Electronic Surveillance is Key to HAI Investigation

Electronic surveillance is becoming a critical tool in an infection preventionist's arsenal of tools with which to fight healthcare-acquired infections (HAIs). ICT presents the following case studies to highlight the benefits of using informatics in infection prevention and control efforts.

TheraDoc

Infection Control Software Facilitates Rapid Analysis and Intervention for Acinetobacter Outbreak

Rhode Island Hospital, a 719-bed facility affiliated with Brown University, has been using infection control software (TheraDoc's Infection Control Assistant) since 2005 for infection surveillance, investigation, prevention and reporting. In 2006, routine surveillance identified hospital-acquired Acinetobacter isolates in two intensive care units (ICUs). Infection prevention staff responded by researching and confirming the presence of an outbreak and developing an infection control plan. The software, with data interfaces to systems such as laboratory, pharmacy, and patient registration, efficiently obtained data, including unit-specific, real-time antibiograms, resulting in rapid analysis and timely intervention.

Following a suspected increase of Acinetobacter isolates in two ICUs, infection preventionists (IPs) at Rhode Island Hospital developed a case definition using infection control software from TheraDoc to efficiently search all positive Acinetobacter cultures obtained in the previous year. The softwares programmed logic allowed for immediate sorting by first positive isolate and "possible hospital-acquired" isolate (positive growth from a culture obtained more than 48 hours after admission). In contrast, review of microbiology results for a second year (prior to the use of the software) had to be reviewed manually. This time-intensive process included requesting reports from microbiology, reviewing patients for possible hospital-acquired Acinetobacter, and manually creating a line list for analysis.

Data collection and analysis for an eight-month period took just three days using the software. In contrast, manual data collection and analysis for the 12-month period using paper microbiology reports took two weeks. "With TheraDoc, it was a huge benefit to have a repository of data readily available to help us determine if there was a problem," says Kerry Blanchard, MT, CIC, infection preventionist at Rhode Island Hospital.

In addition, the system provided real-time, unit-specific antibiograms to aid the investigationimportant since hospital-wide antibiograms, created by the microbiology lab, are only published annually. Antibiogram data were generated and available for analysis within one day, with antibiograms from late 2005 showing increased antibiotic resistance among Acinetobacter isolates in both ICUs.

Armed with data suggesting clonal spread of Acinetobacter, the IPs created an infection control plan, focusing on staff education, compliance with hand hygiene and housekeeping procedures, bathing ICU patients with cutaneous antiseptic agents, and implementing contact precautions. Additional microbiology reports generated by the software updated infection rates and showed the plans success.

According to Julie Jefferson, RN, MPH, CIC, director of the department of epidemiology and infection control at Rhode Island Hospital, and clinical assistant professor of community health at the Warren Alpert Medical School of Brown University, the automated system showed that the antibiogram for Acinetobacter in the two ICUs changed during the outbreak. "Access to real-time antibiograms suggested that something new was introduced," Jefferson says. "After the outbreak was over, the antibiogram reverted back to previous sensitivities. With this tool, we demonstrated that the outbreak was most likely caused by a new organism, and that we eliminated it."

This outbreak investigation confirmed that infection control software can be invaluable not only for routine surveillance, but for investigating suspected infection clusters and outbreaks. The software allowed IPs at Rhode Island Hospital to maintain control of key data, rather than having to rely on other departments. As a result, they were able to:

-- Obtain and analyze data in a more time-efficient manner

-- Create real-time antibiograms for specific organisms in specific units

-- Reduce duplication and increase efficiency

In addition, quickly eradicating the outbreak took on added urgency since one of the units was moving to a new location, and the IPs needed to ensure that the new unit was not contaminated. "The TheraDoc system allowed to us to manage the outbreak, which allowed us to keep the move on schedule," Blanchard says, "thus preventing the introduction of Acinetobacter to the new unit."

Jefferson agrees. "We controlled this outbreak in a matter of weeks. Had we been doing it manually without TheraDoc, we wouldnt have had the quality or quantity of information, and it probably would have taken weeks to months to resolve."

CKM Healthcare

Effective Communication and Collaboration Key to Improved Outcomes

With IPAC Administrator software from CKM Healthcare, a community-based hospital in Southern Ontario was able to quickly identify an outbreak, consolidate and evaluate data, as well as monitor, report and manage an outbreak for continued and improved patient safety.

On Jan. 25, 2010, at approximately 5 p.m., the infection preventionist (IP) on call was notified by the charge nurse from a medical floor that three patients were symptomatic with gastroenteric difficulties. Immediately, all three patients were placed in contact precautions and the unit was closed to further admissions. The index case was a patient whose symptoms had begun six days earlier, on Jan. 19, but since it was only the one patient exhibiting symptoms at the time staff were not overly concerned, attributing this case to something the patient ate. However, between the time the initial case was noted (Jan. 19) and Jan. 29, a total of 13 cases were reported and an outbreak was declared.

Within minutes of declaring the outbreak a line listing of all the symptomatic patients was entered, as suspected cases, in the IPAC Administrator including individual onset time, date and location. Using the IPAC Administrators Outbreak Management interface, a case definition was established to provide continuous unit surveillance. The outbreak management team easily created an automated daily epicurve, using the software, showing the number of new cases, as well as a cumulative record of all cases entered. Customized reports were created throughout the outbreak using the IPAC Administrator and distributed electronically via fax and email to the public health department and outbreak management team for their investigation. When two positive cases of norovirus were returned from cultures sent to the public health lab, a norovirus outbreak was confirmed.

While the IPAC Administrator provided ongoing automatic surveillance, control measures implemented by the staff included:

-- Enhanced cleaning of the department with bleach disinfectant

-- Limiting visitors

-- Closely observing healthcare workers (isolation practices and hand hygiene)

-- Staff and patient education

Because each member of the facilitys outbreak management team had access to the IPAC Administrator software, they were able to monitor the outbreak via real-time updates and with the assistance of the IPAC Administrators Outbreak Management and Contact Tracing tools, no additional cases were identified and the outbreak was resolved within six days.

The IPAC Administrators Outbreak Management tools enable healthcare facilities to automatically receive early warnings of potential outbreaks. Outbreak Line Listings and Epidemic Curves are automatically generated and kept up to date removing duplicate tracking efforts and staff time spent summarizing data results. The software also enables IPs to easily manage all outbreak (pandemic/epidemic/endemic) documentation and protocols including line listings, contact tracing, case definition, costs and control measures. Through the easy-to-use interface, outbreak documentation can be emailed, printed or faxed to provide direct communication to appropriate departments internally as well as off-site, to public health. Another feature of the IPAC Administrator is a real-time lab interface that enables the system to send an e-mail or system alert whenever a predefined threshold or condition (e.g., a positive norovirus culture) is received, enabling the IP to more quickly control an outbreak.

Premier

Catching Clusters That Manual Observation Can Miss

Using SafetySurveillor to complete mandatory state reporting, the director of infection prevention and control at O'Connor Hospital in San Jose, Calif., a facility in the Daughters of Charity Health System, identified a cluster of eight patients over a three-month period on one patient care unit with hospital-onset Clostridium difficile. The cluster would have been missed using manual surveillance. Without SafetySurveillor, it would have been almost impossible to identify cases manually. Most important, cluster recognition prompted immediate intervention to prevent further transmission of C. difficile. Having determined there had been a cluster of patients on a particular unit, infection prevention and control:

-- Formed a C. difficile task force on the nursing unit involved

-- Ordered all rooms on the unit to be thoroughly cleaned with bleach

-- Enforced compliance with proper wearing of gowns and gloves in contact precautions rooms

-- Reinforced proper handwashing techniques

Infection prevention and control also reviewed procedures for SafetySurveillor tags to ensure such issues are caught in the future.

Only one case of hospital-acquired C. difficile was reported on the unit since corrective measures were taken.

"Weve had SafetySurveillor for about 15 months and I just love it," says Suzanne Cistulli, BSN, RN, CIC, director of infection prevention and control at OConnor Hospital. "It helped me identify a problem with C. diff. on a unit that I wouldn't have realized otherwise. We would not have been able to spot that information before SafetySurveillor; we could not have done it manually. . . . One of the things I really like about SafetySurveillor is the alerts. On weekends, for instance, I can get alerts on my BlackBerry, and we can begin looking for these patients and taking proper measures in real time rather than waiting until Monday. I absolutely recommend it to other facilities. I cant imagine doing my work today with out it. Its an absolute necessity."

bioMerieux

Using Informatics for MRSA Containment

In 2008, University Health System, San Antonio, made the decision to partner with bioMérieux to bring the ICNet infection prevention and control platform into its main campus. One of the best examples to illustrate the added efficiency and safety ICNet has brought to UHS is repeat-admission methicillin-resistant Staphylococcus aureus (MRSA) patients. If a previous patient with MRSA is readmitted, this may have been overlooked in the admissions paperwork. The infection preventionist (IP) may not have learned about this for hours, perhaps days. Now, they know immediately and can make sure all proper precautions are implemented.

"Knowing this in real time is incredibly valuable," says Beth Ann Ayala, director of infection prevention for UHS. "Previously, this was next to impossible. Some hospitals try to do this using a tag to the patients medical record number so that readmission of a MRSA patient comes up on the admission screen, but that still doesnt mean the information will reach the IP. Some hospitals actually keep hand-written log books, which are only as good as the staff who perform the hand-written data entry. And some hospitals are starting to screen every patient for MRSA, but they are abandoning all of this historical data that exists in their system because they just cant find it. Its also very expensive to screen every admission. Ive also heard of hospitals hiring 'MRSA nurses' just to track and trace MRSA, but this is incredibly wasteful when so much of the most important data exists within the system."

Aside from the main campus, UHS operates clinics across Bexar County. All lab data from these satellites are pulled into ICNet LabStore database and help inform the IPs when clinic patients are admitted. ICNet provides an automatic tagging process, which is particularly useful when one of its clinics generates a positive MRSA test. When that patient shows up at the hospital which is not uncommon if they need surgery at the wound site the IPs know immediately. This is particularly valuable to UHS because it serves as the Bexar County hospital.

ICNet facilitates the management and surveillance of HAIs at a local level, while enabling regional-level surveillance. It uses laboratory-based data, complemented by unit/ward-level surveillance and links to the hospital information system, thus incorporating alert organisms (e.g., MRSA) detected by the laboratory, alert conditions (e.g., wound infection) detected by clinicians, and alert admissions (e.g., MRSA carrier) detected by the hospital information system, into a single, easy-to-use database.

A case is a period of infection-monitoring interest in a patient. If a patient has an open case, new results are added to the existing case; otherwise, a new case is created. The case is closed when an IP decides that the period of monitoring interest has ended. ICNet can achieve significant time-savings on data entry because a patient's demographic details, ADTs (Admission, Discharge, and Transfer), and organisms are downloaded via electronic interface. The infection prevention team specifies the risk factors and associated features to be collected for action and reporting. The program provides customizable early warning alerts of specific admissions or clusters of results prompting action (e.g., screening or isolation). It provides very rapid outbreak reporting and has been demonstrated to achieve significant time-savings on report creation, as ICNet supports on-demand, customizable data and graphical reporting via a customizable dashboard.

"I rely on it. Its the first thing I look at each morning and its very easy to navigate," says Ayala.

Prior to adopting ICNet, the infection prevention staff would have to wait for all reports to be finalized in order to minimize information overload. With ICNet, they now see preliminary reports, which provide an early alert about an infection or outbreak. The infection prevention staff believes that they now know of critical events, such as MRSA infections, at least 24 hours faster than under the old system. This gives them a huge jump on any clusters, allows them to take action much sooner, and has helped change their perception on the floors. Instead of finding out after-the-fact, they now often deliver the news to clinicians on the floors, as well as initiate the solution.

MedMined

Identifying Trends and Investigating Clusters of VRE and LRE Leads to Changes in Policy and Practice

Preventing and controlling infections from multiple-drug-resistant organisms (MDROs) requires timely identification of outbreaks and emerging resistance, as well as antibiotic stewardship and compliance with best practices. Effective hospital-wide surveillance is key. As demonstrated by the experience at Mobile Infirmary Medical Center (MIMC), a 704-bed hospital in Mobile, Ala., the use of technology is essential.

Joyce Roberson, RN, MSN, CIC, was given the responsibility of managing infection control at MIMC in 1979, when manual data entry and analysis for routine surveillance were time consuming and laborious. "I had reams of data but no way to make sense of it all," Roberson reports. "I needed information that I could act on, hospital-wide, in as close to real time as possible."

In 2005, MIMC implemented an automated, electronic surveillance service (CareFusions MedMined Services, Patient Event Advisor, Data Mining Surveillance Service [DMSS]). Novel data-mining tools and artificial intelligence identify clinically meaningful trends in complex data sets and markers for potential infections.

"The customized alerts inform us right away," Roberson says. "Before, there was always a time lag between when something happened (trends) and when we might be able to identify it or possibly miss it in the midst of all the data." DMSS provides monthly trends that can include possibly inappropriate antibiotic prescribing. These data-based reports provide actionable information, as well as recommendations for reinforcing evidence-based practices proven to be effective in helping reduce the risk of infection.

MIMCs laboratory, admission, discharge and transfer data are sent to DMSS around the clock (24/7). NIMS (nosocomial Infection markers scorecards) are updated weekly and provide hospital-wide rates that are unit- and site-specific. Monthly pattern reports can reveal clusters that may have been missed by routine surveillance.

From April 2009 to January 2010 DMSS reports showed continuing increases in patterns associated with vancomycin-resistant Enterococcus (VRE). In January, while investigating the trend, Roberson identified an even bigger problem.

"The antibiograms for cultures identified by DMSS showed that some VRE was also linezolid-resistant," Roberson says. "I asked MedMined Services to assist me in developing a report to detect vancomycin and linezolid resistance. That report showed increasing linezolid resistance at our facility." Further investigation using the system showed that from summer 2007 to winter 2009 vancomycin use had doubled and linezolid use had more than tripled.

"Our pharmacy was already stretched to the limit," Roberson says, "but the DMSS data documented the urgent need for a medication utilization evaluation (MUE) for linezolid." The MUE results showed inappropriate use of linezolid by some physicians that was undoubtedly contributing to the increasing resistance to this agent. "To preserve linezolid as an effective antibiotic, it was imperative that inappropriate use be stopped," Roberson adds. The DMSS reports and MUE findings were then sent to the pharmacy and therapeutics and infection control committees.

Based on the DMSS data and recommendations, Roberson reinforced with hospital staff the need for compliance with best practices for infection prevention and control. Particular emphasis was given to the recommendations noted by DMSS and to the patient care areas associated with VRE or LRE clusters.

Based on both DMSS and MUE data, the P&T and infection control committees recommended to the medical staff that linezolid ordering be restricted to infectious disease and critical care specialists. Any other physician ordering the agent would be informed that linezolid could not be used or that an infectious disease consult must be obtained.

The recommended policy change, along with the data, is progressing through the necessary committees for approval and implementation. Initial follow-up data on linezolid usage and VRE and LRE patterns are encouraging, and surveillance continues to help assess the long-term impact of policy and practice changes.

At MIMC the ongoing use of CareFusions MedMined automated data mining surveillance service helped identify an unsuspected problem with LRE, as well as continuing increases in VRE. Roberson says, "DMSS allows me to conduct timely, effective hospital-wide surveillance without the burden of manual data-entering and traditional analysis. I can now spend more time reinforcing best practices, educating staff and monitoring outcomes."