Outpatient Management of Diabetes Mellitus in Five Arizona Medicare Managed Care Plans

Abstract

We report findings on the outpatient management of diabetes mellitus in Medicare beneficiaries enrolled in five Arizona Medicare-managed care plans. These findings are the baseline of an ongoing collaboration between the Health Services Advisory Group, Inc., Ari zona's Peer Review Organization (PRO), and the five plans whose object is improved care of diabetes pa tients. The purpose of the study was to determine con gruity between quality indicators identified by the five plans and the care actually received by diabetes patients enrolled in the five plans. The five plans agreed on a common set of quality indicators, including 10 services and 10 measures of patient status. Each plan has iden tified its diabetic population, 75 of whom are randomly selected each quarter by the PRO for chart review and inclusion in the study. The findings in this report cover two quarters of data. Data from chart review were exam ined to determine the extent to which actual practice reflected the indicators. The mean patient age was 71.8, and for most patients onset occurred between 55 and 69 years of age. About 25% had a positive family history, and we estimate the annual incidence of diabetes in this population to be about 1.1%. Mean hemoglobin A_1c (HbA_1c) was 8.9 ± 2.1%; 46% were hypertensive; 42% continued to smoke cigarettes; 36% had retinopathy; 20% had proteinuria; and only 22% were on some kind of exercise program. Thirty-two percent were hospitalized during the 1-year baseline period, and the average num ber of outpatient visits per patient was 11.1 ± 7.4. When care provided to diabetes patients enrolled in the plans was compared with the 10 quality standards identified by the plans themselves, only two of these standards was attained in more than 60% of patients: blood pres sure, 98.7%; and foot examination, 62.7%. Two stan dards were achieved less than one-third of the time: urine dipstick, 10.4%, and appropriate use of angioten sin-converting enzyme (ACE) inhibitors, 31.25%. The others were all between 40 and 55%. Of the 10 service standards, about one-third received 1-4, one-third re ceived 5-6, and one-third received 7-10. Only 5% of pa tients received 9 or 10 services. Outpatient management of diabetes patients in managed-care plans is similar to that in fee-for-service. When compared with fee-for- service or another HMO, a higher proportion of Arizona- managed care patients had HbA_1c, and a much lower pro portion had a dipstick test for urine protein. Values for other variables were usually within 10 percentage points of each other. Regardless of payment scheme, diabetes care is characterized by inconsistencies, omis sions, and a lower than desirable level of services. Al though few patients received most of the indicator services, diabetes patients are nevertheless high uti lizers of medical care, both in and out of the hospital. The hospitalization rate is twice that of Arizona Medi care beneficiaries as a whole, and the number of office visits is three or four times that reported in other stud ies. Further, it seems that many visits are required to achieve even these modest service levels. Had the aver age number of visits been six or less, HbA_1c rates, for example, would have fallen to less than one-third in three of the five plans. We believe that these data are conservative because it is likely that some and perhaps most of these indicators are underreported. It should be emphasized that these are baseline data whose pur pose is to provide a basis against which subsequent im provements many be measured.