Optimal Timing for RRD Repair

Visual function after rhegmatogenous retinal detachment (RRD) repair is influenced by multiple factors, but a key modifiable factor is timing to repair relative to symptom onset. Traditional teaching has recommended treatment should occur within 24 to 28 hours for macula-on RRDs and within 7 days for macula-off RRDs to limit vision loss. Prior studies found that once the macula is compromised, similar visual outcomes are achieved whether repair takes place on day 1 versus day 7. Many studies advocating this interval treated the patients with scleral buckle (SB) procedures with small patient populations. Today, the primary procedure of choice for RRD is more heterogenous with a shift towards pars plana vitrectomy (PPV) of varying gauges with gas or silicone oil endotamponades.   

This meta-analysis by Muni et al examined the relationship between time to RRD repair and visual outcomes. The study included SB, PPV, combined SB and PPV, or pneumatic retinopexy. This approach to inclusion reflects the current approaches to retinal detachment repair and as such may give updated guidance regarding optimal timing and if prior recommendations still apply given advancements in vitrectomy.

The primary objective of this meta-analysis was to evaluate the relationship between time to surgery and visual outcomes for macula-on and macula-off RRDs. Secondary objectives included the intervention type and patient characteristics. Primary outcomes were the following: final best corrected visual acuity (BCVA), change between preoperative BCVA and final BCVA (∆BCVA), and relative risk of final BCVA <0.4 logMAR (better than 20/50 Snellen) between macula-off RRD repair in 0 to 3 days versus 4 to 7 days and macula-on RRD repair in 0 to 24 versus greater than 24 hours. Secondary outcomes included other time points, operating time, the relative risk of primary reattachment, and the relative risk of complications. The authors also had the goal to conduct subgroup analyses within each outcome according to study design, duration of follow-up, surgery performed, lens status, and endotamponade used; however, due to study reporting limitations, these subgroups could not be accurately analyzed and were not ultimately included.

The authors performed a systematic literature search of studies published in the English language using Ovid MEDLINE (2000 to April 2022), EMBASE (2000 to April 2022), and Cochrane CENTRAL (start of database [1993] to April 2022). Randomized controlled trials or observational studies were included if they reported BCVA or primary retinal reattachment rates following RRD repair and analyzed the outcomes from presentation or symptom onset (central vision loss) to surgery. A standardized data collection form was used to extract study identifiers, baseline characteristics (including number of eyes, gender, age, ethnicity, macula status, lens status, preoperative BCVA, duration of symptoms/time to surgery, IOP), intervention characteristics (procedure type, tamponade, vitrectomy gauge size) and outcomes (length of follow-up; BCVA at 3, 6, and 12 months, and final follow-up; change in BCVA from baseline; primary reattachment rate; and complications).

Quality of evidence, conflict of interest, and study sponsorship was assessed by two independent reviewers using the Risk of Bias in Nonrandomized Studies—of Interventions (ROBINS-I) and the Grading of Recommendations, Assessment, Development, and Evaluation (GRADE) tool. For statistical analysis, baseline demographics were reported as a proportion for categorical variables and means with standard deviation for continuous variables. A random effects meta-analysis was conducted using an empirical Bayesian estimator for all outcomes. Continuous outcomes were reported as a mean difference with a 95% confidence interval (CI), and binary outcomes as a relative risk with 95% CI. Heterogeneity was examined using I² statistics. The authors also pursued sensitivity analyses by sequentially removing each study and reanalyzing the remaining studies to identify outliers and influential studies on the results. P values of at least .05 were considered statistically significant. Further details regarding statistical analysis can be found in the article methods but overall appear comprehensive and appropriate.

From their search, 20 observational studies, including 1929 eyes, met inclusion criteria for analysis. Studies were conducted in the United States, United Kingdom, Germany, Japan, Canada, the Netherlands, Taiwan, Turkey, Italy, New Zealand, Pakistan, and Switzerland. Each of the included study sample sizes ranged from 12 to 199 eyes. The mean age of subjects was 58.2 (range: 39.2 to 64.9) and 39% were female (range: 18.8% to 64.5%). Eighty percent of studies (16/20) assessed macula-off RRDs, of which 14 studies defined duration of RRD as time from symptom onset. Time from symptom onset was defined by 10 studies as “central vision loss” and was not defined by four studies. Twenty percent of studies (4/20) assessed macula-on RRDs and all defined duration of RRD as time from initial examination to surgery. Due to this, the authors used time from symptom onset to RRD repair for macula-off RRDs and time from initial presentation to RRD repair for macula-on RRDs for statistical analysis. The mean postoperative follow-up time was 15.9 months (range: 3 to 55).

This meta-analysis found that macula-off RRD repair in 0 to 3 days from symptom onset had superior final BCVA with a mean 0.3logMAR (20/44 Snellen) compared to 4 to 7 days with mean 0.47logMAR (20/59 Snellen) vision (MD –0.06 [95% CI –0.09, –0.03], P < .001). There was no difference between groups for ∆BCVA (MD 0.03 [95% CI –0.18, 0.25], P > .05) or for the relative risk of BCVA <0.4 logMAR (RR 1.27 [95% CI 1.09, 1.49], P < .01). When looking at the tables provided in greater detail, the mean trends certainly do favor repair in 0 to 3 days compared to 4 to 7 days in macula-off RRDs. Three studies were pooled from the 20 to complete these calculations. Macula-on RRD repair in 0 to 24 hours from diagnosis had superior final BCVA compared to greater than 24 hours (MD –0.02 [95% CI –0.03, –0.01], P < .05).

The results were obtained for macula-off RRDs and macula-on RRDs were obtained from articles assigned moderate- and low-quality, respectively, using the GRADE evaluation tool employed by the authors, and they suggest this may be influenced by differences in baseline BCVA. Using the ROBINS-I tool, 55% (11/20) of observational studies had a low overall risk of bias, 40% (8/20) had a moderate risk, and 5% (1/20) had a high bias risk. As mentioned above, 14 of 16 studies included in the macula-off RRD studies included the symptom onset to repair duration. Ten studies defined symptom onset as “central vision loss,” four studies did not provide further details on the nature of vision loss/progression and two studies did not define if the duration of macular detachment was from symptom onset or presentation to the surgeon. Interestingly, exclusion of the studies lacking definitions did not significantly change the results of this analysis. 

The limits of this meta-analysis is that only nonandomized, retrospective observational studies were included. Muni et al report that while the included studies controlled for possible confounding variables through inclusion and exclusion criteria, many did not perform statistical adjustments. The duration of a macula-off RRD, using symptom onset as the marker for onset, can also be less reliable than time from diagnosis. The primary endpoint of final BCVA may also be difficult to judge given differential follow-up durations, but subgroup analysis performed by the authors did not note significant differences. Muni et al noted that adjustments for baseline BCVA would also have been ideal but could not be performed and may affect the estimates for ∆BCVA values. This data set also uses time point thresholds opposed to the definite duration before surgical intervention. Lastly, due to inconsistencies in reporting in the studies, the authors were unable to assess whether the intervention type, lens status, or endotamponade influenced the results.

In summary, this meta-analysis sought to examine the relationship between time to RRD repair and visual outcomes. It was observed that the final BCVA in macula-off RRDs was improved if treated in 0 to 3 days. The improvement, while statistically significant was approximately 0.06 logMAR (∼3 Snellen letters). Earlier repair, while ideal, may be limited by availability of ORs and staff. With increasing emergent surgeries, the retina community may face increased intraoperative complications while working with less experienced support teams and physician fatigue. It is therefore vital to know how long RRD repair can be delayed while ensuring best visual outcomes. A modern understanding of the relationship between visual outcomes and time to repair will help determine how to best triage these patients and advocate for ideal OR access. More studies are needed to help navigate this. This meta-analysis is an excellent step in this direction that does show a trend toward improved outcomes with earlier intervention for both macula-on and macula-off RRD groups.