By Wendy Hess and William Finck
The financial and clinical costs associated with hospital-acquired infections are becoming an increasing concern for hospitals and payers alike. Studies show that each year, approximately 2 million patients admitted to U.S. hospitals acquire an infection while hospitalized, accounting for about half of all major hospital complications. The financial impact of the problem is significant — the 4 percent of patients who acquire an infection while hospitalized erode as much as 185 percent of hospitals' net inpatient operating profits.
In the state of New Jersey alone, patients with hospital-acquired infections require an average additional seven-and-a-half days of hospital treatment, and the average additional cost to treat a hospital-acquired infection (HAI) is approximately $12,500. The cost of treating HAIs exceeds the reimbursement provided by most payers resulting in a net loss for the hospital.
The reimbursement provided comes out of the pockets of healthcare payers and consumers. One study, across 13 hospitals, which compared the financial outcomes of similarly ill patients with and without an infection, found that while the hospitals lost a combined $56.6 million on patients with hospital infections, payers carried an additional burden of:
Medicare: $23.3 million
Medicaid: $33 million
Commercial: $50.9 million
In New Jersey and across the United States, the existing system for identifying hospital-acquired infections can be slow and resource consuming. Because there are thousands of patient-care processes occurring throughout the hospital every day with infection risks (surgery, drawing blood, cleaning equipment, etc.), and billions of data permutations, it is a daunting task for infection control professionals (ICPs) to focus prevention efforts where and when they are most needed. In most hospitals, infection surveillance is targeted to high-risk areas where the most dangerous infections tend to emerge, for example, the ICU. On a daily basis, ICPs review lab reports and cross check positive lab results and patient medical records to identify issues that may require investigation and further action. This system helps identify hospital-acquired infections but rarely helps to prevent future infections or address current outbreaks in a timely fashion. In short, the system is mainly outcomes focused and leaves ICPs little time to investigate processes that, when not performed correctly, may result in infections. This limited perspective fails to provide ICPs with the tools they need to "prevent " infections.
Recognizing the need for a system to support hospital infection control, Horizon Blue Cross Blue Shield of New Jersey established the New Jersey Infection Prevention Partnership, or NJIPP, with 11 hospitals across the state to reduce hospital-acquired infections.
Above all, we wanted to provide hospitals in the NJIPP with a solution that was actionable and could help avert further infections. The ideal solution would provide real-time data so clinicians could focus their efforts on those areas where they can have the largest impact. Most importantly, the system had to provide clinicians with the right information at the right time in order to impact patient care and outcomes. Horizon Blue Cross Blue Shield reviewed several available tools before choosing MedMined, a service provided by San Diego-based Cardinal Health, combining data mining technologies, clinical expert consultation, educational support, and clinical and financial outcomes measurement to reduce the incidence of hospital-acquired infections. MedMined's Data Mining Surveillance service has been adopted in more than 200 hospitals.
The service uses an artificial intelligence technology to monitor an entire hospital and its outpatient population for the incidence of infections — both community and hospital-acquired. The service alerts infection control staff to processes causing increased infection risk so improvements can be made to prevent future infections. Hospitals using this service also have electronic access to their infection-related clinical data in real time to facilitate reporting, charting and automated alerts.
At the center of the service is the Nosocomial Infection Marker, or NIM, an electronic marker for hospital-acquired infections. The NIM provides an electronic measurement of the incidence of hospital-acquired, or nosocomial, infections. The NIM uses algorithms to analyze existing patient clinical data to identify hospital-acquired infections and compute rates of infection. The NIM also provides a repeatable and efficient method for hospitals to track infections of all types and in all locations, as well as to meet mandatory reporting requirements.
Hackensack University Medical Center (HUMC) was selected to participate in the NJIPP pilot demonstration program and received access to the service in October 2005. Prior to using MedMined, our surveillance efforts focused mostly on the ICU and device-related infections. Utilizing the MedMined service, our ICPs are able to compare NIM rates across multiple units and in a wide array of categories, including wound site infections, urinary tract infections, respiratory infections, and many others. With this information, the ICP team can determine how various units are performing, help them to identify where problems are occurring, and provide them with actionable information to prevent future occurrences. The MedMined data gives us a means to prioritize efforts. ICPs now know where to focus their attention to have the greatest impact.
In the early utilization of the MedMined service, HUMC ran NIM scorecards for each inpatient unit and the hospital overall and discovered the most common type of NIMs were urinary tract infections. To determine where to focus, ICPs decided to perform a staff knowledge assessment. Results showed additional staff education was needed on the process for inserting and maintaining a Foley catheter. As a result of these efforts, we have seen the number of urinary NIMs decrease dramatically.
The previous system of surveillance would not have been able to reveal the intricacies of this problem. ICPs wouldn't have been able to identify a specific source to focus a manual system. MedMined has helped our ICP team identify numerous unit-specific issues. Time which was previously spent on data review is now used to review processes that lead to infection prevention.
Outbreak management is critical. MedMined identified the existence of a recent outbreak and allowed infection control staff to identify the index case on the specified unit. We entered the information for the infection we were looking for, according to pathology and time parameters, and MedMined provided a list of all the patients who were affected on the specified unit. The sheer enormity of the data required to perform the outbreak investigation would have made this task impossible using manual surveillance.
Another benefit of the service is how it has enhanced the role of the ICP at HUMC. The opportunity to provide clinicians with actionable data has allowed professionals to have a much greater impact on infection rates. Being able to say to caregivers "Here's the problem and here's how we can fix it," has changed the perception of infection control at HUMC. The infection control staff now spends a great deal of time interacting with the nurses and doctors on a daily basis.
In September of 2005, when Horizon Blue Cross Blue Shield launched NJIPP, we sought to repeat the success of other hospitals participating in similar initiatives. Our goal was a modest single-digit infection reduction rate. The first outcomes analysis of the initiative was performed after approximately nine months of utilization. At that time, Horizon BCBS saw a 10.31 percent decline in the infection rate within the participating hospitals. With a 10.31 percent reduction by June of 2006 — nine months into the program — HUMC had averted more than 150 infections, which resulted in significant cost savings. Based on the early successes of the initiative, we plan to double the size of the program in 2007 with 20 hospitals statewide. Our ultimate goal is to enroll every healthcare facility in the state.
Wendy Hess is director of infection control at Hackensack University Medical Center, and William Finck is director of network initiatives at Horizon Blue Cross Blue Shield of New Jersey.