Prednisolone and Ketorolac vs Ketorolac Monotherapy or Sub-Tenon Prophylaxis for Macular Thickening in Cataract Surgery: A Randomized Clinical Trial

Key Points

Question  Is a combination of corticosteroid and nonsteroidal anti-inflammatory drug (NSAID) eye drops superior to NSAID eye drops alone or dropless surgery with a sub-Tenon dexamethasone depot in controlling postoperative central macular thickening after uncomplicated cataract surgery?

Findings  In this randomized clinical trial with 470 participants, no differences in central subfield thickness or visual acuity across treatment arms were identified, although approximately half of the group given the sub-Tenon depot received additional anti-inflammatory treatment.

Meaning  Therapy with NSAID plus corticosteroid eye drops was not superior to NSAID monotherapy or sub-Tenon depot for postoperative central macular thickening after uncomplicated cataract surgery.

Importance  The choice of anti-inflammatory prophylaxis parallel to cataract surgery is important for patient safety and successful outcome of surgery, but which regimen to choose is contested.

Objectives  To determine whether a combination of prednisolone and nonsteroidal anti-inflammatory drug (NSAID) eye drops was superior in preventing increased central macular thickness (central subfield thickness [CST]) after uncomplicated cataract surgery compared with NSAID monotherapy and sub-Tenon capsule depot (dropless surgery), and to test whether preoperative initiation of eye drop treatment was superior to initiation on the day of surgery.

Design, Setting, and Participants  This investigator-driven, single-center, randomized clinical trial with masked statistical analyses enrolled patients at the Department of Ophthalmology, Rigshospitalet Glostrup, Glostrup, Denmark, from February 1, 2018, to August 15, 2019. Follow-up was completed December 18, 2019. Participants included low-risk patients undergoing phacoemulsification for age-related cataract by an experienced surgeon (1 eye per participant). Data were analyzed from February 17 to June 15, 2020.

Interventions  Participants scheduled for cataract removal were randomized to 1 of 5 anti-inflammatory prophylactic regimens: eye drops with a combination of prednisolone, 1%, and ketorolac tromethamine, 0.5%, with or without preoperative initiation (preoperative prednisolone plus NSAID [control] and postoperative prednisolone plus NSAID groups), ketorolac monotherapy with or without preoperative initiation (preoperative and postoperative NSAID groups), or sub-Tenon depot of dexamethasone phosphate (sub-Tenon group). Eye drops were administered 3 times per day until 3 weeks postoperatively.

Main Outcomes and Measures  CST 3 months postoperatively.

Results  A total of 470 participants (mean [SD] age, 72.2 [7.0] years; 290 women [61.7%]) with 94 participants in each group were included in the analysis. Three months after surgery, the mean CST was 250.7 (95% CI, 247.6-253.7) μm in the preoperative prednisolone plus NSAID group, 250.7 (95% CI, 247.8-253.7) μm in the postoperative prednisolone plus NSAID group, 251.3 (95% CI, 248.2-254.4) μm in the preoperative NSAID group, 249.2 (95% CI, 246.2-252.3) μm in the postoperative NSAID group, and 255.2 (95% CI, 252.0-258.3) μm in the sub-Tenon group. There were no significant differences in CST or visual acuity compared with control and no differences between preoperative and postoperative groups, but 47 of 83 participants (56.6%) in the sub-Tenon group needed additional anti-inflammatory treatment.

Conclusions and Relevance  No differences in CST or visual acuity were detected between the combination of prednisolone and NSAID eye drops vs NSAID monotherapy or sub-Tenon dexamethasone depot, although more than one-half of patients in the sub-Tenon arm received additional anti-inflammatory treatment. Initiating prophylaxis 3 days preoperatively was not superior to initiation on the day of surgery. Monotherapy with NSAIDs may be preferred in uncomplicated cataract surgery.

Trial Registration  ClinicalTrials.gov Identifier: NCT03383328

Pseudophakic cystoid macular edema (PCME) is an important cause of unsatisfactory visual outcome after cataract surgery. It affects visual acuity after approximately 1% of surgical procedures.^1,2 There is no common classification of PCME, and therefore estimates differ between studies. A meta-analysis³ reported that 4% to 25% had some degree of macular edema, even when prophylactic treatment was administered. It is believed that PCME is caused by the inflammatory response after cataract surgery, which disrupts the blood-ocular barrier leading to leakage of fluid into the retina.⁴ To control the inflammatory response and reduce the risk of PCME, prophylactic anti-inflammatory eye drops are prescribed parallel to cataract surgery.

The choice of anti-inflammatory prophylaxis is contested.^5,6 Two types of drugs are available: corticosteroids and nonsteroidal anti-inflammatory drugs (NSAIDs). Corticosteroids inhibit phospholipases and thereby both prostaglandin and lipoxin synthesis,⁷ and NSAIDs selectively inhibit cyclooxygenase enzymes, which convert arachidonic acids to prostaglandins⁷; however, meta-analyses^3,5,6,8 have found that NSAIDs are more effective in preventing PCME. It has been suggested that preoperative initiation of NSAID prophylaxis is beneficial,^9-11 but to our knowledge, evidence from large randomized trials is missing.

Prophylactic treatment parallel to cataract surgery is traditionally administered as eye drops. Many patients have difficulties using eye drops and need assistance.¹² Therefore, several dropless approaches have been suggested, and some studies^2,13,14 have found that the effectiveness of these dropless approaches is comparable to corticosteroid eye drop monotherapy. To our knowledge, no studies have compared a dropless approach with NSAID eye drops. The purpose of this study was to determine whether (1) combination of prednisolone and NSAID eye drops is superior in preventing macular thickening after uncomplicated cataract surgery compared with NSAID monotherapy and dropless surgery using a sub–Tenon capsule dexamethasone phosphate depot and (2) preoperative initiation of eye drop treatment is superior to initiating treatment on the day of surgery.

Quiz Ref IDThis study was a randomized clinical trial comparing 5 different regimens for anti-inflammatory prophylaxis parallel to uncomplicated cataract surgery. The trial protocol and statistical analysis plan are found in Supplement 1 and Supplement 2, respectively. The control group received a combination of corticosteroid and NSAID eye drops, initiated 3 days before surgery (preoperative prednisolone plus NSAID group). Comparison groups received a combination of corticosteroid and NSAID drops initiated on the day of surgery (postoperative prednisolone plus NSAID group), NSAID eye drop monotherapy initiated 3 days before surgery or on the day of surgery (preoperative NSAID and postoperative NSAID groups, respectively), and sub-Tenon depot of dexamethasone administered at the conclusion of surgery (sub-Tenon group). Participants in the sub-Tenon group were not given anti-inflammatory drops. Corticosteroid drops were prednisolone acetate, 1% (Pred Forte; Allergan); NSAID drops, ketorolac tromethamine, 0.5% (Acular; Allergan); and sub-Tenon depot, 0.5 mL of dexamethasone phosphate, 4 mg/mL (Dexamethasone Krka [Krka] or Dexavital [Vital Pharma Nordic]). All drops were administered 3 times per day until 3 weeks postoperatively, and adherence was monitored by asking participants, in an unstructured manner, if they were taking their drops; no formal recording of compliance was collected. Ethics committee approval was obtained from the Committee on Health Research Ethics, Capital Region, Denmark. The study was approved by the Danish Medicines Agency and the Danish Data Protection Agency. Accordance with the good clinical practice quality standard was monitored by the good clinical practice unit at Copenhagen University Hospital, Copenhagen, Denmark. The study was conducted in accordance with the Declaration of Helsinki,¹⁵ and all participants provided written informed consent. Participants received no incentives or compensation for participation. This study followed the Consolidated Standards of Reporting Trials (CONSORT) reporting guideline.

We recruited participants among patients scheduled for cataract surgery at the Department of Ophthalmology, Rigshospitalet Glostrup, Glostrup, Denmark, from February 1, 2018, to August 15, 2019. Inclusion criteria were age-related cataract, capacity to consent, and consent to participation. In addition, women had to be postmenopausal. Exclusion criteria were known allergy to any content of the pharmaceuticals used; medical history of retinal vein occlusion, epiretinal membrane, uveitis, glaucoma, retinal detachment, diabetes, exudative age-related macular degeneration, or age-related macular degeneration with geographical atrophy; and significant complications of surgery such as posterior capsule rupture/vitreous loss, choroidal hemorrhage, or dislocated lens material. Participants with complications were excluded when complications were detected (Figure 1). Sample size was set at 94 per interventional group, which allowed us to detect a difference of 5 μm in central retinal thickness with a power of 90% at a 5% significance level or a power of 80% at a 1.25% significance level. Five micrometers was chosen at the investigators’ discretion.

One eye per participant was included using a computerized coin toss to decide which eye to include if both eyes were eligible. Immediately after receiving signed consent for participation, participants were randomized to 1 of the 5 interventional groups (Figure 1) by the same examiner (J.H.E.). We used the randomization instrument in Research Electronic Data Capture (REDCap)^16,17 hosted at Capital Region, Denmark. A block-randomized list was created at Sealed Envelope (https://www.sealedenvelope.com/simple-randomiser/v1/lists) and uploaded to REDCap by an independent researcher. The length of the list was 470, and block sizes were 5, 10, and 15 in random order.

All surgical procedures were performed by experienced surgeons (including L.M.H.), with experience defined as a minimum of 1000 cataract procedures performed within the 2 previous years. The surgical procedure was phacoemulsification with implantation of a hydrophobic intraocular lens (IOL) in the lens bag. Oxybuprocaine was used as local anesthetic, and eyes were instilled with phenylephrine hydrochloride, 10%, tropicamide, 1%, and ketorolac, 0.5%, before surgery. We used a 2.4-mm main incision and a 1-mm side port incision. Intracameral lidocaine hydrochloride, 1%, and viscoelasticum were instilled, followed by capsulorhexis and hydrodissection. For phacoemulsification, we used a surgical microscope and probe (Infinity Vision System; Alcon) with a miniflared 0.9-mm tip. On conclusion, the anterior chamber was irrigated with 1 mL of cefuroxime, 2.5 mg/mL.

Follow-up was completed December 18, 2019. Participants were examined at the preoperative visit (baseline) and 3 days, 3 weeks, and 3 months postoperatively. The primary outcome was central macular thickness at the 3-month visit measured by optical coherence tomography (DRI OCT Triton; Topcon Europe Medical BV) using the central 1.0-mm zone (central subfield thickness [CST]) of the ETDRS (Early Treatment Diabetic Retinopathy Study) grid obtained from the built-in software (IMAGEnet 6; Topcon Europe Medical BV).^18,19 Secondary outcomes were CST at 3 weeks postoperatively, corrected distance visual acuity (CDVA) in logarithm to the minimal angle of resolution (logMAR) using an ETDRS chart, intraocular pressure (IOP), and subjective tolerance of treatment. Intraocular pressure was measured using a rebound tonometer (iCare Finland) and controlled by Goldmann applanation tonometry if the IOP was greater than 25 mm Hg. Subjective tolerance to treatment was assessed 3 days and 3 weeks postoperatively by asking the participants if they experienced no discomfort or mild discomfort or moderate or severe discomfort. Cataract severity was graded using the Age-Related Eye Disease Study classification.²⁰ Adverse events were defined as events that were not expected as part of routine surgery for cataract or baseline conditions or as events that led to additional treatment. Corneal edema or dryness were not noted as adverse events unless additional treatment was initiated. The decision to institute additional treatment was made by experienced physicians (including J.H.E., L.M.H., and L.K.). Any need for additional topical anti-inflammatory, lubricating, IOP-lowering, or antibiotic treatment was registered. Adverse events with relation to the eyes were grouped as pain/soreness, insufficiently controlled anterior chamber inflammation, cystoid abnormalities on optical coherence tomography, dryness, corneal abrasion, swollen/red externa, elevated IOP of greater than 25 mm Hg, significant corneal edema, or other.

Data were analyzed from February 17 to June 15, 2020. All statistical analyses were conducted according to a prespecified analytical plan using the R statistical software, version 3.6.0 (R Program for Statistical Computing)²¹ and the nlme package (R Program for Statistical Computing). Pairwise comparisons between the control and comparison groups used a constrained linear mixed model with an unstructured covariance pattern and inherent baseline adjustment.²² Dichotomous outcomes were analyzed with Fisher exact test. All hypothesis tests were 2 sided. Primary analyses (change in CST 3 months postoperatively, relative to the control group) were adjusted for multiple testing using Bonferroni correction, resulting in a significance level of .0125. All secondary analyses were corrected for multiple testing using a false discovery rate correction.²³ We considered an adjusted P < .05 to indicate statistical significance. Because some participants were excluded after allocation, analyses followed a modified intention-to-treat approach that included all participants who provided baseline data. The constrained linear mixed model implicitly handled missing data by maximum likelihood estimation. We evaluated the sensitivity of our analyses (eTables 1-6 in Supplement 3). An independent researcher masked statistical analyses by renaming the interventional groups. Full masking was not possible because sub-Tenon treatment was recognizable for participants and outcome assessors. After unmasking, we made analyses of the combination of corticosteroid and NSAID eye drops vs NSAID monotherapy and analyses of preoperative vs postoperative initiation of eye drop therapy by pooling relevant groups. As an exploratory post hoc analysis, we counted cases with a CST increase of at least 10%.

A total of 470 participants (mean [SD] age, 72.2 [7.0] years; 290 women [61.7%] and 180 men [38.3%]) were included and randomized, with 94 participants in each interventional group (Table 1). After randomization, 14 participants were excluded. Nine of these were excluded owing to surgical complications and 5 were excluded owing to other criteria (eTable 7 in Supplement 3); thus, 456 participants provided baseline data and 429 completed 3 months of follow-up (Figure 1).

Three months after surgery, the mean CST was 250.7 (95% CI, 247.6-253.7) μm in the preoperative prednisolone plus NSAID group, 250.7 (95% CI, 247.8-253.7) μm in the postoperative prednisolone plus NSAID group, 251.3 (95% CI, 248.2-254.4) μm in the preoperative NSAID group, 249.2 (95% CI, 246.2-252.3) μm in the postoperative NSAID group, and 255.2 (95% CI, 252.0-258.3) μm in the sub-Tenon group. Differences between the control (preoperative prednisolone plus NSAID) and the comparison groups were 0.1 (98.75% CI, −5.4 to 5.5) μm (P = .97) with the postoperative prednisolone plus NSAID group, 0.6 (98.75% CI, −5.0 to 6.2) μm (P = .79) with the preoperative NSAID group, −1.5 (98.75% −7.0 to 4.1) μm (P = .51) with the postoperative NSAID group, and 4.5 (98.75% CI, −1.1 to 10.1) μm (P = .04) with the sub-Tenon group. Results are presented in Figure 2 and Table 2. The numbers of participants with CST increases of at least 10% are presented in eTable 8 in Supplement 3.

The mean IOP decreased in all groups postoperatively, but it was lower in groups not receiving prednisolone in the early postoperative period. At 3 months, we found no differences compared with the preoperative prednisolone plus NSAID group (−3.3 [95% CI, −3.8 to −2.8] mm Hg) in the postoperative prednisolone plus NSAID group (0.2 [95% CI, −0.5 to 0.8] mm Hg), the preoperative NSAID group (−0.3 [95% CI, −0.9 to 0.4] mm Hg), the postoperative NSAID group (−0.3 [95% CI, −0.9 to 0.3] mm Hg), and the sub-Tenon group (−0.2 [95% CI, −0.8 to 0.5] mm Hg (Table 2 and the eFigure in Supplement 3). Corrected distance visual acuity improved in all groups 3 months postoperatively with no identified differences compared with preoperative prednisolone plus NSAID group (Table 2).

Moderate or severe discomfort was reported by a total of 32 participants (7.3%) at 3 days postoperatively and a total of 13 (3.0%) at 3 weeks postoperatively. We identified no differences compared with the preoperative prednisolone plus NSAID group (Table 3).

Adverse events and addition of anti-inflammatory treatment were registered more often in the sub-Tenon group, but no differences were found between the preoperative prednisolone plus NSAID group and the remaining groups. Two participants were treated for elevated IOP in the immediate postoperative period, but no IOP elevations of greater than 25 mm Hg were found at any postoperative visit. In the sub-Tenon group, there were more unscheduled visits, but no differences were found between the remaining groups and the preoperative prednisolone plus NSAID group. Results are presented in Table 3.

No differences were identified between NSAID monotherapy and the combination of prednisolone and NSAID drops regarding CST or CDVA at any postoperative visit after pooling combination groups (preoperative and postoperative prednisolone plus NSAID) and NSAID monotherapy groups (preoperative and postoperative NSAID). Intraocular pressure was lower for NSAID monotherapy 3 days and 3 weeks postoperatively compared with the combination of prednisolone and NSAID. No differences in IOP were identified 3 months postoperatively (eTable 9 in Supplement 3).

Quiz Ref IDWe pooled groups with preoperative initiation (preoperative prednisolone plus NSAID and preoperative NSAID) and those with postoperative initiation (postoperative prednisolone plus NSAID and postoperative NSAID) of treatment. In this analysis, we identified no differences between preoperative or postoperative initiation of treatment for CST, CDVA, or IOP at any postoperative visit (eTable 10 in Supplement 3).

We found that our overall conclusions were not sensitive to handling of missing values, to restriction to the per-protocol population, or to inclusion, exclusion, or truncation of extreme outliers. In addition, postrandomization confounding from sex, age, and phacoemulsification energy was unlikely (eTables 1-6 in Supplement 3). Characteristics of participants with and without complete data are presented in eTable 11 in Supplement 3.

Quiz Ref IDWe investigated the effect of anti-inflammatory prophylaxis on macular thickening after uncomplicated cataract surgery in a randomized clinical study that evaluated the additive effect of prednisolone to NSAID eye drops in a fashion that allowed us to compare the effect of initiating treatment preoperatively with initiation on the day of surgery. In addition, we evaluated the effect of dropless surgery using a sub-Tenon dexamethasone depot. We found that NSAID plus corticosteroid drops were not superior to NSAID monotherapy or sub-Tenon dexamethasone depot for postoperative central macular thickening after uncomplicated cataract surgery.

One of the primary purposes of prescribing an anti-inflammatory prophylactic treatment in parallel with cataract surgery is to prevent PCME. Pseudophakic cystoid macular edema is infrequent and heterogeneously defined, which makes it difficult to use the incidence as an outcome in clinical trials. Instead, the CST measured with optical coherence tomography is generally accepted as a measure for comparing efficacy of prophylactic regimens. We did not find differences in CST between groups receiving NSAID monotherapy or a combination of NSAID and prednisolone. A large European multicenter trial¹⁹ compared corticosteroid monotherapy (dexamethasone, 0.1%) with NSAID monotherapy (bromfenac, 0.09%) and a combination of both and found no significant differences between the combination of corticosteroid and NSAID and NSAID monotherapy, whereas dexamethasone monotherapy resulted in increased macular thickness 6 weeks postoperatively. However, NSAID monotherapy is still controversial for prophylactic treatment after cataract surgery, especially in the US,²⁴ and NSAIDs are not recommended by the American Academy of Ophthalmology.¹⁰

A main argument for not recommending NSAIDs has been the lack of comparison with prednisolone drops, because they are theoretically more efficient than more potent corticosteroids owing to a greater intraocular bioavailability.¹⁰ Another argument was that the increased effect of adding NSAIDs to corticosteroids was likely caused by increased dosing. To address the first argument, we chose prednisolone rather than more potent corticosteroid drops for the combination groups. Our results showed that NSAID monotherapy was comparable to combined prednisolone and NSAID drops, even though the total dose was higher in the combination groups. This outcome corroborates the conclusions from other randomized trials that compared NSAID monotherapy with combinations of corticosteroid and NSAID eye drops; namely, there is no significant additional effect of adding topical corticosteroids to topical NSAIDs parallel to uncomplicated cataract surgery.^19,25 It could be argued that a result of no statistical significance is not evidence that there is no difference between groups. However, our study was powered to detect differences between groups as small as 5 μm, which we expected to be within normal variation between 2 measurements. Differences in CST and CDVA between eye drop groups and controls were minimal, and CIs were narrow. Therefore, it is unlikely that a larger sample size would uncover any clinically relevant differences.

Preoperative initiation of therapy did not result in decreased CST compared with postoperative initiation, even after pooling relevant groups. A previous randomized clinical trial⁹ concluded that preoperative initiation of ketorolac, 0.4%, was advantageous in the early postoperative period but showed no significant effects on long-term outcomes. No data on central macular thickness were reported, and the groups were small. We found no effect of preoperative initiation on any outcome measures.²⁶

In the sub-Tenon group, CST was significantly increased 3 weeks postoperatively compared with the preoperative prednisolone plus NSAID group, but the difference was no longer significant 3 months after surgery. More than half of the participants in the sub-Tenon group required additional topical anti-inflammatory treatment during the study period, which might explain why no difference in CST was found at 3 months. Other randomized¹³ and nonrandomized^2,27 studies have found comparable results between dropless surgery and corticosteroid eye drop monotherapy, but we were unable to locate reports comparing dropless surgery with combined corticosteroid and NSAID eye drops or NSAID monotherapy. We used a sub-Tenon depot of a highly potent corticosteroid, dexamethasone phosphate, and refrained from using triamcinolone, which has a much longer period of action, because of fear of sustained IOP elevations.¹⁴ Although it was safe to administer the sub-Tenon dexamethasone depot, it was not comparable to NSAID or prednisolone plus NSAID in preventing macular thickening after cataract surgery. In patients with suspected poor adherence with topical treatment, placing a depot of corticosteroid during surgery could serve as adjuvant therapy. However, IOP should be monitored regularly if triamcinolone is used.²⁸

Intraocular pressure was lower during eye drop treatment in groups who did not use corticosteroid drops compared with those who did. However, the mean IOP in all groups was low in the reference range, and there were no elevations of greater than 25 mm Hg at any postoperative visit. There were no differences in visual acuity at any time point, but adverse events and unscheduled visits were more often encountered in the sub-Tenon group.

Quiz Ref IDThis study has some limitations. We used a modified intention-to-treat approach for our analyses because some participants were excluded after allocation. We included patients with an expected low risk of complications, including macular edema, after cataract surgery, and the surgeons were experienced. Thus, our results may only apply to such patients and not patients with risk factors (eg, diabetes).^28,29 The study did not include a corticosteroid eye drop monotherapy group and therefore did not compare corticosteroid drops with NSAID drops. Prednisolone, 1%, is less potent than other commonly used corticosteroid drops. Other approaches to dropless surgery exist that may prove more efficient than sub-Tenon dexamethasone depot as used in this study.^30,31 Follow-up ended 3 months after surgery, and later presentations of PCME may not have been identified. The decision to add anti-inflammatory treatment was made by an unmasked physician.

In this randomized clinical trial, we found that a combination of prednisolone and NSAID eye drops was not superior to NSAID monotherapy and that initiating prophylactic treatment 3 days before surgery was not superior to initiating treatment on the day of surgery in preventing thickening of the central subfield after uncomplicated cataract surgery. In addition, we found that dropless surgery with a sub-Tenon dexamethasone depot was inferior to a combination of prednisolone and NSAID eye drops and that anti-inflammatory eye drops were required in a substantial number of participants in the sub-Tenon dexamethasone depot group. Intraocular pressure was higher in groups that used prednisolone eye drops compared with NSAID monotherapy and sub-Tenon dexamethasone depot in the first 3 weeks postoperatively. Therefore, NSAID monotherapy with initiation on the day of surgery may be preferred as an anti-inflammatory prophylactic regimen in uncomplicated cataract surgery.

Accepted for Publication: June 23, 2021.

Published Online: August 12, 2021. doi:10.1001/jamaophthalmol.2021.2976

Correction: This article was corrected on October 21, 2021, to correct a typographical error in the title of the Visual Abstract.

Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2021 Erichsen JH et al. JAMA Ophthalmology.

Corresponding Author: Jesper Høiberg Erichsen, MD, PhD, Department of Ophthalmology, Rigshospitalet Glostrup, Valdemar Hansens Vej 1-23, 2600 Glostrup, Denmark (jesper.h.erichsen@dadlnet.dk).

Author Contributions: Drs Erichsen and Kessel had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Erichsen, Holm, Kessel.

Acquisition, analysis, or interpretation of data: Erichsen, Forslund Jacobsen, Forman, Kessel.

Drafting of the manuscript: Erichsen, Holm.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Erichsen, Forman.

Obtained funding: Erichsen, Kessel.

Administrative, technical, or material support: Erichsen, Holm, Forslund Jacobsen, Kessel.

Supervision: Holm, Forman, Kessel.

Conflict of Interest Disclosures: None reported.

Funding/Support: This study was supported by grant DFF–7016-00161 from Independent Research Fund Denmark (Dr Erichsen), by Fight for Sight Denmark (Dr Erichsen), and by the Henry og Astrid Møllers Fond (Dr Erichsen).

Role of the Funder/Sponsor: The funding organizations had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Data Sharing Statement: See Supplement 4.