What did the authors study?

There are many different types of genetic tests available to doctors treating epilepsy. Some of them are easy, low-cost tests that are less likely to find the cause of seizures. Some are much more expensive but are much more likely to get answers. When doing genetic tests, most epilepsy doctors start with the lower-cost tests first. If these do not find an answer, they often will order more expensive tests until an answer is found. We do not know if this is the best approach.¹ In their article, “Diagnostic yield of genetic tests in epilepsy: a meta-analysis and cost-effectiveness study,” Fernández et al.² ask whether it would save money if we changed the order of genetic tests and did the most expensive tests first.

How was the study done?

The researchers first used information from many other studies to calculate how often each type of genetic test finds a useful answer. Then they used this information to calculate a cost per diagnosis. For example, if a test costs $100 but only finds an answer in 1 of 20 people tested, then it costs $2,000 to diagnose 1 person. If a different test costs $1,000 but finds an answer in 1 of every 2 people tested, then it costs the same $2,000 to diagnose each person, even though the test is much more expensive. Then they used computer simulations to make the same calculation on different sequences of tests. They compared how much it costs per diagnosis using the current approach of starting with the low-cost test first vs how much it would cost if doctors did the more expensive tests first.

What did the study show?

The first calculation found that the low-cost genetic test commonly used, a chromosomal microarray (CMA), found the cause of a person's epilepsy only 8% of the time. The typical “next step” in testing, or epilepsy panel (EP), found the answer 23% of the time. The most comprehensive and most expensive test, whole-exome sequencing (WES), found the answer 45% of the time.

The authors found that even though WES is the most expensive test overall, it costs the least amount per diagnosis at $15,000 per diagnosis. An EP costs slightly more money ($15,848) per diagnosis, and a CMA costs the most ($17,888). The authors also found that the least expensive sequence of testing is when doctors start with the most expensive test (WES), followed by an EP if the WES is normal, followed by a CMA if both the EP and WES tests are normal. However, when adjusting for the real world, the authors thought that the best sequence would be EP first, followed by CMA, and followed by WES if the others are normal.

What does this mean?

The world of genetic testing is changing very quickly, and the health care system is still trying to figure out the best way to do it. Doctors are also trying to figure out the best system; while we do not want to waste money on unnecessary tests, we want to do what is best for our patients. This study shows that what most doctors have been doing may not be what is best for our patients or for the health care system. It turns out that starting out with the low-cost test first ultimately costs more money overall because so many of those tests do not find an answer.

This study is a great start to convincing doctors and insurance companies to consider doing some of the more expensive tests first. However, the results need to be confirmed in the real world before this becomes common practice. Genetic testing is, and will continue to be, very much customized to each patient. As the costs to all of these tests are continuing to decrease, doctors will continue to ask what the smartest way is to finding answers for our patients.