Collaborative cuts ICU infection rate in half

Rapid cycle changes, checklists were used

A 50% decrease in central-line infections and an increase in compliance with evidence-based practices from 30% to 95%. These are the dramatic results achieved by 10 hospitals participating in the two-year Greater Cincinnati Patient Safety ICU [Intensive Care Unit] Collaborative. The group earned the Joint Commission’s Ernest Amory Codman Award, given for excellence in performance measurement.

Strategies included standardizing the insertion of central lines by using sterile barriers such as gloves, gowns, and full-size bed drapes, using a chlorhexidine antiseptic, and training staff in high-risk units on organizational change methodology.

The 10 hospitals participating in the initiative are Bethesda North Hospital, The Christ Hospital, Cincinnati Children’s Hospital Medical Center, Cincinnati Department of Veterans Affairs Medical Center, The Fort Hamilton Hospital, Good Samaritan Hospital, The Jewish Hospital, Mercy Hospital Mt. Airy, St. Elizabeth Medical Center, and The University Hospital.

"We used a modified Institute for Healthcare Improvement [IHI] model," says Marta Render, MD, the coordinating physician, who developed the Agency for Healthcare Research and Quality project, forming a regional collaborative through the Greater Cincinnati Health Council.

The IHI’s model was modified to include "campaign strategies." The "political" campaign created alliances with leadership across the organization, while the "marketing strategy" focused on selling the change to clinical staff, and a military campaign mapped out timelines, "beachheads," and resources needed to win the campaign.

"The project differs from a usual research project, since if successful, the clinical staff would incorporate these evidence based-practices into their daily approaches forever," says Render.

The IHI model has a role for leadership in initiating the change, but then relies heavily on repeated measurement followed by small tests of change by the front-end staff, Render explains.

Four of the health care systems implemented chlorhexidine and maximal sterile barriers in the first year, while the other five hospitals implemented correct timing of antibiotics acting as a control for the ICU intervention, says Render. In the second year, all hospitals added a second project — either timing of antibiotics in the operating room or implementation of consistent use of sterile gown, gloves, large drape, cap, and mask during insertion of central lines.

The preferred site of insertion of the central line was in the chest or neck, femoral lines when inserted were to be removed in 48 hours, and all central lines were prepped with chlorhexidine rather than betadine.

"Personnel in the ICU were encouraged to use a daily checklist to identify the earliest possible moment when the central line could be removed," says Render.

Here are obstacles and how they were addressed:

• Presently configured kits included supply items that are not recommended for best practices. Items such as small drape and betadine were removed, and quality leaders worked with manufacturers to develop customized kits, says Render.

"The drapes and new kits added some cost to the procedure, which had to be justified to leadership," she says.

The lack of a clinical performance objective for supplies is an obstacle to justifying the use of safer products, Render says. "In the constrained economic hospital environment, a paramount performance objective for supply managers is controlling cost," she says. "The addition of a clinical performance objective for such operations might create a sense of partnership."

• The IHI model was new to staff. "The usual epidemiologic model for change in health care is inefficient and the newer model — rapid action cycles — was unfamiliar. Teaching that method was critical to success," says Render.

Moving from the idea that this was a "study" and short-term intervention for analysis to a "practice" was another challenge, says Render.

Unexpectedly, sharing of data and strategies across health care system boundaries did not prove to be a problem, perhaps because the groups reporting monthly were kept small, usually less than eight people. "Also, the infection control practitioners who were the project leaders had previous experience working together," says Render.

• Time to enter the data, analyze the data and plan the next action was always in short supply. "The teams were magnificent in their ability to put in an extra hour of work when it was needed," says Render. "Inventive and smart nurse managers identified nursing staff that were interested in moving up the nursing ladder, or master’s nursing students who needed an intervention for a focus in order to draw nurses into creating a robust process."

At the Christ Hospital in Cincinnati, there were two goals: reducing central line infections and reducing surgical site infections. For Phase 1 of the project, a "Central Line Insertion Checksheet" was completed by each nurse every time a line was inserted. The nurse checks off whether hands were washed and disinfected, a mask worn, a chloraprep swab used, a sterile glown, gloves worn, full body drape used, and — if the patient had an ultrasound — whether a probe cover was used.

"The project was started in our Medical Intensive Care Unit (MICU) and rolled out housewide after we had worked with the line tray manufacturers and drape manufacturers so we could standardize the protocol throughout the hospital," says Mary Nicholson, RN, BSN, CIC, the project leader and infection control practitioner who led the data collection and analysis efforts.

The checklists are sent to the infection control (IC) department to monitor compliance with the eight indicators and also to identify patients who developed a central line infection.

Each patient record was reviewed by the infection control nurses, and Nicholson then reported compliance and infection rates to the ICU clinical collaborative committee. "The data was also posted in the unit monthly, and celebrations were held to applaud the work being done," she says. "After 12 months, our MICU unit saw a 60% reduction in central line bloodstream infections."

To ensure these gains are sustained, checklists continue to be completed, charts are reviewed by IC nurses, and data are posted quarterly.

In December 2004, Phase 2 of the project began, focused on the timing of antibiotics, beginning with a core group of patient rooms in the OR, using the Plan-Do-Study-Act (PDSA) cycle. A form was created for the same-day surgery nurses to complete, with results tabulated by the IC nurses.

"Through the PDSA cycle, we were able to work through all the kinks in the system before we moved onto other services in the OR," says Nicholson. "The goal was to have the project in all 33 OR rooms by the end of the year. However, we were able to do this much earlier, by September 2005."

Again, data were reported back monthly to the departments and posted for everyone to see. "We are now in the process of automating the reports, so the forms will no longer be used at that point," says Nicholson.

At the participating hospitals, quality professionals were included in the chain of command for reporting the project results. "The role of leader and teacher perfectly suits quality professionals, who could become change agents within their hospitals," says Render. "The systematic implementation of practices to reduce harm to patients and to improve their chances for good outcomes is a very hot area. The need for expertise is expanding exponentially."

[For more information, contact:

Marta L. Render, MD, Cincinnati VA Medical Center, 3200 Vine Street MDP 111 F, Cincinnati, OH 45220. Telephone: (513) 475-6366. Fax: (513) 487-6691. E-mail:]

Mary Nicholson, RN, BSN, CIC, Infection Control Practitioner, Christ Hospital, 2139 Auburn Avenue, Cincinnati, OH 45219. Telephone: (513) 585-1184. E-mail:]