Improving Immunization Rates Using Lean Six Sigma Processes: Alliance of Independent Academic Medical Centers National Initiative III Project

INTRODUCTION

The public and the medical profession acknowledge that quality and safety in healthcare need improvement.¹ The Institute of Medicine has advocated for interventions¹ and followed up with a strategy for health system and medical education redesign.^2,3 Resident quality improvement (QI) efforts have the potential to improve care quickly and effectively.⁴ Residents are often involved in QI projects, but few programs have a systematic approach for integrating these projects with the hospitals' strategic initiatives.

One national healthcare issue receiving attention is the approximately 50,000 adults who die each year from vaccine-preventable diseases in the United States.⁵ Among these preventable diseases are pneumonia and influenza. Combined, these 2 diseases are the nation's eighth leading cause of death overall and the fifth leading cause of death in older adults.⁶ Pneumococcal infections cause an estimated 5,000 deaths from invasive disease annually in the United States.⁷ All pneumococcal infections, including invasive and noninvasive disease, result in approximately 2.4 million days of hospitalization yearly. The overall case-fatality rate for invasive pneumococcal disease is 10%-18% among adults.⁸ Each year, approximately 226,000 people in the United States are hospitalized with complications from influenza, and 3,000-4,900 die from the disease and its complications.⁹ Up to two-thirds of all deaths attributed to pneumonia and influenza occur in patients who are hospitalized during the flu season regardless of age.¹⁰

A sizable proportion of these cases and deaths is potentially preventable through vaccination. In addition to providing substantial healthcare benefits, such as preventing infections and minimizing hospitalizations, vaccination is also a criterion for Centers for Medicare & Medicaid Services (CMS) reimbursement. Healthcare systems are trying to meet CMS core measures such as proper vaccination and to provide excellent patient care, thus improving their Hospital Consumer Assessment of Healthcare Providers & Systems scores.

Crittenton Hospital Medical Center (CHMC) is a 290-bed hospital located in the Metro Detroit area that provides a full continuum of clinical programs. It is nationally ranked for quality excellence and has a medical staff of nearly 500 physicians who represent a wide range of medical specialties and provide primary, secondary, and tertiary care. CHMC is the primary hospital for 4 residency programs (family medicine, internal medicine, transitional year, and otolaryngology). CHMC and its educational partner, Wayne State University Graduate Medical Education, aligned their educational and clinical initiatives. One of the clinical projects identified was to improve immunization rates against pneumonia and influenza prior to immunization being implemented as a core measure. This project was accepted as a Alliance of Independent Academic Medical Centers (AIAMC) National Initiative (NI) III project.

The AIAMC NI is the first national multiinstitutional effort that focuses on the alignment of medical education with hospital quality and safety strategies. The NI provides critical education, team training, and support to participating hospitals and equips participants with the tools and infrastructure necessary to accomplish meaningful improvements within their home institutions.

CHMC's intended clinical improvement was to immunize 100% of hospitalized patients against pneumococcal and influenza infections. The educational objective was to engage the residents in the family medicine residency program to work in interprofessional teams. The teams focused on understanding their workplace, collecting and presenting data, and proposing interventions to improve care. A curriculum was designed to combine QI knowledge acquisition, team building, and experience-based strategies.

RESULTS

Solutions for the problems encountered and error proofing were analyzed using FMEA. We also documented the effect of the pilot solutions compared to the baseline data. The changes involved input from the process owner/sponsor and had to be implemented in the admission assessment form and communicated to the nursing staff. The Xs (inputs in SIPOC) were the prompts on the adult admission assessment form in need of improvement. An alert was added to Cerner prompting for completion of the assessment prior to patient discharge. Nursing staff were educated on CMS standards and the changes in the assessment form.

We next looked at the FMEA (Table 4) to see what could go wrong at each step, why, and how it could affect performance. The hazard scores were calculated by multiplying the severity of the problem by its detectability and by the probability of it occurring. The actions to reduce failure mode are listed in the table and will be considered in the future.

The potential solutions involved the IT and the Clinical Informatics Resource committees updating the prompts on the current adult admission assessment form, particularly for high-risk patients. The meetings held between the residents and the IT committee focused on revising the form to reflect the current indications per CMS guidelines. The revisions to the adult assessment form fixed 80% of the issues identified in the FMEA. In addition, the recommendation to use the Michigan Care Improvement Registry (MCIR) on all units in the hospital and to report adult immunizations to MCIR is a go-forward project. The new adult assessment form (Figure 4) is also now a part of the Downtime Assessment Paper Toolkit. During EMR downtime, the paper forms will be used and immunization status can still be documented.

• Download figure
• Open in new tab
• Download powerpoint

Figure 4.

Snapshot of the new nursing assessment form for immunization in the electronic medical record.

Strategies used to manage this change included regular interaction with the Clinical Informatics Team and the Cerner Super User group to discuss any trends that needed addressing (ie, any failure of the core measure). In addition, daily monitoring of core measures during implementation, consistently attending unit practice council meetings, and assisting with nursing unit education were deemed important to maintain the goal of 100% vaccination and documentation. The new process capability was calculated after the implementation of the solutions. The process to vaccinate high-risk patients against pneumonia operated at a Z score of 3.9 with a DPMO of 9,346 and a yield of 93.5% at the end of our intervention period. The new pneumonia vaccination process capability is shown in Table 5, and the new immunization compliance rates are graphed in Figure 5.

• Download figure
• Open in new tab
• Download powerpoint

Figure 5.

Postintervention data showing improvement in immunization core measure compliance rates.

Core IMM-1a – Pneumococcal Immunization (PPV23) – Overall Rate.

Core IMM-1b – Pneumococcal Immunization (PPV23) – Age 65 and Older.

Core IMM-1c – Pneumo Immunization (PPV23) – High-Risk Population (6-64 years).

DISCUSSION

We demonstrated that aligning educational objectives with the hospital's strategic initiatives can lead to positive educational outcomes and more efficient care delivery through teamwork with faculty, residents, and hospital staff.

By using the Lean Six Sigma process toolset, we successfully identified the underlying problem for reaching maximal immunization rates and providing appropriate documentation in the hospital. The methods involved using the Lean Six Sigma process tools to assess Z scores, which gave us an idea as to how much work was needed to improve the process. During this phase, we realized that the original process was not addressing our high-risk population. We then put our process through the analysis phase and worked through using SIPOC. This activity helped identify the process steps involved, the problems encountered at each step, and the effect of the problem. Barriers to the process flow were addressed using FMEA. The final PI step was to compare data after implementing the new assessment form to the baseline data. The measures were calculated as part of the core measure abstraction process. This and similar projects can have far-reaching implications in terms of promoting patient health and improving the quality of care delivered by healthcare systems. The financial impact of this project will be more fully realized if global immunization measures are selected as part of future Blue Cross and Blue Shield pay-for-performance or value-based purchasing measures.

The results of the QI projects were presented by the teams to the hospital administration and staff, other residents, and faculty at the first hospital-wide Quality Improvement Day on June 20, 2012. More than 150 people attended. Residents engaged with and led interdisciplinary teams. All team members reported high satisfaction with the structured learning experience. They commented that didactic and experiential learning were powerfully synergistic, and the patient care improvements are motivating to the teams. The success of the initiative was grounded in the interdisciplinary teamwork.

The limitations were that our process solved only 80% of the issues identified in FMEA; this result speaks to the complex nature of PI. Our study encompassed a single end-user change during a single influenza season; it is unknown whether this short-term project will have a lasting effect on hospital flow. Lastly, our process was, by design, limited to the inpatient setting; implementing similar improvements in local primary care practices might have a more meaningful impact. Larger studies are needed to evaluate impact. Educational and organizational outcomes need to be tracked longitudinally.

In relation to the evidence, our findings were consistent with other literature on the subject. The American Health Quality Association¹² produces a fact sheet on QI projects designed to improve quality of care for seniors and lists various participatory organizations that have experienced similar success. Also, researchers at the University of Alberta, Edmonton, analyzed 106 studies—nearly 80% of which were performed in the United States—to map an effective way to increase immunization rates.¹³ The researchers concluded that any of a variety of QI interventions could produce modest increases in vaccination rates, and they observed that 3—team change, patient outreach, and clinician reminders—are effective for improving rates of both influenza and pneumococcal vaccinations.

CONCLUSIONS

The immunization PI project was not only beneficial in terms of patient care but was also a positive learning experience, particularly for the residents. CHMC has undertaken QI projects in the past; however, this event was the first time residents were actively involved and participated in the QI and teamwork training. It was a new and exciting opportunity for residents to learn a new skill set that will be useful in their future endeavors in medicine. The project was challenging because the residents worked on improving a process that is not regularly taught as a part of medical training. Nevertheless, this type of experience can have a significant impact on the quality of care that these physicians deliver.