Refining Metrics for Assessing the Quality of Care of Patients Undergoing Cataract Surgery for Use in Big Data Analyses

Abstract

Purpose : The AAO has developed metrics that can be applied to registry or EHR data to assess and compare eye care quality. Two such metrics assess cataract surgery quality: IRIS Measure 191 (≥20/40 BCVA within 90 days) and Measure 192 (complications requiring return to OR within 30 days). In these metrics, patients with various ocular comorbidities are excluded. We determined the success rates using these metrics and examined how success rates change after adjusting the definitions of success and the criteria to be eligible for inclusion in each metric.

Methods : We identified all cataract surgeries in the Sight Outcomes Research (SOURCE) multicenter ophthalmology EHR repository, which captures surgeries performed at academic centers participating in the collaborative. We identified the BCVA for every postoperative visit and used CPT codes to identify return trips to the OR. We calculated the % of cataract surgeries with ≥20/40 BCVA and % with return trips to the OR within 30 or 90 days of surgery. We explored how success rates would change by altering the BCVA cutoff for success, the postoperative period of assessment, and the eligibility criteria for each metric.

Results : Among the 13487 surgeries, 91.6% achieved ≥20/40 BCVA at 90 days. If the definition of success was adjusted to ≥20/30 or ≥20/25 BCVA, the 90 day success rate would be 88.4% and 82.9%, respectively. After excluding 6791 surgeries for patients with ocular comorbidities outlined in Measure 191 (50.4%), the 90 day success rate of achieving BCVA of ≥20/40, ≥20/30, and ≥20/25 was 96.6%, 94.4%, and 90.8%, respectively. Among all surgeries, 99.4% had no return to the OR within 30 days. After excluding surgeries for patients with ocular comorbidities outlined in Measure 192 (40.5%), the success rate was 99.7%.

Conclusions : The success rate for both existing AAO IRIS quality metrics for cataract surgery is very high (>96%) and nearly 50% of all surgeries are excluded due to ocular comorbidities. While applying more stringent definitions of success (e.g. BCVA ≥20/25) would generate a better distribution of outcomes for more meaningful comparisons, current performance falls short of the envisioned 100% goal. Including patients with ocular comorbidities would increase sample size, albeit with the need for added analytical approaches for proper case-mix adjustment.

This abstract was presented at the 2019 ARVO Annual Meeting, held in Vancouver, Canada, April 28 - May 2, 2019.

 

Table 1. Best Recorded Visual Acuity

 

Table 2. No Re-operation