Project Description

MaineNET, a demonstration project funded by the Robert Wood Johnson Foundation was originally implemented as a primary care case management model in three counties to improve clinical and administrative coordination of primary, acute and long term care services for dually eligible (Medicare and Medicaid) elders and adults with disabilities. A case manager was located in the physician's office to improve coordination and planning. Case managers from the regular Home and Community Based Services program served about 100 beneficiaries. Program enrollment was small due to the difficulty convincing beneficiaries to enroll. Although the coordination activities were considered helpful by the participating Medicaid beneficiaries and their physicians, the program did not appear to be financially sustainable due to small caseloads and enrollment.

After the brief care management implementation, the program transitioned to a population-based chronic care management model designed to improve coordination of care for dually eligible beneficiaries by working directly with primary care physicians particularly with an aggressive focus on medication management. The revised program focused on beneficiaries with diabetes, heart disease and congestive heart failure and offered educational materials to beneficiaries and physicians. The primary goals were to:

• Provide physicians with data reports tracking utilization of pharmacy services central to chronic care management and key quality of care events,
• Improve prescribing practices for the frail and chronically ill elderly,
• Encourage and assist physicians to develop interventions to improve outcomes,
• Engage physicians in the development and participation in educational opportunities to address chronic care management.

Demonstration Project

Fifty primary care physicians serving 1,100 elderly and disabled Medicaid (MaineCare) beneficiaries at five pilot sites participated in the project. The MaineNET program manager from the Muskie School of Public Service, University of Southern Maine, met regularly with the physician leader or administrator of each participating group. Dually eligible clients with the target chronic illnesses were identified through Medicare and Medicaid claims data. Physicians received reports based on Medicaid (MaineCare) pharmacy claims and met periodically with the Program Manager to review data and discuss interventions. A pharmacy consultant contracted by the project provided additional academic detailing services to the pilot sites and imparted significant credibility to the reporting activities.

Pharmacy reports were developed that tracked population, practice, physician and individual-level data concerning the medication profiles of target patients. Over an 18-month intervention period, aggregate pharmacy data was presented quarterly for all the dually eligible beneficiaries of the each pilot site and included such criteria as multiple prescribers (patients receiving prescriptions from three to five or more physicians outside of the primary care practice site), polypharmacy (nine or more concurrent medications), potentially inappropriate medications for the elderly (otherwise known as the Beers list criteria), and no prescriptions filled within the last three months. The pharmacy reports given to participating providers also revealed in-patient or emergency room prescribing, patient compliance, duplicative therapies, narcotics use/abuse, as well as potentially harmful combinations of medications. The population of patients being tracked was refreshed each quarter due to the high turnover rate within the eligibility category.

Participating physicians were offered in-service education by the project Consulting Pharmacist to review findings and discuss alternatives to potentially inappropriate medications. All but one of the sites also agreed to have their clinical and office staff participate in educational meetings presented by the project staff. Physicians indicated that chronically ill beneficiaries often received care from multiple sources and coordination of care under the existing fee for service Medicaid system remained difficult for the primary care provider. After receiving input from the project physicians, MaineNET developed a high priority report that directed physicians to those beneficiaries for which coordination of care appeared to be at highest risk. Physicians were compensated at a contractual rate for the actual time spent in the academic detailing or in-service meetings with the project staff.

The MaineNET team developed materials that identified and facilitated referrals to available community resources. These efforts included producing a community resources guide and convening a community resources fair for clinical staff. Strategies to address chronic care management were reviewed with the participating physician groups. These included disease registries and other software tools, group visits, self-management education, disease flow charts and tracking tools. The physicians were directed to free or low-cost resources to assist in the coordination of care for individuals with chronic illness. Educational materials were sent directly to beneficiaries concerning management of their condition under a cover letter from the Medical Director of MaineCare.

The project team presented numerous physician and nursing educational workshops throughout the State of Maine concerning the issues of medication management for the chronically ill and elderly. MaineNET staff created the Continuing Medical Education (CME) product Improved Prescribing for the Elderly – a self-study packet with real-time case examples from the physicians’ panel of dually eligible patients. The activity included completing a Medication Review Checklist with information from the patient medical record and the individual pharmacy report. In the subsequent trial run with the pilot project physicians, 62% of the medication reviews in the packets generated a self-reported provider intervention.

Results

The most significant impact of the intervention was noted in the drop of potentially inappropriate medications (PIM) for the elderly. The rate of PIMs in the study population dropped by 36% from 46.4 to 29.5 over the 18-month intervention period. More importantly, the rate of new PIM prescriptions written dropped by 56%, from 32 to 14 per 100 elderly patients during the same period, reflecting a change in prescribing behavior. Polypharmacy within the study group held steady at 39% over the intervention period. However, the average number of drugs per patient per quarter fell slightly from 8.08 to 7.80. The number of prescribers per patient also held steady over the study period. Although there were significant geographical differences in regions of the state, an average of 22% of the study group received prescriptions from three or more prescribers outside the primary care office. In communities where access to health care is plentiful, there was a much higher rate of multiple prescribers.

A patient sub-group was identified through this project that may benefit from a more intensive intervention in the future. These individuals, the "priority patients" were younger dually eligible, with twice the number of prescriptions and twice the number of psychotropic medications and a four-fold increase in average quarterly pharmacy costs. These individuals represent an opportunity for a coordination of care and quality of care intervention that falls outside of the traditional singular disease management and addresses the complex medication management of the chronically ill with co-morbidities.

Lessons Learned

• Primary care for chronic illness is challenging under a fee-for-service system – there is little incentive for the coordination of care that is crucial for managing co-morbidities in the chronically ill population. The physicians in the project were eager to learn about drug cost issues, potentially inappropriate medications for the elderly, and educational strategies. While there is growing interest for technology solutions for disseminating clinical quality data, reports to providers need to remain simple and improvement activities direct.
• Collaboration with physician groups requires patience and flexibility as many are competing for the time and attention of health care providers. Compensation to providers for time spent on pilot activities was key to cooperation. In terms of the actual program intervention, significant engagement with influential physician leaders is the "make or break" relationship.
• In pilot sites where the physician leader was committed to incremental quality improvement, the measures of change were most significant.

Conclusion

The results of the MaineNET demonstration project indicate that a significant proportion of chronically ill patients are receiving potentially inappropriate medications, filling prescriptions from multiple providers outside their primary care practice, and are receiving nine or more medications on an ongoing basis. The pharmacy reports intervention have indentified potential coordination of care deficiencies (multiple providers), potential clinical concerns (polypharmacy, no prescriptions filled, noncompliance, and duplicative therapy), and quality of care opportunities (potentially inappropriate medications for the elderly, narcotic use/abuse, and disease appropriate therapies). In addition, through the collaborative work of the participating physicians, a priority report was developed to alert physicians to patients whose medication profiles may warrant a more immediate focus.