Hospital uses high-tech infection detection
Sunday, May 14, 2006
The Times

Like a skulking intruder, their presence is felt rather than seen. They wreak havoc on their victims and sometimes take an inestimable toll.

But unlike a criminal, these interlopers don't break into your house, they invade your body. They are the germs that breed in hospitals, causing an estimated 2 million infections per year in 5 percent of the patients there.

But infection control specialists at one area hospital, Robert Wood Johnson at Hamilton, have begun to use a new high-tech tool to track infections with an eye toward heading them off before they take hold.

Traditionally a paper-intensive process, infection tracking often has meant poring over stacks of charts to spot patterns. The job was entirely reactive, instead of preventive, and focused on controlling infections rather than pre-empting them.

But by using a new computer surveillance technology -- known as data mining -- administrators are able to establish patterns and pinpoint infections occurring throughout a hospital using existing patient data.

Developed by MedMined, an Alabama-based technology firm specializing in data mining, the tool takes infection control to an entirely new level. By determining where infections arise, administrators can spot faulty practices, incorrect use of new equipment and higher than expected incidences of infection.

The tool automatically identifies the patterns and does not require the user to search for specific information or target an area of interest.

"The service looks for patterns and areas where there might be a problem and alerts the infection control staff. Hospitals are able to reduce infections, length of stay and cost," said Amy Geus, director of marketing for MedMined.

RWJ, which admits about 14,000 patients a year and sees another 53,000 in its emergency room, is one of 10 hospitals in the state to adopt the MedMined technology.

While data on the number of patients who acquire infections at RWJ was unavailable, Healthgrades, a hospital grading firm, ranked the hospital below average in preventing infections from 2002 to 2004.

The MedMined service is paid for by Horizon Blue Cross Blue Shield of New Jersey under a pilot program to reduce the rate -- and cost -- of hospital-acquired infections in 10 hospitals throughout the state. The system also is in place in hospitals in Newark, Somerset, Hackensack and elsewhere.

"The overall costs (of infections) to the health-care system are huge and the impact on the patient is significant," said Bill Finck, director of physician and ancillary networks for Blue Cross/Blue Shield.

The numbers are startling.

About 5 percent of all hospital patients -- about 2 million per year -- will get an infection during their hospital stay. Of those, about 90,000 a year will die, Finck said.

According to MedMined's 2004 testimony before the House Ways and Means Committee, hospital-acquired infections are the fourth-leading cause of death in the United States behind heart disease, cancer and strokes.

In addition, an infection adds an average of about $14,000 to the cost of treatment and can add as much as $40,000 to $50,000, Finck said, costing hospitals, insurers and patients millions each year. Acquiring an infection in the hospital also will add an additional 8.7 days to the average hospital stay.

Because the majority of hospital-acquired infections occur in Medicare or Medicaid patients, who are paid for on a fixed sum basis -- meaning additional infection treatment is often not covered -- hospitals stand to save millions if even some of those infections are prevented.

"We think we can save more than $100 million over the next five years for the hospitals (in New Jersey)," Finck said.

Blue Cross expects to save about $15 million to $20 million over that time, he said.

And there is data to back up those claims.

Studies conducted by MedMined on a group of Alabama hospitals using the service showed a 19 percent reduction in infections after two years of use. That reduction has amounted to millions in savings for the Alabama hospitals.

Officials from the company stress that the numbers for each hospital will vary based on size and other factors, but infections, they say, will be reduced.

Because of the success of the Alabama program, Blue Cross is conducting similar pilot programs in several other states, including California, New York and Texas.

But cost savings are not the only benefit and, some may say, not even the most important.

Used correctly, the program also can spot false trends and faulty procedures. By spotting a pattern of false positives, a hospital can cut down on unnecessary treatment for patients and teach staff members to perform some functions more efficiently and safely.

While it is far from glamorous, Anne Dikon, infection control coordinator at RWJ, uses this example.

Hospitals have a special procedure nurses and doctors are required to follow when collecting urine specimens. And if the specimen is collected incorrectly, it can be contaminated with germs -- possibly showing up as an infection.

The system is able to collate the data showing a spike in urinary infections and clue the researcher that something -- likely the procedure -- is wrong.

By identifying the faulty procedure, the hospital can save the patients from undergoing treatment for a urinary tract infection and train employees to use the proper procedure.

"It allows us to take the information and show it to the people collecting the specimens and ask ourselves, `What can we do to make sure the process is followed every time?' It provides an educational opportunity," said Dikon.

The new system is a big step toward the ultimate goal in her field.

"This is true infection prevention, instead of infection control," she said. "We will now be able to put resources where they are needed much more quickly than before."

Finck tells a more dramatic success story from another hospital using the tool.

The hospital showed an increase in infections in open heart surgery patients. The MedMined service was able to show that all of the patients had come from the same operating room. A search of that room led investigators to a bucket where ice used in surgery was stored.

The researchers learned the bucket had a bacteria in it that was continually infecting patients.

"That's the real advantage," Finck said. "It gives the infection-control professional a tool to react to. As a result, there have been success stories around the country where MedMined has been introduced. We have seen a 10 to 20 percent reduction in rates of infection."

The 10 New Jersey hospitals are in the early stages of use and likely won't have data and results for several months.