This section of the report addresses the question, "What did we learn?" by drawing on the experiences of the National Project Team (NPT) over 4 years, and with interviews conducted with 11 State leads.8 To better understand the facilitators and barriers to the project, two HRET staff conducted semi-structured interviews lasting 60 to 90 minutes each with the State leads of six high-performing and five challenged States. One staff person led the interview, while another recorded and coded responses. High-performing and challenged States were defined by the degree of CLABSI rate reduction or the number of units in a State that sustained a zero CLABSI rate for the last two reporting periods and degree of unit team engagement determined by the number of recruited and retained hospitals, data submission rates, and unit attendance on content and coaching calls. The interview questionnaire was designed to elicit information about a number of topics: why the association joined the initiative; association leadership engagement; association experience with QI collaboratives; association QI infrastructure (staff, qualifications, and QI committees); State consortium stakeholders; educational programming, coaching calls, and overall faculty support; data collection and use; NPT support; key challenges and successes; and sustainability plans. The 40-question interview guide is provided in Appendix A.
1. Have Well-Defined, Evidence-Based Interventions
A national campaign has distinct requirements that do not exist in smaller-scale improvement efforts. Only well-defined interventions with demonstrated significant results can be successful in national efforts where there are limited resources and capacity to execute them. The published success of the MHA Keystone: ICU project contributed significantly to HRET's ability to recruit States to On the CUSP: Stop BSI. The published results demonstrating the impact of the MHA Keystone: ICU project legitimized the On the CUSP: Stop BSI program in two ways: 1) by heightening awareness of the significant morbidity and mortality associated with central line infections, and 2) demonstrating that these infections could be reduced to almost zero. The magnitude of this success in 127 ICUs across the entire State of Michigan helped alleviate the skepticism and fatigue that surround many quality initiatives that have proliferated in recent years. On the CUSP: Stop BSI was adapted closely from the MHA Keystone: ICU project and had well-defined interventions—the CUSP model to address safety culture and the technical change package.
2. Build a Solid Implementation Structure and Project Plan
National Experts with Proven Ability
Having highly credible national experts with proven ability to achieve project goals is a critical element of any successful national quality improvement campaign, and this was certainly the case with On the CUSP: Stop BSI. All State leads interviewed by HRET stated that the national expertise and leadership of members of the NPT was a primary factor in their recruitment to this program.
The results of On the CUSP: Stop BSI demonstrate that MHA Keystone and the Armstrong Institute had developed an education and coaching program that could be scaled up nationally. The ability of MHA Keystone to clearly articulate what worked in their State was also extremely important. Knowing that a peer organization could and did achieve success encouraged other States/regions to try this in their home territory.
Endorsement and Coordination of Key Stakeholders
The NPT communicated frequently with leaders working on the elimination of HAIs at CDC and HHS in order to inform them of the goals and progress of On the CUSP: Stop BSI. National content calls and State lead meetings featured CDC and HHS as faculty to explain data definitions and discuss the HHS national campaign to eliminate HAIs. CDC and HHS representatives also served on the program's TEP which, as mentioned earlier, provided periodic feedback to the NPT on the project's direction, implementation and evaluation. As noted above, States/regions were strongly encouraged to work with State and regional stakeholders to coordinate successful strategies for eliminating CLABSI.
The creation of a national infrastructure to set goals, coordinate efforts, and assure accountability was vital to the successful implementation of On the CUSP: Stop BSI (see "National Project Team" and "Project Stakeholders" sections). As the project progressed, the NPT learned to improve communication channels to ensure effective coordination among all groups. To keep State leads informed, HRET created a weekly update, which State leads could use in turn to communicate with their unit teams. The NPT held monthly State lead calls in which State leads could share their experiences and strategies with their peers.
3. Collect and Use Timely, Accurate, and Actionable Data to Improve Performance
A Web-based data system that units could easily use to upload teamwork and CLABSI rate data and which States could use to generate reports was critical to the success of this initiative. At the start of the project, few hospitals were entering infection data into NHSN so most participants submitted data to the NPT through the Care Counts data repository operated by MHA Keystone. As time progressed, the NPT worked with State hospital associations and CDC so that participant data could be pulled directly from NHSN into Care Counts, avoiding the need for duplicate data entry. Once data was in the system, States/regions viewed data reports generated by the NPT on their monthly coaching calls to monitor progress. The NPT used these reports to track State and national progress. In addition to reporting progress to AHRQ on a semi-annual basis, the NPT was able to use the database to identify units with a rate of 3/1,000 central line days or higher in the last reporting quarter. The NPT provided an action planning kit for States and regions to assist them in doing one-on-one follow up with each of these units to reduce their CLABSI rates. Of the units identified, about half dropped their rates since targeted efforts were implemented.
4. Tailor National Program for Local and Unit Audiences
Hospitals and regions do not all begin improvement efforts with the same level of knowledge, attitudes, and skills, and it became apparent early on in the project that the NPT needed to accommodate these differences. Examples of how the NPT addressed local and special needs included holding Critical Access and Long-term Acute Care affinity group calls, developing a neonatal CLABSI elimination collaborative and holding conference calls on central line maintenance which surpassed line insertion as the biggest opportunity for CLABSI reduction. The NPT also made itself available to States and regions with less regional quality improvement experience, limited staff and/or those without clinical backgrounds. State leads in States with a history of successful quality improvement collaboratives were asked to share their experiences and resources with other States and asked their unit teams to present on national calls. In the case of Puerto Rico, it soon became apparent that Spanish translation services were needed on monthly coaching calls and that the CUSP Manual and CLABSI Elimination Toolkit needed to be translated into Spanish. The "Science of Safety" video was also made available in Spanish to support staff from participating Puerto Rican units.
States also asked for a focus on the neonatal intensive care unit (NICU). In fall 2011, HRET partnered with the Perinatal Quality Collaborative of North Carolina (PQCNC) to leverage existing, State-based neonatal networks to recruit nine States that registered 100 NICUs. Using a slightly different model than On the CUSP: Stop BSI, this national neonatal central line-associated bloodstream infection project, also known as the NCLABSI project, employed neonatologists and clinicians as the State leads. Participating NICUs received CUSP education on a shortened timeframe, a technical bundle that was geared for the neonatal population and submitted their rate and teamwork data into a database developed by PQCNC. A separate, final report on the results of NCLABSI was submitted to AHRQ.
Supplemental funds were provided by AHRQ to States/regions to support local needs. States/regions could apply for these supplemental funds by outlining their unique or local needs and a plan of action. For example, Missouri saw a need to tailor the CUSP curriculum for their hospital units, and Wisconsin decided to make improvements to their State hospital association listserv and database. For their final in-person meeting, Ohio focused the agenda on immuno-suppressed patients and central line maintenance because they saw an opportunity to make large gains in that area. It was important to recognize and financially support States/regions and hospitals that had special content needs and to adapt the project to fit their interests and capabilities.
5. Evolve Project Strategies and Emphases Over Time
Work with Late Adopters
On the CUSP: Stop BSI was a multi-phase effort spanning four years. Over that time, there were changes that required the NPT to adjust its strategies and emphases. The knowledge, skills, and attitudes of the State and regional participants changed over time. For example, the last cohort of States did not consist of early adopters, and they needed additional support. The NPT met frequently to try to address the lower level of engagement of these State leads and their unit teams. And while no particular solution was developed, the NPT did attempt to focus on the "late majority" and "laggards" of the Rogers Innovation Adoption Curve with a "higher intervention technique with more frequent tracking and communication and coaching services."9
Seek New Opportunities to Reduce CLABSI
As rates declined, the NPT, States/regions, and unit teams looked for opportunities to reach zero CLABSIs. This included changing focus on central line insertion to central line maintenance and disseminating guidelines on dressing changes and hub scrubbing. It involved disseminating CUSP and CLABSI elimination interventions to different areas of the hospital such as non-ICUs and hemodialysis units, and focusing on special patient populations such as those who were immuno-suppressed.
Adapt to Changing External Environment
The NPT learned to be flexible in responding to the changing environment at all levels—from the national level to the unit level. This manifested itself in how the NPT advised States on monthly State lead calls and in using faculty from HHS and CDC to present to unit teams and at State lead meetings. States and unit teams worked to adapt, coordinate, and integrate other improvement techniques and tools such as TeamSTEPPS®, Just Culture, Lean, and Six Sigma programs with the CUSP efforts.
In some States, there was competition for hospital recruitment with the State QIO. Many States made the decision to have some hospitals do the QIO CLABSI intervention and others participate in the On the CUSP: Stop BSI initiative. Other States were able to work with their QIO and not compete for hospital unit recruitment.
When CUSP became an explicit component of the 10th Scope of Work for QIOs around the country, members of the NPT met with CMS and developed a short series of national calls with QIOs to explain the CUSP model and make QIOs aware of all of the On the CUSP: Stop BSI resources located on the project Web site.
In States that had health departments with CDC contracts to reduce HAIs, State health department representatives attended and sometimes served as faculty at in-person meetings. By the end of the project, many participating hospitals were also being recruited to join a CMS-funded Hospital Engagement Network (HEN). Because the HENS have received considerable resources to promote improvement in 10 areas, including CLABSI, the NPT has worked to share resources and expertise with HEN contractors to support their CLABSI improvement activities.
8 HRET staff interviewed State leads from Florida, Georgia, Hawaii, Illinois, Maryland, Minnesota, Missouri, Nebraska, Oregon, South Dakota and Wisconsin.
9 Rogers EM. New Product Adoption and Diffusion. Journal of Consumer Research. 1976; 2(4):290-301. Available at: http://www.jstor.org/stable/2488658 accessed on August 1, 2012.