ADC keynote: Big data in ESRD can be used to improve outcomes, but ties to policy may be troublesome
DALLAS — The complexity of treating kidney disease and debate in the early days of the ESRD program regarding concerns about high mortality rates have led to the establishment of several large, publicly accessible databases that can help clinicians assess quality of care, Kevin Erickson, MD, said during his keynote address at the Annual Dialysis Conference. However, using data from these sources to create health care policy has some risks.
“The modern era of big data is using that data to create policy,” Erickson, an assistant professor of medicine and nephrology at Baylor College of Medicine in Houston, said. “But there are problems with that approach.”
Some of these issues include the potential bias that can be formed based on the knowledge gained from data. Likewise, data derived from observational studies vs. randomized, controlled trials may not tell the same story.
“We have seen a number of studies looking at dose of dialysis,” Erickson said. “Providing a higher dose of dialysis shows benefit in some studies, but not in others,” such as in the HEMO trial, which showed patients prescribed a higher dialysis dose than recommended by U.S. guidelines or use of a high-flux membrane did not have a better rate of survival or lower morbidity.
Results of analyses of large datasets are “only as reliable as the data; size (of the data set) cannot address all the problems,” Erickson said.
There is a temptation to use big data to find ways of curbing the high cost associated with treating kidney failure, which consumes 7% of Medicare’s budget yet only covers care for about 1% of the health care plan’s patient population.
That effort, along with the data collected by the two largest dialysis providers in the country – DaVita Kidney Care and Fresenius Medical Care – gives clinicians an opportunity to mine the data, evaluate their practice patterns and make needed changes.
“In the evolution of big data, you will find a lot about ways that big data will change medicine,” Erickson said. – by Mark E. Neumann
Erickson K. Big data! Past limitations and the future. Presented at the Annual Dialysis Conference; March 17, 2019; Dallas.
Disclosures: Erickson reports no relevant financial disclosures.