Diabetes and pre-diabetes affect an astonishing 43 percent of the country’s population at a cost of $237 billion in treatment and $90 billion per year in indirect costs such as absenteeism. The U.S. spends more treating diabetes than the entire GDP of Portugal.
The earlier the disease is caught, the more likely treatment costs will be kept down. But testing is expensive and time consuming, so providers need to be wise about who they test. Usually, the patients who receive a diabetes test already have a symptom, meaning the chances of reversal are low and treatment costs are more likely to be high.
But new research from SMU may change all that. Scholars are using a statistical model to predict potential patients for diabetes testing up to ten years before there are noticeable symptoms. If the disease can be caught in the pre-diabetes stage or even earlier, treatment costs can be reduced.
The current strategy to stop the development of diabetes is to give everyone a test, but false positives are common because they can be impacted by what the patient ate recently. The costs and multiple visits necessary for diabetes testing make it difficult to test everyone. In addition, there would be many tests of people who don’t have diabetes nor are likely to develop it.
There is a 9-12 year period where patients don’t show any symptoms, but testing during that period can help prevent its development years before the disease becomes significantly more expensive to treat.
The researchers worked with Parkland Health and Hospital System to look at data for 12,000 patients, creating a model that analyzes factors like gender, weight, race, family history and others to map out a future for patients. The model can make accurate predictions five to ten years out, and could help providers develop a plan about who to screen, when to screen, and how often.
The model wouldn’t require any additional technology or know how, and at a hospital like Parkland where treatment costs are funded by taxpayer dollars, reducing diabetes costs impacts businesses and individuals alike.
The research concludes that applying their model to identify and screen individuals who are likely to develop diabetes years in advance will add 2.06 Quality of Life Years, and would save Parkland alone $130 million. If these models are extended to 10 percent of the population, long term savings could $380 billion, the report reads.
“There is no new technology,” study author Vishal Ahuja says. “You already know how to screen it, this just make your resources a bit more effective.”
The report was authored by Hossein Kamalzadeh, Ph.D. Candidate at theLyle School of Engineering at Southern Methodist University Vishal Ahuja, Ph.D. at the Cox School of Business, Southern Methodist University, Michael Hahsler, Ph.D. Lyle School of Engineering, Southern Methodist University, Dr. Michael Bowen Divisions of General Internal Medicine and Outcomes and Health Services Research. University of Texas Southwestern Medical Center.