Aetna Mines Ethnic Health Data

To supply the needed data for the Hartford-based health insurer's new initiatives to track racial and ethnic disparities in healthcare, Aetna's US Quality Algorithms division will use data marts and data mining tools to sort the data for the company's health services researchers.

According to Tina Brown-Stevenson, president, US Quality Algorithms (USQA), at Aetna ($40 billion in assets), managing the expected increase in data volume will not be especially difficult. "Managing healthcare data is like managing grains of sand on a beach," she says. "Taking on a little more ethnic and racial data is not really that hard."

USQA already manages more than seven terabytes of live, online healthcare data, with 30 terabytes of additional live, archived data, says Brown-Stevenson. "The collection process will not be complete until year-end," because many flat data files come from Aetna's TPAs and most of Aetna's insureds still use paper enrollment forms, she says. "We will have the most robust study on racial and ethnic health disparities. Considering that studies with 50 patients are considered good, we will have data on thousands of patients."

And since race and ethnicity of patients do not change over time, unlike the way a person's other medical conditions might change, "the history lets us to do a lot ofanalysis into the past," allowing Aetna to leverage its 37 terabytes of historical data.

USQA primarily uses SAS (Cary, NC) data mining tools to form data marts that researchers use to look for patterns in disease, adds Brown-Stevenson.

Greg MacSweeney is editorial director of InformationWeek Financial Services, whose brands include Wall Street & Technology, Bank Systems & Technology, Advanced Trading, and Insurance & Technology. View Full Bio