The Good News in the CBO’s Medicare Reform Studies

Supporters of health care reform were discouraged this week by a Congressional Budget Office report that synthesized research on Medicare pilot programs focusing on disease managment and care coordination. The CBO concluded that most such projects have not reduced Medicare spending and that, on average, they had little or no effect on hospital admissions. A similar CBO review of studies on value-based payment initiatives similarly found that only one of four demonstrations yielded significant savings for Medicare.

Because such innovations are considered to be among the best hopes for constraining Medicare's growth rate while improving the quality of care—and are key components of the Affordable Care Act—those deflating headlines provided new ammunition to politicians who want to repeal the legislation. But a closer reading of the CBO reports, and some of the particular studies they review, provide all kinds of useful insights about relatively successful strategies that can be more broadly emulated while avoiding approaches that clearly don't work. 

The Medicare pilot programs were a collection of experiments that varied in a multitude of ways. Like trial-and-error processes in laboratories, they are extremely valuable in pointing the way toward future modifications in the next wave of experiments drawing on the lessons gleaned from the first set. That is how progress is made. It would have been miraculous if the pilot programs had been uniformly successful given the incredible complexity of the health care system and the Medicare program, the deeply entrenched practices and culture of health care providers, the inherent difficulty of treating older patients with multiple chronic ailments, and the limited experience with these kinds of reforms. Moreover, most of the pilot projects lasted only three years and affected a relatively small number of patients, leaving little time to ramp up fairly dramatic changes, allow for adaptation, and produce meaningful results in relation to a control group.

Even still, some of the pilot programs were found to have produced impressive outcomes. Four of the 34 disease management and care coordination initiatives reduced hospitalizations by 15 percent or more, and another seven yielded hospitalization reductions of 6 percent to 15 percent.  Because reducing hospitalizations of chronic care patients save substantial costs while demonstrating that those individuals have been able to maintain better health, they are a key measure of success. So it is good news that a substantial subset of the pilot programs were able to make significant progress on that front, providing models that might be pursued and expanded  upon in other settings. At the same time, the approaches used elsewhere with less success are now known to be ineffective—so we have learned about mistakes to avoid.

One example of a relatively successful pilot program, analyzed in great depth in this 162-page study, was Massachusetts General Hospital's Care Management for High-Cost Beneficiaries Demonstration. Relative to a control group, it succeeded in reducing hospital admissions by 19 percent to 24 percent (the different figures were connected to different time periods and different groups of patients).  In addition, the program achieved significant cost savings: for every dollar invested in the program's management fees, Medicare received $2.65 in savings on beneficiary health care services. Lyle Nelson of the CBO explains why the Massachusetts General program may have produced better results than most of the other pilots:

Unlike nearly all other programs tested in these six demonstrations, the program at Massachusetts General is closely integrated with the health care delivery system. The program has the strong backing of the hospital’s senior management and the physician group, and physicians have been involved in the program’s initial design and evolution. The care managers are staff members of primary care physicians’ practices, and they have access to patients’ electronic medical records. Patients of Massachusetts General Hospital and its affiliated physician group reportedly obtain the vast majority of their health care within that integrated system, so the electronic medical records provide care managers with current information on nearly all of their patients’ medical care.

Moreover, the hospital notifies care managers when their patients are hospitalized or admitted to the emergency department. The care managers interact with patients by telephone and in person during physician office visits and hospital stays, and they have access to a pharmacist to address potential problems with patients’ medications.

Other nuggets from the longer study of the Massachusetts General  program that may help to explain its relative success:

• In the early stages of the CMHCB demonstration, CMP leadership learned that many
  highcost, complex patients have mental health issues that were not effectively addressed by the current model of health care delivery or its pilot program. As a result, the program allocated greater resources to support mental health, hiring a social worker to assess the mental health needs of CMP participants and support them in accessing psychiatric care as needed or provide treatment if appropriate.
• Since many members of the CMP population have complex medication regimens, MGH
  enlisted the support of a pharmacist to review the appropriateness of medication regimens and assist patients with access to medications. The pharmacist also evaluated medication regimens to identify opportunities to reduce the number of medications and to suggest alternative therapies. Lessons learned during the early stages of the demonstration helped to motivate a change in MGH’s medication delivery services—MGH began providing home delivery of medications 5 days per week rather than 2 days per week.
• Case managers assigned to each practice met with physicians at the practices to describe
  the program, the skills that they bring to the physician practice, and their interest in collaborating to support patients in their efforts to manage their medical conditions. Case managers collected information from providers about how they could add value to the medical practice.
• MGH invested considerable personnel resources to develop and implement its CMP. At
  the time of the initial site visit, the program was staffed with 11 nurse case managers who
  received guidance from the program leadership and support from the project manager, an
  administrative assistant, and a community resources specialist. Each case manager was located in a physician practice and provided case management support to a group of 180 to 220 patients who received their primary care from a provider in that group. In addition, almost all case managers also “floated” to one or two additional small physician practices, which had five or fewer MGH CMP eligible patients. Responsibilities included conducting patient assessments, visiting patients who were hospitalized at MGH (when feasible), contacting patients who visited the emergency department or were recently discharged from the inpatient services, calling patients scheduled for office visits each week, following up with patients who missed office visits each week, making followup calls to provide case management services, and promoting the MGH CMP to physicians.
• MGH developed a series of clinical dashboards using data from the MGH electronic
  medical record (EMR), claims data, and its enrollment tracking database. The dashboards
  allowed MGH to examine trends in health care utilization and outcomes, overall and by
  enrollment status, physician practice, and/or case manager, as well as activities of its case
  managers. Examples of indicators included in the dashboard are: number of assessments
  completed within 90 days, number of referrals or interventions conducted, number of advanced directives in place, number of participants screened for depression, number of participants with a positive screen for depression referred to mental health, and the top 10 discharge diagnoses.
• MGH enlisted physician support to help ensure the success of its CMP in providing highquality care to patients. Physicians were asked to conduct the following activities: encouragebeneficiaries to participate in the program and enroll them in the program when possible; collaborate with case managers to review initial assessment findings and develop care plans for each patient; inform case managers about patient events and refinements to patient care plans during the demonstration period; and discuss advance directives with enrolled patients.

Obviously, such ambitious interventions will be difficult to replicate elsewhere in settings that lack the physical and human resources available to Massachusetts General Hospital. But what they did there genuinely worked and was quite different from what the other pilot programs studies did. So the challenge going forward is to try to get more of these initiatives to do what they can to imitate Massachusetts General's approach, to the extent that they can. The payoff is very real. From the study's conclusion:

Based on extensive qualitative and quantitative analysis of performance, we find that
MGH’s CMP had success at improving primary care providers’ satisfaction with their quality of work life and improving some measures of beneficiary experience with care and functional status. We also find that MGH’s CMP had substantial success reducing acute care hospitalizations and ER visits and mortality, and achieving substantial cost savings. We find these latter successes within both the original and refresh intervention groups. The financial savings is particularly noteworthy given the relatively small sample sizes and regression to the mean effects. 

That's pretty terrific news for supporters of health care reform.

 

 

 

 

 

 

 

achusetts