Genetic Testing Is Being Used More Often. What Does That Mean for Patients?

I’m really excited about a new study that evaluated trends in genetic testing across a span of 20 years. Here’s the upshot: more and more patients are getting tested, and those tests are leading to more accurate diagnoses.

Back in the days of the Human Genome Project, when scientists were entrusted with nearly $3 billion in taxpayer funds to reveal our genome sequence for the first time, the great hope was that a deeper understanding of our genetic code would eventually lead to better healthcare. This report is a great illustration that the strategy is paying off.

A couple of caveats about this study: first, it was performed with data from a single health system, and a major academic medical center at that. Results won’t match what’s happening at your local community hospital, but the overall trends are still informative. Second, for various reasons the team excluded several large categories of molecular testing, including infectious diseases, pharmacogenetics, and certain prenatal and cancer tests. As some of the major uses for molecular diagnostics, those are huge exclusions, so we’ll have to understand that the test numbers reported are artificially low.

With those disclaimers out of the way, let’s take a look at what the researchers learned.

Just the stats, ma’am

The study was conducted on electronic health records from 1.86 million patients treated at Vanderbilt University Medical Center from January 2002 through December 2022. In 2002, just 1% of patients had any kind of genetic test result in their file; that increased to 6.1% by 2022. (Personally, I would have expected that to be higher. We still have a long way to go.) Encouragingly, the number of diseases diagnosed with genetic testing increased from 51 to 509.

Testing trends

During the course of two decades, it wasn’t just that genetic testing was used more often — the types of tests evolved quite a bit. In 2002, physicians usually ordered the simplest tests; more than 60% of tests looked at either a single gene or a single genetic variant. By 2015, tests covering multiple genes began to be used more frequently than single-gene tests. These gene panels were used in nearly half of all tests run in 2022. Much more comprehensive testing based on whole exome or whole genome sequencing was introduced in 2011; these approaches diagnosed 410 diseases, including 168 that were not (and most likely could not be) diagnosed with any other technique. While the vast majority of tests were deployed to diagnose a disease or condition, about 15% of tests were given for carrier screening or for testing family members when a relative was diagnosed with a hereditary disease.

Diagnoses … and lack of diagnoses

For all those (um … millions of?) Salisbury’s Take readers who feel apprehensive about genetic testing, this may be a comforting statistic: across the two decades, more than 63% of tests returned negative results. Less than 16% supported a diagnosis, about 6% gave a carrier status, and less than 3% reported a genetic risk factor. Most of the diseases diagnosed were rare, with 85% of them diagnosed in 10 patients or fewer and a little less than half diagnosed in only one patient.

But the data also pointed to a challenge that’s growing right alongside the rise of genetic testing: the return of inconclusive results. When a genetic variant is detected but there’s not enough clinical or scientific evidence to classify it as benign or pathogenic, it’s put into another category: a variant of unknown significance, or VUS. Across the entire study period, genetic tests turned up more than 28,000 pathogenic variants, which could be useful for a diagnosis, and more than 25,000 variants of unknown significance. But as tests broadened out beyond single genes and variants, those variants of unknown significance became more common. By 2018, the researchers found, patients got this kind of inconclusive result more often than a pathogenic variant.

The trend has continued since then and it represents a huge challenge in patient care and management. These VUS results typically have no immediate value, but over time, new evidence may help tip the scales. That same variant that has no known significance today could be considered a disease-causing variant a few years from now. Who’s on the hook to track down all the patients who had this variant and give them a new diagnosis when that happens?

The hunt for data

Aside from a clear increase in the use of genetic testing, another theme jumps out from this study: the challenge of finding and analyzing all of this data. The amount of hunting and data scrubbing that had to be performed to yank all of these genetic test results out of electronic health records represents quite the labor of love. Information was culled from pathology reports and clinical notes, but the scientists suspect that other data may have been missed since there is no consistent method for describing genetic tests in patient records. “Researchers have long envisioned a system where genetic test results are stored in a searchable format and easily flow between systems, similar to other laboratory test results,” lead author Lisa Bastarache and colleagues from Vanderbilt University Medical Center wrote in the paper published in the American Journal of Human Genetics. “In reality, genetic test results are often stored in the [electronic health record] as scanned documents that are not machine readable and are hidden from easy accounting.”

While the Vanderbilt team had the expertise and resources to mine their data, most health system teams won’t. A better system, with structured data and consistent formatting, would make it easier to perform this kind of deep dive at other institutions.

This could well be the most comprehensive view of genetic testing across such a long time period within a single healthcare system. It clearly shows the growing value of genetic testing for patient care, but also raises important questions about the complexity of some of these tests, how physicians and patients understand the results, and what more we could gain with a better system for reporting and storing genetic test data.