0.76]), and Soft Shell (SMD = 0.58, CI: [0.30, 0.86]) was computed. On the other hand, results implied a nonsignificant increase in postoperative IOP with Healon GV (SMD = 0.07, CI: [−0.28, 0.41]), Healon5 (SMD = 0.15, CI: [−0.33, 0.64]), 2% HPMC (SMD = 0.32, CI: [−0.0, 0.64]), and OcuCoat (SMD = 0.26, CI: [−0.37, 0.9]). Additionally, a nonsignificant reduction in postoperative IOP was inferred with Viscoat + Provisc (SMD = −0.28, CI: [−2.23, 1.68]). Conclusion Improvement in IOP was shown with Viscoat + Provisc. Additionally, IOP nonsignificant upsurge was observed with Healon GV, Healon5, 2% HPMC, and OcuCoat compared to significant upsurge with Healon, Viscoat, and Soft Shell.


Introduction
A small-incision cataract surgery is the preferred method of cataract surgery by most surgeons. Ophthalmic viscoelastic devices (OVDs) have numerous advantages during smallincision cataract surgery. OVDs protect corneal endothelium against fluid turbulence, oxygen free radicals released during ultrasound [1], contact with surgical instruments, air bubbles, and lens fragmentation [2]; facilitate surgical procedure; reduce the risk of collateral damage to delicate intraocular tissues [3], maintain the anterior chamber space and stability to avoid capsular rupture [3]; and provide clarity to avoid complications. ese properties may vary based on physical, chemical, and rheological characteristics of OVDs [4]. An ideal OVD [5]-which does not exist yet-would be easy to inject into the eye, would maintain anterior chamber [6], would not impair vision by trapping air bubbles, would not increase intraocular pressure (IOP), and would be easy to remove from the eye after the surgery.
However, a major disadvantage of OVDs is longer retention time in an eye after cataract surgery resulting in IOP spikes [7][8][9][10]. OVDs remain in the eye resulting in mechanical obstruction of the trabecular meshwork, impeding outflow and causing IOP spikes within 24 hours after surgery which have become a concern [10], specifically for glaucoma patients. In glaucoma patients, IOP spikes may cause significant damage to the optic disc [9].
In the literature, numerous prospective randomized control trials (RCTs) have been conducted to compare safety, efficacy, and performance of various OVDs used during routine small-incision cataract surgeries and intraocular lens (IOL) implantation. Most studies compared preoperative characteristics including best corrected visual acuity (BCVA), central corneal thickness (CCT), endothelial cell count (ECC), and IOP with postoperative characteristics in various OVDs at several time intervals. Few studies suggested that super viscous and cohesive OVDs take longer time to remove from a normal eye leading to IOP elevation for a greater time period [3,4,11,12]. On the other hand, few studies state that longer aspiration time is required to remove dispersive OVDs [1] compared to cohesive OVDs [1,10,13].
In 2009, a single study [14] evaluated the protective effect of different viscoadaptive, super viscosity cohesive, viscous cohesive, medium viscosity dispersive, very low viscosity dispersive, and Soft Shell devices on ECC during cataract surgery by conducting a meta-analysis. Our research expands by evaluating the effect of various OVDs on IOP and BCVA during small-incision cataract surgery by performing a systematic review and meta-analysis. For systematic review, published as well as unpublished (grey) literature is systematically searched and data is synthesized from the included articles to compare various OVDs in order to investigate the answer to two key questions: (1) Which OVD causes the greatest risk of an IOP spike? (2) What is the specific postoperative time point at which an IOP spike is most likely to occur?

Search Strategy.
In this work, we adhered to the Preferred Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [15]. Bibliographic [15]; for more information, visit http://www.prisma-statement.org).  Reuters), and the Cochrane Library (Wiley) till December 2018, were searched. Database specific subject headings and key words for "ophthalmic viscoelastic device" or "ophthalmic viscosurgical device," and "increased IOP" or "endothelial cell loss" were employed in the search strategy. e searches were modified to accommodate syntax of each database (S2). OVID AutoAlerts were set up to send monthly updates with any new literature.   Journal of Ophthalmology 5

Selection Criteria.
Randomized controlled trials, published in English language, discussing unilateral and bilateral cataract surgery on human subjects above the age of 19 and older were included. No restriction was placed based on study location. Figure 1 summarizes the PRISMA flow diagram. In total, 1114 records were retrieved from multiple databases including MEDLINE (213), EMBASE (639), ISI Web of Science (195) Cochrane Library (5), and CINAHL (62). An additional 2,199 records were identified through grey literature searches. EPPI-Reviewer 4 gateway (by EPPI-Centre, Social Science Research Unit, the Institute of Education, University of London, UK) was used to conduct the systematic review. All identified records were imported to EPPI-Reviewer 4 to remove duplicates. After removing duplicates (217 records), 3,096 records were included for the three-level screening process.

Screening.
Tittle (Level 1) screening involved reviewing titles, while abstract (Level 2) screening involved reviewing abstracts, and full-text (Level 3) screening (S3) involved fulltext reviews of included articles, independently by two reviewers (AF and YS). At each level, agreement and disagreement between the two reviewers were assessed by Cohen's kappa (κ) coefficient. Differences between the reviewers were discussed and resolved by consensus. In cases where consensus was not achieved, a third reviewer (MM) intervened to provide a decision.

Data Extraction.
Data was extracted from the 36 eligible articles using a data extraction form. Data included study design, location, total patients enrolled, total patients enrolled in and completed the study, number of females, patient demographic characteristics, follow-ups, and baseline and postoperative characteristics including IOP, BCVA, CCT, and ECC. For missing data, various pieces of available information (such as the range, p value, and confidence interval) were utilized and converted to the common effect measure-SD. Quality of each included article was checked using modified Downs and Black checklist [16].   Standardized mean difference (SMD) was computed as the effect size. e extracted mean and standard error of the IOP at baseline and end point were used to compute the mean IOP reduction (IOPR), percentage of IOP reduction (IOPR%), and standard deviation percentage of IOP reduction (SD IOPR% ) [17].
SMD was calculated as the treatment effect since it is a mean difference standardized across all studies. Weights were assigned to each SMD according to the inverse of its variance, and then average was computed. SMD for each study was then aggregated using the fixed or random-effect model based on the presence of heterogeneity to estimate the summary effect.
To test heterogeneity, I 2 statistics, Z-value, and χ 2 statistics were computed. Additionally, a high Z-value, a low p value (<0.01), and a large χ 2 value imply significant heterogeneity and, therefore, a random-effects model using DerSimonian and Laird methods was computed. Funnel plots were generated to check publication bias.
Subgroup analysis was conducted to ascertain the influence of OVD following the cataract surgery on    Journal of Ophthalmology postoperative IOP and BCVA. Causes of heterogeneity were also explored.

Search Results.
Total of 36 RCTs were eligible for metaanalysis (Table 1). From these 36, 28 RCTs had data on IOP (S4), 13 studies had data on central corneal thickness (CCT), 16 studies had data on endothelial cell count (ECC), 9 studies reported data on best corrected visual acuity (BCVA) (Table 1), 10 studies listed wash-out times for OVDs (Table 2), and 9 studies reported the adverse events that occurred postoperatively (Table 3). In the end, a total of 36 studies (3893 subjects) were included for qualitative synthesis and 28 studies (2613 subjects) for quantitative synthesis (Figure 1). Figures 2 and 3 show the funnel plots (S4) for studies reporting preoperative and postoperative IOP and BCVA, respectively, for various OVDs. Figure 2 shows studies scattered from top to bottom right of the plot. erefore, publication bias could not be concluded. Partially, the reason was difficulty in interpretation of funnel plot due to high heterogeneity and small effect sizes.  Figure 5). More studies evaluating Viscoat + Healon GV and Viscoat + Provisc are required. Figure 6 shows a forest plot for subgroup analysis by OVDs for SMD of pre-and postoperative IOP in 6

Discussion
A systematic review was conducted to evaluate the effect of various OVDs in patients with cataract. A total of 36 RCTs (3893 subjects) were included for qualitative synthesis and 28 RCTs (2613 subjects) for quantitative synthesis. Percentage of reduction in IOP and standardized mean difference (SMD) in IOP as well as BCVA were computed. Meta-analysis results showed significant increase in postoperative IOP up to 5-hour follow-up irrespective of OVDs used. erefore, removal of OVD is essential to avoid the IOP spikes. Results suggested a nonsignificant increase in postoperative IOP in 6-hour follow-up with Healon compared to a significant increase with Healon GV, Healon5, Viscoat, 2% HPMC, and OcuCoat. Additionally, postoperative IOP significantly increases with Healon, Viscoat, and Soft Shell compared to a nonsignificant increase with Healon GV, Healon5, 2% HPMC, and OcuCoat, compared to a nonsignificant reduction with Viscoat + Provisc in 1-day follow-up. Postoperative IOP nonsignificantly decreases with Healon, Viscoat, 2% HPMC, and OcuCoat compared to a significant increase with Healon GV and Healon5 even after 1-week follow-up.
Meta-analysis results implied significant improvement in postoperative BCVA in a day regardless of OVD. However, a single study showed a nonsignificant improvement in postoperative BCVA in 6-week follow-up with Vitrax and Anterior Chamber Maintainer (ACM). erefore, additional research is required. e reason behind substantial between-study heterogeneity could reveal different study populations, demographics, inclusion/exclusion criteria, study location, design, OVDs used, surgeon's experience, available facilities to       perform cataract surgery, rates of complications, and years when the surgeries were performed, as well as years when the studies were conducted. e results imply that good quality studies with longer follow-up periods need to be reported to better understand the optimal role of various OVDs in IOP management.
e limitations for systematic reviews and meta-analyses such as this one are necessary before conclusions are made. Firstly, included articles were of high, medium, and poor quality. However, few studies evaluating each OVD were available for analysis; all were included, irrespective of their quality. is is a recognized, but necessary, limitation due to the few clinical studies currently available examining each OVD. Secondly, meta-analysis of observational studies is influenced by inherent biases in the included articles [41].
For example, a multitude of other factors such as level of education, ethnicity, income status, socioeconomic status, previous ocular and nonocular surgeries, family history, other ocular and nonocular diseases, preoperative and postoperative medications, number of medications, and comorbidities (e.g., high blood pressure, diabetes, stroke, heart conditions, etc.) could influence the estimates in the original studies.
Our analysis indicated improvement in IOP with Viscoat + Provisc in 24-hour follow-up. Additionally, IOP nonsignificant upsurge was observed with Healon GV, Healon5, 2% HPMC, and OcuCoat compared to significant upsurge with Healon, Viscoat, and Soft Shell in 24-hour follow-up. e reason for this could be careful removal of OVD after cataract surgery. erefore, it is not possible to  differentiate OVD in increasing IOP. Further, additional research is needed to better understand how to maximize the utility of OVD in cataract management.
In conclusion, results indicated that postoperative IOP does significantly increase irrespective of OVD up to 5 hours of follow-up. erefore, careful removal of OVD is essential to avoid the IOP spikes. More good quality RCTs are needed to better understand and define the position of OVDs in cataract management.

Data Availability
e data used to support the findings of the study are available from the corresponding author upon request.

Conflicts of Interest
e authors declare that there are no conflicts of interest regarding the publication of this paper.

Authors' Contributions
Malvankar-Mehta and Hutnik contributed to study concept and design; participated in administrative, technical, or material support; and supervised the study. Fu and Subramanian were responsible for data acquisition. Malvankar-Mehta drafted the manuscript, carried out data analysis and interpretation, and takes responsibility for the integrity of data and the accuracy of data analysis. Malvankar-Mehta,