Paper: New Uses of Medical Records to Help Diagnose Genetic Disorders
OnFocus – Over a lifetime, a person may experience a wide variety of symptoms and diseases. Any one of these conditions by themselves may seem like an everyday event from a doctor’s perspective. When considering these symptoms and diseases together, it may be an indication of a more sinister ailment that is influenced by genetics but less obvious to the doctor.
Marshfield Clinic Research Institute (MCRI) and data from Marshfield Clinic’s biobank were part of a study that developed a computer-based method to search patient medical records to understand the possible impact of genetic diseases on health. More specifically, medical histories hidden within patient electronic health records may hold the key to diagnosing rare genetic diseases that are misdiagnosed as more common ailments such as heart or kidney disease, according to research published last week in the journal, Science.
Authors used clusters of symptoms to create a “phenotypic risk score” for many different genetic diseases. Those risk scores were then calculated and combined with genetic information collected from de-identified individuals and electronic health records from Vanderbilt University and Marshfield Clinic. It was hypothesized that individuals with high risk scores may have a genetic disorder; those with low scores were less likely to have a genetic disease.
Researchers found 18 relationships between genetic variants and high phenotypic risk scores. Some of these 18 genetic-disease relationships are well known like those related to cystic fibrosis. Some of the genetic-disease relationships are new and may provide new insights on how genetic diseases travel in families. The surprising part of this research was that some patients with high disease risk scores, and had a genetic variant that likely explained their high risk scores, were never diagnosed with genetic diseases.
“This innovative research shows we can use a clustering of symptoms and diseases to help identify individuals who should have genetic testing to help a doctor diagnose there conditions appropriately,” said Dr. Scott Hebbring, a Marshfield Clinic scientist and co-author on this study.
In addition to identifying individuals who may need genetic testing, the authors also show that the phenotypic risk scores may be used to prevent severe outcomes. Long ago, coal miners used canaries to see if the mine was safe for work. If the canaries died, the miners knew early that the mine was filled with poisonous gas and unsafe. In the instance of this research, study results showed that 14 percent of patients who had a genetic variant that affects the kidneys ended up having a kidney transplant whereas 10 percent with another variant had liver transplants. Had the patient’s genetic disease been diagnosed sooner by using the phenotypic risk scores and the DNA information, some of these transplants may have been preventable.
Researchers also learned that only one variant was needed to affect a patient’s health; not two mutations as previously thought for recessive genetic diseases. With medical genetic testing and commercially-available direct to consumer genetic testing more common, additional research may be done to better understand disease-associated genetic variants.
This study used historical data to evaluate the phenotypic risk scores and genetic information. “This is an example how we can use electronic medical records to better serve our patients,” said Dr. Murray Brilliant, Director for the Center for Human Genetics at Marshfield Clinic and co-author on the study. “Future research will be needed to see how best we can implement these computer-generated scores in actual clinical practice.”
In the end, it is hoped that this study will build on the growing evidence that extensive medical and genetic data may be able to improve the diagnosis of genetic disorders to better treat our patients.
-Submitted to FOCUS by Marshfield Clinic