Do we need to hold aspirin before cataract surgery? A systematic review and meta-analysis of 65,196 subjects

Authors: Zeid MA, Elrosasy A, Alkheder A, Alkousheh H, Azzawi MADA, Mohamed SF, Alkhateeb A, Serhan HA.

Geographical coverage: Israel, United Kingdom, United States, China, India, and Japan.

Sector: Cataract surgery

Sub-sector: Aspirin

Equity focus: Not reported

Study population: Patients undergoing cataract surgery

Review type: Effectiveness review

Quantitative synthesis method: Meta-analysis

Qualitative synthesis method: Not applicable

Background:

Cataract surgery is frequently performed on older adults with various medical conditions. Many of these patients may be on aspirin therapy for cardiovascular reasons. There is an ongoing debate about the use of aspirin during the perioperative period due to concerns about the risk of bleeding versus thromboembolic events. Despite its common use, there is no consensus on whether to continue or discontinue aspirin before cataract surgery.

Objective:

To assess the safety of continuing or discontinuing aspirin in the setting of cataract surgery.

Main findings:

The review concluded that continuing aspirin before cataract surgery significantly increases the risk of subconjunctival haemorrhage, while no other postoperative outcomes differed significantly between continuing and discontinuing aspirin or between aspirin and placebo.

The systematic review included nine studies, encompassing a total of 65,196 patients and 65,507 eyes. The studies originated from various countries, including Israel (n=2), United Kingdom (n=1), United States (n=2), China (n=2), India (n=1), and Japan (n=1). Study designs included prospective and retrospective cohort studies, as well as case series. Sample sizes ranged from approximately 40 participants to over 45,000 participants. Follow-up periods ranged from one week to one month.

Meta-analysis revealed that continuing aspirin significantly increases the risk of subconjunctival haemorrhage (RR: 1.74, 95% CI: 1.22, 2.50, p = 0.002). No statistically significant differences were found in the incidence of hyphema, retrobulbar haemorrhage, vitreous haemorrhage, intraocular pressure spike, corneal oedema, posterior capsule rupture, or best-corrected visual acuity. This suggests that aspirin continuation increases the risk of subconjunctival haemorrhage but does not significantly affect other postoperative outcomes.

Methodology: 

The authors searched PubMed, Scopus, Web of Science, and EMBASE from inception to January 25, 2024. Terms related to aspirin and cataract surgery were used. Two authors independently screened articles, and disagreements were resolved with a senior author. A manual search was conducted to check citations and similar articles.

Inclusion criteria: patients undergoing cataract surgery (Population); aspirin or aspirin withdrawal (Intervention); placebo or continued aspirin (Comparator); any reported outcome (Outcome); RCTs or observational studies (Study Design).

Data were extracted by two authors using a predefined Excel sheet. Quality was assessed with the Newcastle-Ottawa Scale and NIH quality assessment tool. Meta-analysis was performed with RevMan and OpenMetaAnalyst. Risk ratios (RRs) and 95% CIs were calculated, and heterogeneity was assessed using χ² and I² statistics. Fixed-effect models were used for homogenous data and random-effects models for heterogeneous data.

Applicability/external validity:

The review acknowledges limitations that may affect generalisability, such as differences in study design, population, aspirin dosage, and surgical techniques. To improve applicability, future research should include more RCTs, evaluate newer anticoagulants, and perform long-term follow-up. Regular meta-analysis updates are recommended.

Geographic focus:

Israel, United Kingdom, United States, China, India, and Japan.

Summary of quality assessment:

Overall, the review’s conclusions are presented with moderate confidence. Language bias is present, as only English-language studies were included. Additionally, results were not stratified by risk-of-bias status for included studies.