NCQA tool doesn't adequately evaluate quality of diabetes care at health centers

UCLAFeb 16 2012

On the health front, the poor often have at least two things going against them: a lack of insurance and chronic illnesses, of which diabetes is among the most common.

The federal Affordable Care Act would expand the capacity of the nation's 8,000 community health centers to provide care for low-income, largely minority patients — from the current 20 million to about 40 million by 2015. The federal government is also trying to ensure that these community health centers deliver high-quality primary care, including diabetes care.

A crucial part of this is the implementation of what is known as the "patient-centered medical home model," which provides comprehensive, coordinated care among patients, their physicians and, sometimes, family members through the use of registries, information technology and other resources. It is intended to ensure that patients receive care on a continuous basis — rather than just during periodic visits to the doctor's office, for example.

The assessment tool used by federal government programs to measure whether a community health center is functioning as a "medical home" was developed by the nonprofit National Committee for Quality Assurance (NCQA). But, according to a new UCLA study, there's a problem: The NCQA tool doesn't adequately evaluate the services that determine the quality of diabetes care in community health centers. In fact, the study found, there is no relation between how well a health center scored on the NCQA assessment and the quality of diabetes care it provided.

The study is being released Feb. 15 as a "Web first" publication by the journal Health Affairs and will appear in the journal's March print issue.

"The major issue here is that the NCQA assessment tool was developed based on evidence of what worked for private primary-care practices that delivered care to insured patients," said lead author Dr. Robin Clarke, a physician in the Robert Wood Johnson Foundation Clinical Scholars program in the division of general internal medicine and health services research at the David Geffen School of Medicine at UCLA. "Because we have limited experience in applying the NCQA tool to community health centers, there is a question of what effective, patient-centered care for low-income patients actually entails."

For their study, the researchers had 30 Los Angeles County community health centers complete the 2008 NCQA Physician Practice Connections/Patient-Centered Medical Home (PPC-PCMH) tool, which assesses the operational systems the health centers use to identify, track and treat their patients. Health centers are scored on a 0-to-100 scale and based on their scores are given recognition levels ranging from Level 3 on the high end to "not recognized" on the low end.

To measure patients' quality of care, the researchers looked at medical records for 50 randomly selected patients from each of the 30 health centers. They assessed whether five recommended diabetes screening tests were conducted in the previous year: hemoglobin A1c, low-density lipoprotein cholesterol, blood pressure, a urine protein test, and a dilated eye examination. They also looked at three "intermediate outcomes" — risk factors that are precursors to a serious condition, such as elevated blood pressure preceding a heart attack. In this case, they checked for control of blood glucose levels and cholesterol, as well as blood pressure.

Of the 30 participating community health centers, eight earned Level 3 recognition on the PPC-PCMH assessment tool, three were at Level 2, and 19 were at Level 1. There was a wide range of NCQA scores on the tool, indicated that some health centers had many more medical home components than others. There were also substantive differences in the quality of diabetes care within the sample.

"We found that there was a broad distribution of NCQA scores and a broad distribution in the quality of diabetes care that these health centers deliver," Clarke said. "But there was no statistically important relationship in how well a clinic scores on NCQA and the quality of care it provides."

Clarke and his co-investigators write in the study that these federal programs launched by the health care reform law represent a special opportunity: a combination of stable insurance through health care reform and genuine patient-centered medical home care through a community health center, which could potentially help reduce illness and premature death among low-income diabetes patients. In addition, if these patients receive better primary care, they may be less likely to use emergency departments and require hospital care.

But as it now stands, this study raises the question of whether the NCQA tool, when applied to community health centers, can lead to those goals.

"There is a lot of potential for the positive effects that the patient-centered medical home model could have on community health center care," Clarke said. "But the NCQA's tool itself seems to miss the services that are important for low-income diabetes patients."

There are some potential limitations to the findings, the researchers said. For instance, while this study was cross-sectional and observational, a randomized longitudinal study would be needed to determine if a higher NCQA score can lead to better diabetes care. The study used the 2008 version of the NCQA's assessment tool, and an update version was released in 2011.