Marie Smith, PharmD

The provision of high-quality healthcare is a priority for US health plans and the CMS. Health plans can implement quality initiatives to ensure appropriate healthcare for their beneficiaries and use quality measures in their various initiatives to assess quality of care and outcomes.

 We reviewed publicly reported diabetes quality measures for Medicaid fee-for-service (FFS) programs in all 50 states and the District of Columbia. Our objective was to provide insights into the extent of diabetes quality measure reporting (ie, which diabetes quality measures are being reported and what percentage of states are using the diabetes Healthcare Effectiveness Data and Information Set [HEDIS] measures) and into the quality of care being provided (ie, achievement of established goals in diabetes care and state-level performance on diabetes quality measures).

 Overall, the number of Medicaid FFS programs reporting these data was low: 23 states reported data on at least 1 HEDIS measure. The most common measures reported were process measures (eg, retinal eye exams and HbA1c testing); outcomes measures such as the proportion of patients achieving goal were much less frequently available. Few states had historical data available, which limits the opportunity for retrospective analyses or quality benchmarking.

 This research is important, as an understanding of the current landscape of quality measures for the Medicaid diabetes population is critical to designing and implementing initiatives aimed at improving the quality of care for individuals living with diabetes.

The provision of high-quality healthcare is a priority for the United States, in large part as a result of the Affordable Care Act (ACA) and its focus on 3 goals to improve the experience of care, improve the health of populations, and reduce the per capita costs of healthcare. These triple aims also guide the National Quality Strategy that was established as part of the ACA, which describes the national priorities and strategic plan for quality improvement in healthcare. 

This reformative health policy environment has spurred many organizations to call for methods to measure, track, and improve the provision of healthcare and engage in the nationwide conversation regarding the definition of “value.”

Medicaid is responsible for providing healthcare coverage to more than 72 million beneficiaries.

 The population served by Medicaid will continue to expand under the ACA. The CMS has been a major forerunner in implementing quality healthcare measures and has implemented ongoing initiatives to ensure quality healthcare for both Medicare and Medicaid beneficiaries through accountability and public disclosure.

The ACA required the implementation of a core set of healthcare quality measures for adults participating in Medicaid. The initial set of measures was released in 2013.

 Reporting by states is voluntary; however, many states now require Medicaid managed care plans to report quality measures using the Healthcare Effectiveness Data and Information Set (HEDIS) tool. Although Medicaid is shifting away from a traditional fee-for-service (FFS) model, a substantial number of complex beneficiaries remain within this structure (eg, those with multiple chronic conditions and high resource utilization) and account for a disproportionate amount of total Medicaid expenditures.

 As such, we sought to explore the availability and breadth of data from quality measurement programs that traditional FFS Medicaid plans and Medicaid managed care organizations (MCOs) have instituted for patients with diabetes.

Diabetes-related quality data were obtained as of July 2015 through a Web-based search of publicly reported data, such as state Medicaid and related websites (list of websites provided in Table 1). This research included a review of all 50 states’ Medicaid FFS programs, the District of Columbia, and the largest Medicaid MCO in the 10 states with the highest Medicaid enrollment through May 2015: California, New York, Texas, Florida, Illinois, Ohio, Pennsylvania, Michigan, Georgia, and New Jersey. For purposes of this research, a Medicaid MCO was defined as a contracted entity that administers Medicaid benefits on behalf of a state. Because of the inconsistencies observed in the degree and time frames with which Medicaid FFS plans reported their diabetes-related quality data on public websites, the investigators reached out personally to a 10% sample of states to validate the completeness of data that were captured after the initial Web-based research was completed. Medical and quality program directors at 5 state Medicaid FFS programs (California, New York, Florida, Ohio, and Michigan) were contacted via phone and/or e-mail by the research team. For each state, a director of policy and quality, or similar role, was sought to provide confirmation on the data. These 5 states were chosen for the sample because they include the greatest number of covered lives under their FFS models.

The extent of diabetes-related quality measures found included mostly HEDIS Comprehensive Diabetes Care indicators (both current and previous measures), such as:

Blood pressure (BP) control (<140/90 mmHg)

Low-density lipoprotein cholesterol (LDL-C) screening

Other non-HEDIS diabetes-related quality measures were also identified and documented when found. When provided, the percentage of Medicaid patients meeting each quality indicator in each program was captured. Quality data were analyzed by examining the states’ individually published information for trends in Medicaid program reporting and performance while also noting those Medicaid plans that did not report any data.

All 50 states and Washington, DC, reported using HEDIS as their quality reporting system for either Medicaid FFS or MCO patients. For FFS programs, publicly reported data on diabetes quality measures were poor, with only 23 states (45%) having published data on at least 1 HEDIS measure for their FFS Medicaid lives or total Medicaid lives (Figure 1). We also found that the data-reporting period (ie, the date of the data available on the websites) varied widely, from as early as 2002 through March 2015. Only 6 states reported data from the previous year (2014; Colorado, Illinois, Tennessee, Florida, Vermont, and Connecticut). Only Colorado, Florida, and Vermont reported 7 or more HEDIS diabetes quality measures. Other non-HEDIS diabetes measures identified included hospital admissions rate for short-term complications associated with diabetes, hospital admissions rate for long-term complications associated with diabetes, proportion of patients with microalbuminuria, proportion of patients receiving screening for microalbuminuria, proportion of patients classified as obese or overweight, receipt of influenza vaccine in adults aged 50 to 64 years, health literacy, tobacco use, and composite measure of HbA1c, low-density lipoprotein cholesterol (LDL-C), blood pressure, and tobacco use.

Of the 23 states with publicly reported data for the FFS programs, the most commonly reported process measure was “percentage of patients receiving HbA1c testing” (21 of 23 states), followed by the other screening measures: LDL-C screening (19 of 23 states) and eye examination (12 of 23 states). The least frequently reported measures were those that require states to conduct medical record reviews or additional data analysis (eg, outcome measures such as percentage of patients achieving HbA1c <7%, proportion of patients achieving BP control, and proportion of patients achieving LDL-C <100 mg/dL).

 Moreover, few states had published consecutive data year over year that would enable examination of changes in reporting following Medicaid expansion as a result of the ACA, and only 5 states had historical FFS program quality data available (Colorado [2013], Illinois [2013], Kansas [2003-2008], North Carolina [2010], and Vermont [2012]). It is worth noting that the 5 states that had historical data generally reported more measures (ie, at least 3 measures: Colorado [9 measures], Illinois [3 measures], Kansas [4 measures], North Carolina [3 measures], and Vermont [10 measures]). As shown in Figure 1, the overall quality of diabetes care was low, although generalizations should be made with caution because of the small number of states with publicly reported data and the variance in the age of the reported data. In comparison, however, national benchmark data from the National Committee for Quality Assurance, the organization responsible for the development and maintenance of the HEDIS measures, shows that the proportion of members receiving HbA1c testing was 94% in 2014; the percentage of members with poor control (HbA1c >9%) was 19%.

Contrary to FFS programs, in the assessment of the 10 Medicaid MCOs, they all used the HEDIS core set of measures and had more recent data reporting time frames (all but 1 MCO reported data from 2013 or 2014). The most commonly reported measures were the process (screening) measures: HbA1c testing, eye examination, and LDL-C screening.

The degree and timing of reporting diabetes quality measures differed significantly among states, and historical data were found to be scarce. There was sometimes a significant lag in the time from measurement to reporting of data (eg, 3 years from time of data measurement to report date in Wyoming, Louisiana, and Arizona). Overall, states did not consistently report quality measures on an annual or other regular interval. Moreover, a significant difference exists between Medicaid FFS program populations and MCO populations in terms of their available data. Despite many states’ transitioning their Medicaid population from FFS plans to MCO plans that monitor performance via well-established HEDIS quality measures, the current state of reporting for Medicaid FFS programs remains inadequate. The issue of poor or low levels of public reporting of quality measures is an important one. One of the strategic goals and objectives of the US Department of Health and Human Services (HHS) plan for 2014 through 2018 is to strengthen and improve healthcare quality and patient safety.

 However, improvements can be recognized only if the measures are standardized across the various stakeholders and organizations and the performance data on quality measures are shared regularly. It is not known whether the states identified in this report were influenced by the ACA to report quality measures; although the ACA states that the HHS must publish and periodically update the Medicaid Adult Core Set, participation and reporting by Medicaid programs is voluntary. Although only 4 states had public reports that predated the availability of the Medicaid Adult Core Set (2012), it is not known whether the remaining 19 states that reported ≥1 measure began reporting because of this or whether the timing was coincidental.

There are ongoing initiatives that seek to enhance reporting within various Medicaid populations and more, such as the CMS Innovation Center’s State Innovation Models (SIMs) and other state-level multipayer initiatives, which, perhaps, indicates that changes are on the horizon. In this assessment, 40 states were identified as either participating in SIMs or otherwise establishing larger statewide quality initiatives, with 22 of them tracking at least 1 diabetes-related measure. Because funding for these plans began in April 2013, these initiatives are still in their infancy, with only baseline outcomes data available for review at the time of publication. However, it is anticipated that the level of visibility for progress in diabetes care will continue to increase as states continue to leverage the availability of these funding awards to test innovative health system management approaches.

With the continued expansion of Medicaid and the increasing number of eligible beneficiaries, continuous quality improvement and the impact on outcomes for patients will remain a top priority in the US healthcare system. The recognition of improvement, however, will begin only with the establishment of quality-driven benchmarks and consistent reporting within Medicaid programs.

: Robert Bailey and Janice Lopez are employees of Janssen Scientific Affairs. Marie Smith is a paid consultant to Janssen Scientific Affairs. Herman Chen is an employee of Xcenda, which received funding from Janssen Scientific Affairs to conduct this research.

: Jessica Black, Xcenda, for research and Evelyn Sarnes, Xcenda, for manuscript development assistance.

: Concept and design (JMSL, RB, HC); acquisition of data (HC); analysis and interpretation of data (HC, JMSL, RB, MS); drafting of the manuscript (HC, MS, RB, JMSL); critical revision of the manuscript for important intellectual content (HC, JMSL, RB, MS); statistical analysis (HC); provision of patients or study materials (HC); obtaining funding (JMSL, RB); administrative, technical, or logistic support (HC); and supervision (JMSL, MS, RB).

1. Centers for Medicare & Medicaid Services. Medicaid & CHIP: August 2015 monthly applications, eligibility determinations and enrollment report. https://www.medicaid.gov/medicaid/program-information/medicaid-and-chip-enrollment-data/monthly-reports/index.html. Published October 26, 2015. Accessed November 4, 2015.

2. National Quality Forum; Measure Applications Partnership. Strengthening the core set of healthcare quality measures for adults enrolled in Medicaid, 2014. Final report. http://www.qualityforum.org/WorkArea/linkit.aspx?LinkIdentifier=id&ItemID=78283. Published August 29, 2014. Accessed November 5, 2015. 

3. The Henry J. Kaiser Family Foundation. Medicaid and managed care: key data, trends, and issues. https://kaiserfamilyfoundation.files.wordpress.com/2012/

02/8046-02.pdf. Published February 2012. Accessed November 4, 2015.

4. US Department of Health & Human Services. 2014 annual report on the quality of health care for adults enrolled in Medicaid. https://www.medicaid.gov/medicaid/quality-of-care/downloads/2014-adult-sec-rept.pdf. Published November 2014. Accessed November 5, 2015.

5. National Committee for Quality Assurance. 2014 accreditation benchmarks and thresholds — mid-year update. http://ww2.ncqa.org/Portals/0/PolicyUpdates/Trending%20and%20Benchmarks/2014_BENCHMARKSANDTHRESHOLDS.pdf. Accessed November 3, 2015.

6. US Department of Health & Human Services. Strategic goal 1: strengthen health care. http://www.hhs.gov/about/strategic-plan/strategic-goal-1/index.html. Updated February 9, 2016. Accessed November 3, 2015.

Articles

In a review of publicly reported diabetes quality measures for Medicaid fee-for-service, overall reporting was low and reports of outcomes measures were rarely available.

Low Reporting of Medicaid Diabetes Quality Measures: Room for Improvement?

Since the publication of the report “Crossing the Quality Chasm” by the Institute of Medicine,

hundreds of new healthcare quality measures have been developed at the national, state, health plan, system, and practitioner levels to improve healthcare and promote accountability for better healthcare processes and outcomes.

In the United States, medication-related quality measures are included in various measure sets, including Healthcare Effectiveness Data and Information Set, Electronic Health Record Meaningful Use, accountable care organizations (ACOs), and the Pharmacy Quality Alliance (PQA). Most, however, relate to prescribing processes, adherence rates, or electronic health record capabilities (

). While these measures may lead to improved prescribing and adherence, they do not address medication optimization for effectiveness and safety, medication coordination across multiple prescribers and pharmacies, and medication follow-up required between care transitions or office visits.

Beyond the administrative burden, the sheer volume of current measures makes it difficult to identify the right measures for assessing improvement in medication management that encompass appropriate prescribing, effective dosing, avoiding medication interactions and adverse events, and improving adherence. While there has been keen interest in focusing on patient outcome measures (eg, blood pressure or diabetes control), there is still a need for meaningful process measures for medication management and monitoring

—especially in new team-based care delivery models.

Although hundreds of new quality measures exist today, many “care gaps” still do not have measures. One approach to measure development has been to identify measure concepts for “measures that matter.” Some examples are the need for measures that help to:

		Differentiate provider performance: clinical processes, effectiveness, and diagnostic/treatment accuracy

		Identify patient safety improvements and all-cause harm causes

		Track care coordination and transitions

		Monitor practice transformation progress

		Document trusted patient-provider relationships.

There is ample room for a more deliberate prioritization of measure development to close gaps in care that involve medication therapy decision making and management processes. As new care delivery models evolve to focus on value, accountability, and team-based care, many care gaps mentioned above remain unaddressed. These important care gap measures can be more challenging to develop; once developed, consistent measuring that also adjusts for differences in patient populations or practice settings can remain challenging.

In 2011, the National Priorities Partnership and more than 50 public- and private-sector organizations worked with the HHS to release the first report from the National Quality Forum (NQF): “2012 NQF Measure Gap Analysis.”

 This NQF report includes 6 priority areas, each with aspirational goals and specific targets around which to focus public- and private-sector performance measurement and improvement: health and well-being, prevention and treatment of leading causes of mortality, person and family-centered care, patient safety, effective communication and care coordination, and affordable care. More work is needed to advance the nation’s measurement capabilities across all priority areas—especially in the gaps identified by the NQF report. 

 outlines some medication management solutions that address these gaps.

Measure development that identifies the extent to which comprehensive medication reviews are implemented at primary care visits and care transitions can address many of the gaps that were identified. When a practitioner conducts a comprehensive medication management review, the patient’s narrative is valued and incorporated into their care planning—including such information as which medications they are actually taking, the impact of medication experiences on healthcare beliefs, the challenges they face with medication access or adverse events, and the outcomes they have experienced.

In addition, we need to measure the extent to which: 1) patients who are at high risk for medication problems are being identified in population health data or referred by healthcare personnel for comprehensive medication reviews; 2) medication optimization recommendations based on evidence-based guidelines are sent to and accepted by prescribers; and 3) culturally and linguistically appropriate patient medication action plans are being developed to address patient self-management goals for effective medication use and safety.

There are several ways to categorize quality measurement. One way is to look, in combination, at the purpose of measurement, how measures will be used, and who will use them. Measures can be used to: 1) provide information to consumers in public reporting; 2) information for payment (as in pay-for-performance or shared savings programs); 3) compel quality improvement through requirements from an external source; and 4) improve internal quality improvement programs.

Another way to categorize quality measurement is by what is being measured. Donabedian suggests using 3 categories to evaluate medical care, each referring to a component in the provision of patient care: structure, process, and outcomes.

		care is provided and whether specific desired components are present. Examples include whether certain qualifications are met (licensing or certification), whether certain technology is available, or whether policies and procedures are defined and used. Such measures utilize specific structural components that have demonstrated improvement in health outcomes in scientific research.

 are useful when a specific service or action provided to the patient is strongly associated with improving patient outcomes. Measurement of care processes is the most common type of measurement.

Examples include whether a guideline is met or if identified patients receive recommended medication for a specific condition.

		Outcome measures are often the best indicator of the quality of medical care. Outcomes can be as clearly defined as death or survival. Others include clinical end points such as blood pressure measures, patient-related assessment of pain, and patient care experiences. Outcome measures are often more difficult and complicated to measure, and can be influenced by factors not controlled by the healthcare provider. An example of an outcome measure is the 30-day readmission rate following hospital discharge.

We need to ensure that medication-use measures improve care and lead to better outcomes by addressing medication optimization for effectiveness and safety, and coordination across multiple prescribers, sites, and pharmacies. A measure value set is a grouping of measures that together can better assess the value (quality and cost) of care than any single measure can.

Some new approaches to measure set development and usefulness are discussed below.

One new approach, introduced by the NQF Measure Applications Partnership (MAP),

is “measure families,” groups of measures related to the same condition, patient outcome, or healthcare scenario. The MAP has defined 10 families of measures assessing all parts of the NQF: cancer care, cardiovascular disease, care coordination, diabetes, dual-eligible beneficiaries, hospice care, patient safety, population health, patient- and family-centered care, and affordability. The great majority of existing measures fit into these families and are being used in public or private programs. 

 depicts a family of measures for medication safety (a subset of patient safety).

Families of measures are intended to promote alignment. Increased alignment of performance measures for healthcare delivery may provide substantial benefits, including: 1) increased clarity on the highest priority areas; 2) reduced confusion in interpreting the results of similar, yet slightly different measures; and 3) decreased burden associated with data collection and reporting for various measures addressing similar topics.

Measure families are useful tools for alignment because they can be cross-cutting in nature, with each family including measures that span other families and settings of care. Cross-cutting measures can be used to measure the outcomes of several different conditions, patient-reported therapy outcomes, or patient experiences with the healthcare system. 

 includes cross-cutting measures for care coordination and care effectiveness.

The measures of medication use are applied across conditions and sites of care. For example, a medication selection for care transitions from hospital to home impacts effectiveness and affordability, and improves population health. Medication adherence also impacts effectiveness and affordability, and offers opportunities for person- and family-centered care shared decisions.

A layered measurement approach recognizes that different measures are needed for providers, organizations, and external payment and reporting purposes. One set of measures is applied for providers’ internal quality improvement, another related set for internal organization management, and a third related set for public reporting and payment (

). Healthcare organizations will eventually be required to report common population-based performance measures, and can choose whichever metrics serve their purposes for internal improvement. For example, hypoglycemiarelated events may be required for care of patients with diabetes, whereas measures of hospitalizations and emergency department visits may be mandated for external accountability and performance requirements.

Developing and Implementing Measures That Matter

 is a multi-stakeholder, consensus-based measure development organization that collaboratively promotes appropriate medication use and develops strategies for measuring and reporting performance information related to medications. PQA has been developing measures since its inception in 2006, with the most well-known measures being those included in the Medicare Star Ratings program. Data from the Star Ratings program have been used to compare the quality of different health plans. Health plans have begun to collaborate with pharmacies, since pharmacists and pharmacy teams have the ability to influence improvement on the measures.

Since the use of PQA-endorsed measures in the Star Rating program, gradual improvement has occurred in appropriate prescribing and adherence rates.

However, a need still remains to develop medication use measures to address gaps in patient care. As part of the measure development process, PQA continually looks to the future and assesses measurement needs. As the healthcare system and new care delivery models evolve, so do the types of measures that are needed to support team-based care delivery systems and care transitions.

Recently, PQA members approved the first Quality Improvement Indicators (QIIs) for pharmacist-provided medication therapy management (MTM) services to patients recently hospitalized, to address readmissions.

The descriptions for these QIIs are: 1) the percentage of high-risk patients who have been discharged from the hospital and who receive MTM from a pharmacist within 7 days; and 2) the percentage of high-risk patients who received MTM from a pharmacist within 7 days post discharge who are readmitted within 30 days of their discharge.

These QIIs are not intended to be performance measures for payment or public reporting. Rather, they are intended to be used by healthcare organizations for internal quality improvement as part of a layered, cross-cutting approach to improve effectiveness of care, affordability (ie, avoid costly readmissions), population health (ie, prevent harm to patients), and patient- and family-centered care (ie, medication reconciliation).

 is currently developing a set of QIIs focused on a team-based approach to identifying patients with hypertension, and then referring for and providing comprehensive medication management (CMM) to achieve blood pressure control. This set of metrics is intended for use by integrated care teams in patient-centered medical homes, ACOs, or community-based health teams that will address structure (referral for CMM), process (provision of CMM), and outcomes (improvement in blood pressure and blood pressure control).

Most current pharmacy-related measures are medication-related performance and adherence measures. These measures have been adopted by Medicare Part D, quality organizations, health plans, and pharmacies as part of pay-forperformance programs, fulfillment of contractual obligations, plan comparisons, research, and public reporting. However, with the evolution of new healthcare delivery models, there is a need to develop additional quality improvement measures that focus on medication management processes and patient outcomes that can be used by pharmacists, healthcare providers, and quality improvement specialists to implement strategies to better understand the efficiency and outcomes of internal medication management processes.

As new care delivery models evolve to focus on value, accountability, and team-based care, new measure development opportunities exist to address care gaps and to support team-based care delivery systems and care transitions.