Dr. Geoffrey Sher just published an insightful blog post outlining some of the deficiencies in the current SART/CDC IVF success rate model. Specifically he mentions:
- Currently all data collected is self reported by the member clinics and is not audited by a third party.
- Spot checks are performed on only approximately 10% of member clinics
- These spot checks verify the number of live births but do not verify the total number of cycles performed by the clinic. This means that the success rate reported by the clinic still remains unverified.
- The current method of measuring success categorizes success rates by woman's age and does not consider other important variables such as prior IVF failures and cause of infertility.
I think Dr. Sher is spot on with these criticisms. While it is important to give prospective patients an easy starting point in evaluating the success rates of the clinic they are considering for treatment, a system that would enable a more accurate and personalized exploration of that data would only serve the patient better.
What do you think? Does the current system of data collection work well enough? How would you make it better?