Data mining shows trends in trauma care

The first trauma registries in the United States — developed in the 1970s along with trauma centers and systems — were small, hospital-specific reference databases of seriously injured patients. Today, the American College of Surgeons' National Trauma Data Bank (NTDB) contains detailed data on over 5 million cases from more than 900 participating U.S. trauma centers.

The NTDB and an affiliated data bank, the Trauma Quality Improvement Program (TQIP), use the data to track outcomes, provide feedback to trauma centers and identify processes that improve patient care.

Donald Jenkins, M.D., medical director for the Level I Trauma Center at Mayo Clinic Hospital, Saint Marys Campus in Rochester, Minn., says outcome data helps centers spot trends that may indicate problems. "If the national data report comes back saying that 4 percent of our patients developed pulmonary embolus and nationally only 2 percent did, then we know we have to do a better job. So we look closely at our process and write a protocol or do whatever it takes to bring the numbers down," he explains.

Debra L. Anderson, R.N., the lead trauma registrar at Mayo Clinic in Minnesota and architect of the hospital's trauma registry, agrees. "We have used the TQIP data to provide benchmark rates of complications such as deep vein thrombosis (DVT), pulmonary embolus (PE), pneumonia, surgical site infections and more. While Mayo typically fares well when compared with other trauma centers, we still strive to make improvements. For example, adding a standardized practice guideline for venous thromboembolism chemoprophylaxis resulted in a significant decrease in DVT and PE," she says.

Data is power

Anderson started looking at patient data in 1990. Back then, only patients arriving by medical helicopter were tracked and information on them was limited. Today, she and six full-time abstractors capture data on all admitted emergency department patients as well as all activations that aren't admitted — about 2,000 patients each year. And they now use 200 fields that track everything from demographics and procedures to outcomes, complications, deaths and autopsies.

Like other participating Level I trauma centers, Anderson submits data annually to the NTBD and every quarter to TQIP, which focuses on critically injured patients. She also sends monthly reports to the state. Referring hospitals send registry data to the state, too. Anderson tracks those reports and gives smaller hospitals feedback the government doesn't provide. "If they have questions, they can look at the data — injury severity score, diagnosis, complications — and use it for their own performance improvement activities," she says. "Overall, the data show our referring hospitals are doing a wonderful job stabilizing patients."

The most recent Mayo Clinic data is also impressive. The national mortality rate for trauma patients is 4 to 4.5. At Mayo, it is 2. Still, Dr. Jenkins says, "We are continually looking at where patients are losing their lives. Are we making mistakes in the field? We routinely look at the triage level to make sure it's appropriate to meet patient needs."

Anderson says a robust database combined with ongoing process improvement has made a tremendous difference. "We must be doing something right; our rates keep coming down," she points out. "Right now, there are two or three areas where we feel we can improve care even more. So we're looking at those in-depth."