Researchers develop a new way to screen for diabetes using EHRs
February 19, 2016
Health IT
Population Health
Primary Care
Christina Hwang, Contributing Reporter
By tapping into the data of approximately 10,000 electronic health records (EHRs), UCLA researchers have created an algorithm to help diagnose individuals with Type 2 diabetes. The researchers have also found correlations between diabetes and new risk factors, such as an earlier diagnosis of sexual and gender identity disorders, viral and intestinal infections and chlamydia, to name a few.
The team – behind Ariana Anderson, lead study author, and Mark Cohen, a Semil Institute professor – set out to design an algorithm that would be both accurate and efficient, even with an incomplete medical record. “We intentionally selected all patients in the database. We didn't omit any records which had limited information,” Anderson told HCB News.
They discovered that individuals with any diagnosis of sexual and gender identity disorder have the same, increased, chance of developing diabetes as someone who has high blood pressure — roughly 130 percent more likely. Similarly, people with a history of certain viral infections, such as chicken pox, have as much of an increased likelihood of developing diabetes as someone with high cholesterol.
There also discovered factors that lower the risk of diabetes. People who take anti-anxiety and anti-seizure medications were found to have a much lower risk, and people who get migraines also have a lower risk.
Diabetes is usually screened by doctors through conventional methods like blood pressure, BMI, age and gender. However, the researchers believe the algorithm has a 14 percent better chance of diagnosing an individual than a doctor.
As to why certain factors correlate with greater or lesser diabetes risk, the group said more research needs to be done. Since the analysis was based largely on diagnostic codes rather than on actual diagnoses, the findings are not “fine-grained” enough to precisely link certain conditions to diabetes, they noted.
Anderson and her team plan on verifying their findings by using a different database of EHRs in the University of California’s medical system. If the algorithm is used on a national scale, 400,000 new individuals may potentially be diagnosed with diabetes.
“Given that one in four people with diabetes doesn't know they have the disease, it’s very important to be able to say, ‘This person has all these other diagnoses, so we're a little bit more confident that she is likely to have diabetes,’” said Anderson, in a statement.
By using technology to analyze medical records, new health patterns and diagnoses could be investigated. “This is a treasure trove of information that has not begun to be exploited to the full extent possible,” added Cohen.
Back to HCB News
By tapping into the data of approximately 10,000 electronic health records (EHRs), UCLA researchers have created an algorithm to help diagnose individuals with Type 2 diabetes. The researchers have also found correlations between diabetes and new risk factors, such as an earlier diagnosis of sexual and gender identity disorders, viral and intestinal infections and chlamydia, to name a few.
The team – behind Ariana Anderson, lead study author, and Mark Cohen, a Semil Institute professor – set out to design an algorithm that would be both accurate and efficient, even with an incomplete medical record. “We intentionally selected all patients in the database. We didn't omit any records which had limited information,” Anderson told HCB News.
They discovered that individuals with any diagnosis of sexual and gender identity disorder have the same, increased, chance of developing diabetes as someone who has high blood pressure — roughly 130 percent more likely. Similarly, people with a history of certain viral infections, such as chicken pox, have as much of an increased likelihood of developing diabetes as someone with high cholesterol.
There also discovered factors that lower the risk of diabetes. People who take anti-anxiety and anti-seizure medications were found to have a much lower risk, and people who get migraines also have a lower risk.
Diabetes is usually screened by doctors through conventional methods like blood pressure, BMI, age and gender. However, the researchers believe the algorithm has a 14 percent better chance of diagnosing an individual than a doctor.
As to why certain factors correlate with greater or lesser diabetes risk, the group said more research needs to be done. Since the analysis was based largely on diagnostic codes rather than on actual diagnoses, the findings are not “fine-grained” enough to precisely link certain conditions to diabetes, they noted.
Anderson and her team plan on verifying their findings by using a different database of EHRs in the University of California’s medical system. If the algorithm is used on a national scale, 400,000 new individuals may potentially be diagnosed with diabetes.
“Given that one in four people with diabetes doesn't know they have the disease, it’s very important to be able to say, ‘This person has all these other diagnoses, so we're a little bit more confident that she is likely to have diabetes,’” said Anderson, in a statement.
By using technology to analyze medical records, new health patterns and diagnoses could be investigated. “This is a treasure trove of information that has not begun to be exploited to the full extent possible,” added Cohen.
Back to HCB News
1(current)
You Must Be Logged In To Post A CommentRegisterRegistration is Free and Easy. Enjoy the benefits of The World's Leading New & Used Medical Equipment Marketplace. Register Now! |
|
