Innovative computer programs utilized for disease surveillance and infection control, but ultimate goal is to link with the EMR.
An ICU patient at Texas Children's Hospital comes down with a serious hospital-acquired Serratia marcescens infection shortly after another patient with the same condition is transferred to the room next door.
By traditional epidemiology it would appear the infections are related.
Texas Children's, however, uses microbial genotyping as one method of controlling and tracking hospital-acquired infections.
"In this case, genotyping showed the two organisms to be completely different, proving that transmission between patients had not occurred, and relieving staff of the burden that they may have caused transmission of the organism," says Jeffrey Starke, M.D., infection control officer at the hospital. Malpractice litigation can result if staff negligence is suspected.
Genotyping is just one of a variety of information technology methods emerging to help hospitals meet growing demands for better infection control measures. A 2009 Association for Professionals in Infection Control and Epidemiology survey of 2,000 members showed only 20 percent of U.S. hospitals currently use some form of electronic surveillance or data mining to monitor HAIs.
MultiCare Health System, which operates four hospitals in Tacoma, Wash., uses a mining system that interrogates data across the entire enterprise and includes such departments as microbiology, admissions and finance.
"We're informed of such things as patients testing positive for bacteria in the first 48 hours of admission, indicating the patient likely was admitted with the infection," says Florence Chang, vice president and chief information
Data mining systems also provide actionable data for each nursing unit, without which hospitals must perform targeted manual surveillance, looking for device-related infections, ventilator-associated pneumonias, or central line bacteremias. Sometimes, manual surveillance only targets ICUs.
"Now, we have data that can be used to improve care for every unit," Chang says adding that her organization was able to demonstrate a drastic reduction in catheter-related urinary tract infections over an 18-month period, preventing over 180 infections and saving the organization an estimated $5 million.
The Goal: Integration
One future data mining challenge is integrating infection control methods with the electronic medical record. Currently, most EMR systems are designed for individual patient care, not for aggregation of data from multiple patients. As legislation at both the state and federal level drives further infection control requirements, and as EMRs edge closer to universal adoption, hospitals should "make sure the EMR provides a solution to integrate infection control," says Lior Blik, CIO, Hoboken (N.J.) University Medical Center.
Bothwell Regional Health Center, Sedalia, Mo., is ahead of the curve here. Their EMR pulls specific data, gathers statistics and identifies infection control trends.
"Infection control staff are now able to access patient history from one hospital stay to another, and also see what pre-hospital treatment consisted of, all of which helps the infection control process," says Ja Ellen Hickman, director of resource management at the 180-bed hospital.
Still, until EMRs, genotyping, and data mining get better traction, surveillance remains the gold standard for gathering HAI data. On a national level, the Centers for Disease Control and Prevention provides a widely used Web-based application called National Healthcare Safety Network that 2,600 hospitals in all 50 states use to help with infection surveillance and prevention. Hospitals participate by entering data into the system and using NHSN algorithms to analyze it. The system also lets CDC aggregate data on a national scale.
"NHSN provides national level data about particular types of HAIs by hospital unit or surgical procedure," says Daniel Pollock, M.D., chief of the surveillance branch in CDC's division of health care quality promotion. In this way, hospitals can compare themselves to other sites.
Participation in NHSN is currently voluntary, but that's changing. Pollock says over 20 states are enacting legislation requiring hospitals to submit data through the CDC system to state agencies. Washington is one.
MultiCare's Chang says the organization's commercial infection control data mining program now offers direct upload to NHSN.
"This saves data entry time," she says.