Clinical context

Cataract surgery outcomes are not defined by best-corrected visual acuity alone. Refractive accuracy, vision at different distances, spectacle dependence, contrast sensitivity, dysphotopsia and procedural safety all matter. Corneal and macular status, capsular support, cataract density and IOL calculation also shape the final result. A statistically convincing average therefore still has to be interpreted against the characteristics of an individual eye.

What was studied

Routine laboratory tests and ECGs are often ordered before low-risk cataract surgery despite no demonstrated benefit in patients without an indication. At UCLA Health, physicians conducting preoperative evaluations were randomized to usual care or one of three interruptive EHR alerts. Messages cited recommendations and emphasized lack of safety benefit, financial burden or psychological stress. Encounters within 30 days of surgery were included. The primary outcome was change in the proportion of visits with any signed testing order relative to the preceding 12 months.

How the evidence was produced

Randomization reduces the influence of known and unknown confounders and makes comparisons between interventions more reliable. Readers should still examine sequence generation, allocation concealment, masking of outcome assessors, the prespecified primary endpoint, losses to follow-up and intention-to-treat analysis. These details determine how confidently an observed difference can be attributed to the intervention.

Key result Of 776 randomized physicians, 229 had at least one eligible encounter. The intervention period included 1,045 visits from 967 unique patients; mean age was 72.1±9.1 years and 56.9% of encounters involved women. Mean ordering declined from 83% to 79% in the combined alert arms and from 84% to 82% in control. In mixed-effects models, any nudge did not significantly reduce overall testing: adjusted OR 0.68, 95% CI 0.34–1.34. In secondary analyses, 16% of orders were cancelled after an alert. The psychological-stress message was associated with fewer complete blood counts, OR 0.37, and ECGs, OR 0.42, both P=.01.

How to interpret the result

The primary conclusion is negative: an EHR alert alone did not solve overall low-value testing. Secondary signals should not replace that result, though they may inform a future intervention combined with workflow, payment or ordering changes. Physician-level randomization reduces contamination, but effectiveness depends on who signs tests and local workflow. The trial concerns reducing routine orders, not withholding evaluation when symptoms, unstable disease or a specific anesthesia indication is present.

Clinical applicability

This evidence is most useful as a basis for preoperative discussion and surgical planning. It helps identify which outcomes should be measured and which trade-offs should be explained to patients. The average effect should not be generalized to every IOL, technique or patient without checking eligibility criteria, surgeon experience, biometry, ocular comorbidity and duration of follow-up.

Limitations

Only 229 of 776 randomized physicians contributed eligible encounters, reducing the effective sample and limiting generalizability. This single academic system had very high baseline testing. The before-and-after outcome within a clustered randomized design requires appropriate modeling, and the primary confidence interval was wide. Test-specific analyses were secondary and multiple. Verify allocation, cluster power, physician balance, alert exposure and fatigue, missing data, clinical exclusions, unintended harms, cost, NCT04104256 registration, funding and conflicts.

What to verify in the source

Use the DOI below to verify complete methods, result tables, funding and conflicts of interest.