Determination of valid benchmarks for outcome indicators in cataract surgery: a multicenter, prospective cohort trial

To evaluate a systematic approach to derive valid benchmarks for 2 outcome indicators intended to ascertain quality in cataract surgery and to propose benchmark levels drawn from the study results.

Prospective, multicenter cohort trial.

A total of 1685 patients (206-239 eyes per trial site) were recruited consecutively at 7 study sites. The patients featured age-related cataracts and were undergoing unilateral cataract surgery in the period between January 2007 and August 2008.

Only patients with uncomplicated age-related cataracts were included. Cataract surgery was performed by phacoemulsification. The SN60AT (Alcon, Inc., Fort Worth, TX) intraocular lens (IOL) was used as a study lens. The IOL power was calculated using the SRK-T formula with a standardized A constant. Biometry was performed with the IOL Master (Carl Zeiss Meditex, Jena, Germany). Only highly experienced senior surgeons were involved.

The outcome indicators 1 month and 3 months after surgery were the respective achievement of: (1) maximum absolute deviation of 0.5 diopter (D) between target refraction and postoperative spherical equivalent (primary end point, refractive accuracy); (2) best-corrected visual acuity of at least 0.8 (secondary end point, visual acuity outcome).

In the pooled data, maximum absolute deviation of ± 0.5 D from target refraction was achieved in 80% (95% confidence interval, 78%-82%) of cases. Visual acuity of 0.8 or more was reached in 87% (95% confidence interval, 80%-93%) of cases. The results from the trial centers differed significantly in the outcomes of the primary and secondary end points (P<0.001).

The study quantified benchmark levels for 2 outcome indicators in a standardized cataract surgery procedure. External confounding factors such as the comorbidity of patients, which cannot be influenced by the surgeon, were excluded. The derived benchmarks selectively illustrate the quality of the surgery and are superior to success rates published in the literature from unspecific data collections. This method is more suited for improving outcome quality by benchmarking. General methodologic problems are discussed, leading to recommendations for future study designs.

The author(s) have no proprietary or commercial interest in any materials discussed in this article.