October 13, 2004

VISX Accused of Submitting Falsified Data to the FDA

Understanding pre-market approval and labeling differences of two leading customized ablation platforms: a call for reform at the FDA.

Journal of Refractive Surgery 2004 Sep-Oct; 20(5):S588-92.
Boxer Wachler Vision Institute, Beverly Hills, CA 90210, USA. bbw@boxerwachler.com

PURPOSE: To analyze the Food and Drug Administration (FDA) pre-market approval reported data for two currently available wavefront-guided excimer laser systems.

METHODS: VISX and Alcon myopic wavefront laser in situ keratomileusis (LASIK) trials were analyzed based on public information available in the pre-market approval documents.

RESULTS: Clinical tests and outcomes reporting varied between pre-market approvals. One pre-market approval reported contrast sensitivity whereas the other did not, although data was collected in both studies. Based on the FDA definition of "accountability," one pre-market approval reported 12-month accountability of 97.7% although only 24.5% of patients were analyzed at that interval. One pre-market approval reported 6-month results whereas the other reported 6- and 12-month results. One pre-market approval reported grouped pre- and postoperative subjective survey results; the other reported individual changes in pre- to postoperative subjective symptoms. CONCLUSION: We recommend an FDA revision to the definition of "accountability" and also recommend future FDA refractive surgery trials be guided by standardized criteria.

Excerpt from full article

VISX reported that 97.7% (84/86) of eyes had UCVA of 20/20 or better at 12 months. These 86 eyes represent only 24.5% of patients. Accountability is reported to be 95.6%, despite the fact that 70.7% (248/351) of patients were labeled “not yet eligible” for analysis at 12 months.

In the VISX trial, 13 eyes were excluded from the data analysis because they required retreatments (12 within the study, 1 outside the study) as these 13 eyes had UCVA of 20/32 or worse. Adding these eyes to the 12-month data, UCVA of 20/20 can be recalculated to be 84.8% of eyes (84/99) at 12 months; this is in contradistinction to the 97.7% reported with exclusion of these 13 eyes that saw 20/32 or worse without correction.

The addition of these 13 eyes still only accounts for 28.2% (99/351) of 12-month results from patients enrolled in the clinical trial; the refractive outcomes for >70% of the study population are not reported. 

Source:  www.ncbi.nlm.nih.gov

Posted by Admin at October 13, 2004 05:54 PM