Diabetes HBA1C {Poor Control}

Part 1: Introduction

Part 2: Characteristics for Success: Diabetes HbA1c

Part 3: Implementation of Clinical Quality Measure: Diabetes HbA1c

Part 4: Improvement Strategies: Diabetes HbA1c

Part 5: Holding the Gains and Spreading Improvement

Part 6: Supporting Information

Part 3: Implementation of Clinical Quality Measure: Diabetes HbA1c

Before following the steps in Part 3, an organization should first make a commitment to decrease the number of poorly-controlled diabetic patients and complete the initial steps outlined in the previous section that include:

• Developing an aim statement
• Creating an infrastructure for improvement
• Gaining commitments from leadership

Performance on this measure indicates how effectively all the steps of the processes used to deliver care work together so that glycemic control is optimized. Because there are so many factors that can have an impact on glycemic control of patients, it helps to visualize how these steps are mapped. The next section defines Critical Pathway and illustrates the application of this concept to test improvements to improve the HbA1c in poorly-controlled diabetic patients.

The case study continued...

Case Study: The Approach

A critical pathway, also known as a clinical pathway, is a visual depiction of the process steps that result in a particular service or care. The sequence and relationship among the steps are displayed, which reveals a map of the care process. Additional information, including tools and resources regarding the mapping of care processes, can be found in the Redesigning a System of Care to Promote QI module. In an ideal world, the care process is reflective of evidence-based medical guidelines. Evidence-based medicine aims to apply the best available evidence gained from the scientific method for medical decision making. ⁽¹¹⁾ A map of the care process steps that incorporates all of the known evidence and follows respected evidence-based medical guidelines can be considered the idealized critical pathway.

While the needs of individual patients should always be considered, clinical guidelines synthesize the best evidence into a pragmatic set of action steps that strive to provide the optimum health care delivery system. It is important to emphasize that clinical evidence and guidelines will evolve as knowledge progresses; therefore, the idealized critical pathway may evolve over time and not meet the needs of every individual.

In Figure 3.1, the schematic for Critical Pathway for Diabetes HbA1c incorporates available evidence and represents an idealized critical pathway for care to optimize glycemic control. The boxes represent typical steps in care delivery. If these steps happen reliably and well, effective care is delivered.

Walkthrough of the Idealized Critical Pathway

The steps illustrated in the schematic reflect a system that is working well. It is helpful to understand these steps in more detail and how they relate to glycemic control:

1. It is important to know the HbA1c. If an organization follows current clinical guidelines, it needs to ensure that this test is completed (not just ordered) at appropriate intervals depending on the patient's risk.

2. Next, an organization needs to ensure that completed test results are viewed by the correct staff member. In some organizations, all results are routed to the provider. In others, a designated staff member is responsible for reviewing the results as guided by a protocol created by the provider.

3. An organization needs to assess the value of the HbA1c against the goal for the patient. Goals are recommended by clinical guidelines and tailored to the patient's risk and co-morbidity. Regardless of individual variation, a value greater than 9 percent is considered poor glycemic control and is the threshold for the poor control measure.

4a. If the value is not what it should be for any given patient, steps must be taken to lower the HbA1c. There are a number of contributing factors that may cause a value to be high. These can be organized into patient-related, care team-related or system-related factors. Individual patient needs should be addressed to drive the HbA1c down. Systematic implementation of improvement strategies in all three areas reduces the HbA1c for individual patients and decreases the percent of the population served with HbA1c greater than 9 percent.

4b. Patient achieves target HbA1c level. Reinforce the care plan to ensure that good glycemic control continues. Any anticipated challenges should be discussed.

5.5. Interim and follow-up care is then discussed to ensure proper monitoring and that the patient has what is needed to manage his or her care until next seen by the care team. Guidelines are emphasized so the patient understands what screening and examinations are to be done. Appropriate follow-up screening occurs in a timely manner and the cycle repeats.

A quality improvement team benefits from mapping out how care is actually provided. Once it is able to evaluate where there are potential opportunities for improvement, it can use some of the improvement ideas that have worked for others, as outlined in Table 4.2: Sample Changes That Work.

A couple of important notes:

• An organization may adopt additional diabetes guidelines that include important care parameters. The American Diabetes Association Guidelines ⁽¹²⁾ and The American Association of Clinical Endocrinologists/American College of Endocrinology describe guidelines for comprehensive diabetes care. ⁽¹³⁾
• A critical pathway can also be constructed to illustrate how care is currently provided within an organization (the existing pathway). Understanding the gap between an organization’s existing critical pathway (how you provide care now), and the idealized critical pathway (how to provide reliable, evidence-based care aligned with current guidelines) form the basis for improvement efforts.

In addition to understanding the steps for providing care for diabetic patients, factors that interfere with optimal care should be understood. As there may be several of these factors, a QI team may find it helpful to focus its attention on factors that interfere with ideal outcomes. This becomes especially useful as plans are developed to mitigate these factors.

Factors that have an impact on Diabetes HbA1c can be organized into those that are patient-related, relative to the care team, and a result of the health system. Overlaps exist in these categorizations, but it is useful to consider factors that have an impact on care processes from each perspective to avoid overlooking important ones.

Patient factors are characteristics that patients possess, or have control over, that have an impact on care. Examples of patient factors are age, race, diet, and lifestyle choices. Common patient factors may need to be addressed more systematically, such as, a targeted approach to address low health literacy, or a systematic approach to educate staff on the cultural norms of a new refugee population. Examples of how patient factors may influence glycemic control include:

• Age- diabetes is a progressive disease and becomes more challenging to control with age
• Cultural differences- food choices, affinity for physical exercise, and cultural norms about weight vary
• Health literacy- creates barriers in understanding and following a care plan
• Work status- may create care access issues; shift work may influence care plan
• Co-morbid diagnosis- complicates treatment choices and ability to follow the care plan
• Socioeconomic status- may have an impact on access to medications and food choices

Care team factors are controlled by the care team. These types of factors may include care processes, workflows, how staff follows procedures, and how effectively the team works together. Care team factors that may influence Diabetes HbA1c include the processes and procedures that:

• Staff use to outreach to patients to ensure periodic care based on level of risk
• Provide culturally-competent care to address the patient's cultural norms about diabetes care
• Provide planned, comprehensive care for patients who are seen regardless of reason for visit

Health system factors are controlled at the high level of an organization and often involve finance and operational issues. Health system factors that may influence Diabetes HbA1c include:

• Cost-co-pays and access to affordable medications
• Scheduling systems-availability of evening and weekend appointments and wait time may have an impact on access
• Location-no transportation or unsafe location may present barriers to keeping appointments

These factors, when added to the critical pathway, create another dimension to the map as shown in Figure 3.2:

Next, a team may identify specific factors that pertain to the way care is provided for its patients. The team may look at Step 1: HbA1c measured at appropriate interval, and Step 2: Results received and routed to the designated person of the critical pathway. What factors have an impact on how effectively, timely, and reliably Step 2 follows Step 1? It is tempting to consider the first thoughts that come to mind, but teams are best served by systematically thinking through the potential impact of each category. Example 3.1 illustrates a team's output:

 

Example 3.1: A Team’s Brainstorming Session

The team did some quick checking and found that HbA1c tests were ordered appropriately but that patient follow through was erratic. Once the test was done, the results were reliably transmitted electronically and given directly to the provider of record. Using this information, the team brainstorms on factors that would likely have an impact on the arrow (or opportunity) between Steps 1 and 2 of the critical pathway for Diabetes -HbA1c.

Factor Category	Factors pertinent to our organization - Steps 1 and 2
Patient	Patients do not have a clear understanding of the disease and the importance of regularly monitoring their HbA1c levels; patients experience transportation issues
Care Team	No staff, workflows, or prompts dedicated to HbA1c testing frequency; available educational materials are not culturally appropriate for the population; no provider consensus about how frequently to test HbA1c
>Health Systems	Patients needed to have test done at another location and required an additional co-pay; "no news is good news" policy about lab results

The team continues to look at different parts of the pathway to identify relevant impacts for each part. Once it is able to evaluate where there are potential opportunities for improvement, it can use this information to target its efforts. Additional examples of strategies to improve care for the measure, Diabetes HbA1c, are described in the Part 4: Improvement Strategies of this module.

Once the team visualizes the pathway and identifies opportunities for improved care, the next step is to collect and track data to test and document them. First, a QI team needs to determine how to collect data to support its improvement work. This step is essential for understanding the performance of its current care processes, before improvements are applied, and then monitoring its performance over time.

There are three major purposes for maintaining a data infrastructure for quality improvement work:

• To know the starting baseline
• To track and monitor performance as changes are implemented
• To perform systematic analysis and interpretation of data in preparation for action

The first step to creating a data infrastructure for monitoring the performance measure is to determine the baseline. A baseline is the calculation of a measure before a quality improvement project is initiated. It is later used as the basis for comparison as changes are made throughout the improvement process. For the Diabetes HbA1c measure, an organization can determine the percentage of patients with an HbA1c value greater than 9 percent. Performance reflects the current organizational infrastructure and the patient's interactions with existing care processes and the care team.

Baseline data is compared to subsequent data calculated similarly to monitor the impact of quality improvement efforts. The details of how to calculate the data must be determined to ensure that the calculation is accurate and reproducible. The difference between how an organization provides care now (baseline) and how it wants to provide care (aim) is the gap that must be closed by the improvement work.

The next step of data infrastructure development involves a process in place to calculate the measure over time as improvements are tested. A QI team's work is to make changes, and it is prudent to monitor that those changes result in achieving the stated aim. This involves deciding how often to calculate the measure and adhering to the calculation methodology.

Finally, an organization’s data infrastructure must include systematic processes that allow analysis, interpretation, and action on the data collected. Knowledge of performance is insufficient for improvement. It is important for an organization to understand why performance is measured and to predict which changes will decrease the number of poorly-controlled diabetics based on an organization’s specific situation. Collecting data related to specific changes and overall progress related to achieving an organization’s specified aim are important to improvement work. The next section describes in more detail how to develop a data infrastructure to support improvement.

1. Step 1 - Determine and Evaluate the Baseline
   
   As discussed above, a baseline for improvement is a calculation that provides a snapshot of the performance of the systems of care for a measure before improvements are applied. The baseline is determined by calculating the measure and collecting the information for the numerator and denominator.
   
   Determination of a baseline is accomplished by actually calculating the measure and requires that the information for the numerator and denominator be collected. There are several methods to collect this information. While electronic methods are more efficient once established, manual chart audits utilizing random sampling techniques are equally valid.
   
   Consistent data collection sources and methodologies are critical to ensure reliable data. Please note that the tables referenced in this section are from the NQF-Endorsed National Voluntary Consensus Standards for Physician-Focused Ambulatory Care Appendix A- NCQA Measure Technical Specifications (April, 2008 V.7. Pages 20 - 23 and 26 - 28). The methodologies suggested are also from NQF and can be found here.
   
   The following tables and figure depict a decision algorithm for the measure, Diabetes HbA1c. The algorithm outlines the steps that an organization follows to determine its baseline and monitor improvements for Diabetes HbA1c.
   

   Identify the Denominator
   The denominator for this measure is the number of patients aged 18 through 75 years of age with a diagnosis of type 1 or type 2 diabetes mellitus during the measurement year
   a. Use a one-year date range, hereafter called the measurement year. See choosing a date range to audit
   b. Choose a selection method	Pharmacy method-patients who were dispensed insulin or oral hypoglycemics/antihyperglycemics during the measurement year or year prior to the measurement year on an ambulatory basis. [Table CDC-A ] Do not include patients who take metformin in the denominator without another reason to do so. Metformin is used for other conditions as well as to treat diabetes. Claim/Encounter Data-patients who had two face-to-face encounters with a diagnosis of diabetes [Table CDC-B ] on a different date of service
   c. Exclude those who have a diagnosis of polycystic ovaries, steroid induced diabetes, or gestational diabetes but do NOT have a diagnosis of diabetes from the denominator	Exclude patients where the HbA1c value is suspected to be inaccurate. The value of HbA1c needs to be considered in the context of the patient as the assay is not foolproof. Depending on the assay method being used, certain hemoglobinopathies may interfere with results. This problem is highly method-dependent. Inaccurate results may be obtained in the presence of salicylates, chronic alcohol or opiate use, hyperbilirubinemia, liver or renal disease, iron deficiency, vitamin C, vitamin E, hypertriglyceridemia, lead poisoning, and when there are conditions of abnormal red blood cell turnover such as in anemia.

   
   

   Identify the Numerator
   a. Based on an organization's systems, evaluate all of the individuals who remain in the denominator and choose an Electronic Method or the Medical Record Audit method to determine the numerator. For Electronic Method, use electronic data from an Electronic Medical Record or registry to identify the most recent HbA1c test during the measurement year. The patient should be included in the numerator if the: Result of the HbA1c is greater than 9 percent [Table CDC-E ] Most recent test result is missing (even if documentation of test ordered exists) HbA1c test was not done during the measurement year
   b. Medical Record Audit: Audit all patients in the denominator or use valid sampling methodology. The records audited may be electronic or paper. Include the patient in the numerator if the: Result of the HbA1c is greater than 9 percent Most recent test result is missing (even if documentation of test ordered exists) HbA1c test was not done during the measurement year

   
   

   Calculate the Measure
   Divide the numerator by the denominator and multiply by 100 to get the percentage of the diabetic population with poorly-controlled HbA1c. Note: This percentage also includes those whose test results are unknown or not done within the measurement year, both of which require attention in order to improve diabetes management and outcomes.

   Compare an organization's performance to national benchmarks and other available data. The NCQA website [PDF | 3.58MB] updates national and state performance on this measure on an annual basis. Note that there is considerable variation among practices reporting. Other opportunities for comparison data are from payers, state diabetes control programs, state and regional quality improvement organizations, as well as aggregate reports for specific HRSA-funded programs.

   Decide if the performance is satisfactory based on available data from reliable sources. It is important to consider the organizational capacity and constraints, but it is recommended that an organization's aim is high. An organization with a low performance may want to allow a longer time to achieve excellence, but striving to reach an HbA1c value less than 9 percent is feasible for most. If the performance is satisfactory, an organization may wish to choose another measure and focus on other systems of care.
   
   If the performance is unsatisfactory, consider adopting the measure and using it to monitor improvements to the care delivery system. An organization should understand if a measure is adopted for improvement, ongoing and regular measurement is necessary to reach and sustain its organizational goals. More information regarding measurement can be found in the Managing Data for Performance Improvement module.

Evaluate the baseline. Initially, a team compares its baseline to the performance it hopes to achieve. It is important to remember this gap in performance is defined as the difference between how the care processes work now (baseline) and how an organization wants them to work (aim). An organization may often modify its aim or timeline after analyzing its baseline measurement and considering the patient population and organizational constraints.

As an organization moves forward, the baseline is used to monitor and compare improvements in care over time. While it is important for an organization to stay focused on its aim, it is equally significant to periodically celebrate the interim successes.

  2. Step 2 - Create a reliable way to monitor performance over time as improvements are tested. An    
     organization should:

    Standardize its processes and workflows to ensure the team collects and calculates performance data the same   
    way over time. An organization should document exactly how the data is captured so staff turnover does not 
    interfere with the methodology.

1. 1. Determine the frequency performance is calculated. Frequent data collection is often associated with higher levels of improvement. Monthly measurement is recommended, if feasible, as it is associated with a higher level of team engagement and success. If monthly is infeasible, quarterly measurements may be obtained. Less frequent performance measurements are adequate for reporting purposes, but not for supporting improvement efforts. An advanced discussion can be found in the Managing Data for Performance Improvement module.
   2. b. Chart and display results. A simple chart audit form is appropriate for manual audits and can be repeated frequently as desired. Results of multiple audits can be presented in a graphic format to demonstrate trends.
   Note: The frequency of team meetings is not necessarily prescribed for success. Many successful teams meet once a week while others may meet bi-weekly when focusing their improvement efforts on any given measure. Success of these meetings is rather the output of the team members’ active engagement in the meeting and being prepared to report on recent improvement findings. More information, including resources and tools supporting developing and implementing effective team meetings can be found in the Improvement Teams module.

2. Step 3 - Create systematic processes that allow an organization to analyze, interpret, and act on the data collected.

   Having the data is not enough. Improvement work involves thinking about the data and deciding what to do based on that analysis. A QI team needs to put processes in place - team meetings, scheduled reports, and periodic meetings with senior leaders, to use the data tracked. This section describes how a QI team may accomplish the work of creating actionable plans based on the data collected. In Example 3.2: QI at Team Excelsior Health, the scenario illustrates how a team may use these concepts to act on its data.

   1. Analyze: What are the data trends? Tracking performance over time for the measure, Diabetes HbA1c, is critical to successful improvement, but calculation of performance is not enough. It is important for a team to meet to analyze the data on a regular basis. QI teams that are experienced in looking at data recognize these common patterns:
      • Performance is improving
      • Performance is decreasing
      • Performance is flat
      • Performance has no recognizable pattern

      Additional examples of common data patterns are provided with further explanation in the Managing Data for Performance Improvement module. It is typical for a team to see little movement in its data over the first several months. If a team has chosen to monitor an associated process measure, such as, the percent of no-show diabetic patients who are rescheduled, performance improvement may be evident more quickly. Regardless, it is important that a QI team review performance progress regularly. A QI team that meets regularly and calculates performance monthly should spend part of one meeting each month reviewing its progress to date.

   2. Interpret: What do these data trends mean? A QI team needs to then interpret what these data trends mean within the context of its own organization. If performance is increasing, but has not yet reached the numerical aim, perhaps the changes in place are having the desired effect and the aim will be reached over time. If performance is decreasing, what has changed? Are there new care process changes, a failure of registry data input, or a large increase in those patients included in the registry? If performance is flat, did the organization maximize the benefits from changes implemented or was there some regression to the former way of doing things? Improvement trends that have reached a plateau may indicate that an organization needs to think differently about future changes. A few suggestions that an organization may consider when experiencing a plateau in improving are listed below:
      1. Consider looking at outliers to determine barriers to patient access to care for diabetes, for example, lack of insurance, transportation, or language and cultural differences.
      2. Consider changes in a different part of the framework to get improvement back on track. If using a critical pathway approach, an organization may look at the steps prior to where the problem seems to be. If a Care Model approach is used and the team worked hard on delivery system design issues, opportunities to better leverage the clinical information systems or engage the community may be considered.

      Interpretation of data over time is critical in determining where a team will target its efforts. Additional tools that can assist a team in understanding underlying causes for data trends are beyond the scope of this manual but are discussed in detail in a monograph that was published by the NQC, A Modern Paradigm for Improving Healthcare Quality.

   3. Act: Make decisions based on data. Once a QI team has a better understanding of what the data means, efforts should be targeted to further advance the performance toward the aim. Often the decisions are made at the team level about what to tackle first. Then small tests of change can be accomplished to determine what improvements could be implemented to enhance performance. The practice of using small tests of change actually allows multiple changes to be tested simultaneously.
      
      Note: An advanced discussion on how to use the data collected to advance an organization's improvement, including resources and tools to support improvement, can be found in the Managing Data for Performance Improvement module.
      

   Example 3.2: QI Team at Excelsior Health

   ---

   The Quality Improvement (QI) Team at Excelsior Health worked diligently to improve HbA1c levels for its diabetic patients over the last several months. The team focused on patient education, following testing guidelines, and streamlined those processes. But during the last three months, the performance remained the same at 30 percent, which was below its aim of having less than 20 percent of its patients with an HbA1c greater than 9 percent.

   Analysis: The team noted improvement initially. Registry input, care processes and patient volumes seemed to be stable but performance was flat for the last three months.

   The team leader asked for a list of those patients who had an HbA1c ordered but did not have the test completed-outliers for the measure. Further study of these specific cases found that over half of those patients were uninsured.

   Interpretation: Because there was initial improvement followed by several months of flat performance, the team leader looked for obvious changes in processes that would have an impact on performance, but found none. The team leader interpreted the data to mean that initial changes provided some improvement, but not enough to achieve its aim and have the desired impact. More work was needed. The team leader employed a common strategy to find additional opportunities; i.e. he looked at the population not in compliance (the outliers) for a common cause to be addressed. In this case, a common thread was that patients were coming in for care but were not able to follow through with testing.

   This information allowed the team to consider ways to assist uninsured patients with following through on lab testing. They looked at Sample Changes that Worked (Table 4.2) for ideas then added suggestions based on its own patient population. The team decided to increase focus on access to testing. A proposal was submitted to the organization leadership to purchase a machine that would allow it to perform HbA1c testing in the health center. A cost analysis was done that included cost of the machine, materials and staff training, as well as potential revenue. The purchase was approved and systems designed for implementing its use. The improvement team will continue to monitor its performance to determine if this change contributes to achieving its aim statement goals.

   Act: The information gathered from the analysis and interpretation of the data allowed the team to focus its next efforts. Since numerous patients were not following through with testing, the team targeted its efforts on improving access to affordable testing. This enabled the team to focus on PDSAs to test changes specific to these areas and monitor its progress.

   
   A QI team leader needs to monitor the pace of the progress over time. If there is insufficient progress to meet the specified aim, reasons should be analyzed and addressed. One organization may choose to accelerate its improvement efforts; another may decide to extend its initial allotment of time to achieve its aim and consider other constraints within the organization.