Population Health Insider February 2009
Second interim report
CMS: MHS did not save money or improve health
The second interim report of the Medicare Health Support (MHS) project was not what disease management (DM) hoped, but industry leaders say the analysis is not an indictment against the DM concept.
The Centers for Medicare & Medicaid Services (CMS) released the review of the first 18 months of the DM-inspired project to Congress shortly before the new year began.
The report found that DM programs did not save money, improve beneficiary self-management and physical and mental health functioning, or cut acute care utilization or mortality during the first half of the three-year project.
The second interim analysis found that none of the companies involved in the pilot reached the targets that would warrant program expansion.
The programs did not improve clinical quality of care and beneficiary satisfaction or achieve budget-neutral spending targets, according to CMS. (For a list of the companies included in the project, see “Medicare Health Support organizations” on p. 3.)
A CMS official says the interim findings show that in the first 18 months of the MHS project, DM was not successful in the Medicare fee-for-service (FFS) population.
Note: CMS policy does not permit public attribution of comments by its staff members.
“The first finding is that the approaches that have been used in commercial and the managed care populations, at least early on, have not achieved the same results in the Medicare fee-for-service population,” says the CMS official. “We haven’t found a lot that has really worked in terms of what’s worked and what hasn’t worked. The jury is still out until we see what happens in the end result.”
The CMS report, which Washington, DC–based RTI International prepared for the agency, provided the following key findings:
- Several vulnerable subpopulations of Medicare FFS beneficiaries were less likely to participate in the MHS pilot
- The DM companies’ interactions with participating beneficiaries are unlikely to produce significant behavioral change and savings
- There was limited effect in improving beneficiary satisfaction, care experience, self-management, and physical and mental health functioning during the first 18 months of the project
- Seven of the eight MHS organizations (MHSO) had a positive effect on one or more process-of-care measures, but no positive effect on reduction in acute care utilization or mortality
- Fees accrued during the first 18 months far exceeded the savings produced
The CMS official says the MHSOs learned from the project and modified their offerings to better match the population during its course. “The different participants in the program have learned that they need to adapt to the needs and demographics in the Medicare fee-for-service,” the official says.
Although some may see the interim results as a judgment against DM’s nurse call center model, the results did not represent one particular care model. Instead, each of the eight MHSOs offered different programs, which were primarily call center–based, but some focused on physician outreach and face-to-face interactions with beneficiaries, the CMS official says.
The second MHS project report included deeper analysis that encompassed the project’s first six months, says Gordon K. Norman, MD, MBA, executive vice president for science and innovation at Alere, a DM company in Reno, NV, and chair of DMAA: The Care Continuum Alliance, which represents the DM/population health industry. However, the most recent analysis did not include the details of what did and didn’t work, which is what DM is interested in, he says.
Norman also questions CMS’ selection of Medicare beneficiaries who took part in MHS, many of whom were too sick or not sick enough to participate in a traditional DM program.
Health Dialog was one of the eight DM companies that participated in the MHS project, and one of the five that completed it. Christopher Coloian, senior vice president of health services at the Boston-based DM company, says the DM industry can learn from the MHS project, which is why Health Dialog stayed with it.
Coloian says he hopes healthcare leaders don’t view the interim report as the definitive study about DM’s effectiveness. DM has grown to a more than $2 billion business in commercial and Medicare Advantage populations because its customers know of the concept’s success in improving member health, he says.
“It would be a shame if folks took a very vibrant industry that we think is at the core of helping with medical cost management and quality improvements and really helpful in mitigating one of this country’s real concerns, which is the spiraling cost of healthcare, and put it through the sole lens of the Medicare Health Support finding,” says Coloian.
Key findings from second report
In analyzing the first half of the MHS project, RTI discovered that:
Several vulnerable subpopulations of Medicare FFS beneficiaries were less likely to participate in the MHS pilot. Although MHS was designed as a broad population-based FFS program, the participating populations were healthier, less costly, and used acute care services less than beneficiaries who did not participate in the program, RTI reported.
The interim review suggested that MHSOs’ recruitment strategies did not reach the most at-risk and sick populations, such as dual Medicare/Medicaid eligibles and beneficiaries with disabilities. In response, MHSOs said contacting costly beneficiaries was difficult because some were in nursing homes or other institutions and at the end of life when the project began.
The DM companies’ interactions with participating beneficiaries is unlikely to produce significant behavioral change and savings. In its report, RTI questioned MHSOs’ limited telephonic outreach. During the seven to 18-month time period, most beneficiaries received between two and five months of telephonic support, according to the report.
This limited outreach caused RTI to wonder whether the nurse call center model can work in the Medicare FFS population. “Given the lack of consistent monthly or bimonthly interaction with many of the MHS participants, it is unlikely that the MHSOs will be successful at changing beneficiary behavior with respect to self-management of their chronic illness,” wrote RTI.
RTI also found that beneficiaries who were surveyed did not experience “meaningful improvements” in the area of self-care activities. In fact, of the 27 gauged beneficiary survey measures, 13% showed significant positive intervention and 2% alleged negative intervention effect.
The remaining indicators were not statistically different between the intervention and comparison beneficiaries.
MHSOs achieved success helping beneficiaries set goals to address their care needs, but there was little improvement in self-efficacy or self-care activities. The combination of limited calls to beneficiaries coupled with self-care behaviors not improving 18 months into the project caused RTI to state that the MHS project will not significantly improve behavior.
“To positively affect acute care utilization, one would expect to see improvement in self-care behaviors by the midway point of the pilot, and savings have proven illusive to date,” wrote RTI.
There was limited effect in improving care experience, beneficiary satisfaction, self-management, and physical and mental health functioning during the first 18 months of the project. RTI reported that only two of the MHSOs, Health Dialog and Aetna, improved beneficiary satisfaction, which gauged how well a company was helping beneficiaries cope with their chronic condition.
In addition, none of the MHSOs “demonstrated consistent positive intervention effects across the four domains of satisfaction, care experience, self-management activities, and physical and mental health functions,” according to the 18-month interim report.
The MHS authorizing legislation stated that the organizations needed to improve quality of care and beneficiary satisfaction and achieve savings targets so CMS could expand the pilot. The organizations did not achieve those goals in the first 18 months, RTI wrote.
“The focus of the pilot program interventions was largely on impacting beneficiary behavior to better manage their chronic illness,” wrote RTI in its review. “Yet these results show little evidence of changes in self-efficacy or self-care. We did not observe any consistent pattern of positive intervention effects by disease cohort of heart disease only, diabetes only, and heart failure and diabetes.”
Seven of the MHSOs had a positive effect on one or more process-of-care measures, but no positive effect on reduction in acute care utilization or mortality. To achieve quality improvement measures and improve health outcomes, MHSOs needed to increase rate of receipt of claims-based process-of-care measures and improve health outcomes by reducing hospitalizations, readmissions, ER visits, and mortality rates relative to their comparison group.
One of the provisions of the MHS legislation was that the project improve quality and health outcomes. RTI tested five quality measures and found that only 40% of the measures were achieved in the patient population. Seven of the MHSOs demonstrated at least one positive intervention effect, but the rates of improvement in the quality-of-care measures were only 2 to 4 percentage points.
None of the organizations demonstrated positive intervention effects related to health outcomes. RTI found no reduction in mortality rates or time to death during the first 18 months of the pilot.
Fees accrued during the first 18 months far exceeded the savings produced. The third requirement for program expansion was financial savings defined as budget neutrality when factoring in MHSOs’ accrued management fees.
At the MHS project’s halfway point, RTI said none of the MHSOs achieved gross savings. Savings at the halfway mark offset 12%–26% of estimated accrued fees for four of the MHSOs, with the remaining showing no evidence of savings.
“The four MHSOs with modest savings would have to increase their rate of savings by roughly three- to sevenfold, while the second group of four nonsavers has a significantly higher hurdle,” according to the second interim report. “This seems unlikely given performance over the first 18 months.”
The five findings create a fairly negative view of the MHS project. “Given the limited gains regarding quality of care and savings to offset accrued monthly management fees, it will be difficult to justify these private disease management models on cost effectiveness grounds—at least for chronically ill Medicare FFS beneficiaries,” wrote RTI in its analysis to CMS.
Two years away from third report
The CMS official says the third report for the MHS project will provide the findings for the whole three-year project. That analysis is expected in February 2011, although the official says CMS will release the findings sooner, if possible.
For the final report, Norman says he would like to see which programs worked and which did not so the DM/ population health industry can learn from the experience. The analysts should not only compare programs, but delve deeper to find the results from subgroups, such as those with dementia and depression, he says.
“I think that’s how you tease out the combination of factors that may explain the totality of the results … I am very hopeful that the final report that looks back over the whole three-year experience does a better job of probing into areas that will maximize our learning from this very large, very expensive trial,” Norman says.
The CMS official says the interim report highlighted some subpopulation comparisons, such as heart failure beneficiaries versus those with diabetes, but the researchers didn’t find much difference between the results. If CMS and RTI don’t provide the level of detail that allows the DM/population health industry to learn which programs worked and which did not, Norman says he hopes CMS provides the data to a third party to analyze.
“We can ill afford to reject looking at any promising solution just as we can ill afford to throw out any favorable baby with the undesirable bathwater from MHS,” says Norman. “There has to be some valuable learnings hidden by this sort of blanket up or down evaluation bias that I think has been portrayed in the first two reports.”
The ultimate message from the MHS project is the difficulty of intervening with the Medicare FFS population, says Norman. One reason is that, unlike Medicare Advantage, many FFS beneficiaries don’t have primary care physicians.
Is this the end of DM Medicare projects?
Despite the MHS project’s negative interim results, CMS will continue to test DM programs. DM/population health elements are evident in other demonstration projects.
For example, LifeMasters Supported SelfCare, Inc., has been working on a successful demonstration project in Florida that is being offered to beneficiaries who are dual-eligible for Medicare and Medicaid and have congestive heart failure (CHF) or a combination of CHF, diabetes, and coronary artery disease.
The Medicare Medical Home Demonstration that is scheduled to begin in 2010 could include a DM/population health element.
Patients need the various elements of healthcare working together to create a better system, Norman says, adding that one area of healthcare can’t do it alone.
“I don’t think either party working alone can optimize population health outcomes,” says Norman. “That’s obviously a bias I have, and it may or may not be shared in the provider community, but it would be a shame to see magic-bullet attempts on the provider side to go through the same learning curve as DM over the last 15 years only to realize that we need to converge and integrate our respective roles and capabilities in order to maximize the outcomes.”
Coloian says he hopes CMS and the DM industry take what they learned from the MHS project and forge ahead with another DM pilot because MHSOs better understand the Medicare FFS population.
“It is incumbent upon Medicare and the industry to really learn not only about the first 18 months, but the entire 36 months, and look for the next opportunity,” says Coloian.
Tale of MHS
The Centers for Medicare & Medicaid Services (CMS) kicked off the Medicare Health Support (MHS) project in 2005 as a way to test the effectiveness of disease management (DM) programs in the Medicare fee-for-service (FFS) population for heart failure and diabetes beneficiaries. CMS selected eight organizations to take part in the first care management project in the Medicare FFS population. The DM companies focused on eight geographic areas and supplemented the care of about 287,000 chronically ill Medicare beneficiaries.
Before long, companies involved in the project began raising concerns, and three MHS organizations (MHSO) ultimately dropped out before the project’s completion. The organizations that dropped out complained that the patient population wasn’t as sick as expected, the initial 5% savings requirement was too ambitious, and CMS’ collaboration and transparency were lacking.
After a disappointing first interim report, CMS decided to end Phase I of the MHS project with no plans for a Phase II. The news was met with much consternation from the DM/population health industry and caused many to wonder about the effectiveness of DM programs. The announcement also affected one of the MHSOs—Healthways—which saw its stock drop after the announcement.
Industry leaders petitioned Congress and the Senate to push for a Phase II, but the MHS project ended quietly in 2008 without talk of reauthorization.
Those involved with MHSOs and DM leaders have raised several questions about the project, such as the composition of the target group (claiming DM could not effectively help those who were near death and those who were healthy but living with a chronic disease), the lag time in CMS getting current claims data to the MHSOs, and the lack of timely updates provided by CMS so the DM industry could learn from the project.
The final report that will analyze all three years of the MHS project is expected in February 2011.
Medicare Health Support organizations
Organization Target geography End date
Healthways Maryland and Washington, DC July 31, 2008
LifeMasters Supported SelfCare Oklahoma December 31, 2006
Health Dialog Services Corp. Western Pennsylvania August 14, 2008
McKesson Health Solutions, LLC Mississippi May 31, 2007
Aetna Life Insurance Co. Chicago (surrounding area) August 31, 2008
CIGNA Health Support Northern Georgia January 14, 2008
Green Ribbon Health West-central Florida August 15, 2008
XL Health Corp. Tennessee (selected counties) July 31, 2008
Source: Centers for Medicare & Medicaid Services.
Health Dialog implements what it learned from MHS
As one of the leaders in disease management (DM), Boston-based Health Dialog knows how to reach and engage patient populations in the commercial setting.
In signing up for the Medicare Health Support (MHS) project, the DM company—which had extensive experience serving senior and Medicare lives—hoped to learn more about working in the specific Medicare fee-for-service (FFS) setting and implement that knowledge into its wider book of business.
The MHS project allowed Health Dialog to “dive deeper” into the senior population, where it learned about the importance of targeting the right people, supporting them appropriately, getting timely information, and the critical role that providers play in healthcare, says Molly Doyle, vice president of government programs at Health Dialog.
Health Dialog quickly found many unexpected beneficiaries involved in the project. There were seniors who were near death and institutionalized who DM could not help and also those who experienced a chronic condition at one time but were in fine health at the start of the MHS project, both of whom were less likely to benefit from a DM program. These groups needed program refinements, and Health Dialog and others enhanced programs throughout the course of the pilot to better serve their population.
The 10% most costly beneficiaries in Health Dialog’s project were seven times more costly than the average cost of a Medicare beneficiary, Doyle says. Many of them were cognitively impaired and/or institutionalized, so the organization was only able to engage them through health partners (i.e., family and friends) when they could be identified. Meanwhile, more than half of beneficiaries cost about one-third of an average Medicare beneficiary, meaning DM could not save much money by avoiding hospitalizations and other medical care.
Having patients at either extreme reconfirmed the importance of choosing the right people for DM. “Targeting the right population is essential, and having the data to target people when they need support is critical,” Doyle says.
When signing up for the project, Christopher Coloian, senior vice president of health services at Boston-based Health Dialog, says the company expected that heart failure patients would have at least two medical claims for heart failure, which are the kind of people DM programs would usually target. But CMS added beneficiaries who didn’t fit that criteria and were, in fact, healthier.
One of the problems was that CMS chose beneficiaries approximately 18 months before the MHS project began, which meant those patients could have experienced wide varieties of health improvements or setbacks, Coloian says. In the commercial setting, Health Dialog receives hospital admission information daily and full claims data within 30 days after an episode, as compared to 60–90 days in the MHS project. The time lag in the MHS project meant that Health Dialog could not connect with at-risk beneficiaries during the critical window right after diagnosis or hospitalization.
In commercial-setting hospital admissions, data are received daily, whereas CMS was not able to provide this information at all. Health Dialog developed a work-around to receive notification of hospitalization through the fiscal intermediary for its region, but still resulted in a 10–14 day delay in notification from the date of admission.
However, the work-around wasn’t in place until six months into the three-year project, which means the second interim report may not have been affected by the data exchange improvement. In addition, medication claims data were not available from CMS until after the 18-month evaluation, according to Health Dialog.
“In the first 18 months, you’re looking through a window where some of the MHS pilot infrastructure and design issues were having to be mitigated by innovation and learning,” Coloian says.
Having more up-to-date information allows the DM company to reach out to the person when he or she could most be open to change behavior, such as at the time of hospitalization or a new diagnosis, Doyle says. “If you find out about someone going to the hospital five or six months later, the window is closed,” she says.
Health Dialog actively engaged physicians in the MHS program through outreach and sharing clinical data, and physicians were receptive to this outreach. However, this population would typically require much more intensive physician engagement.
CMS randomized the beneficiaries rather than the providers, which meant that Health Dialog couldn’t communicate with the physicians without contaminating the control and intervention group separation, Doyle says.
If the randomization scheme was different, Health Dialog would have had more extensive integration with providers. Because the population was randomized at the individual rather than provider level, members of the control and intervention group would be affected by the physician activities, Coloian says.
“[Randomizing beneficiaries rather than physicians] also diminished the number of lives that were within any individual practice, so there was very little concentration of intervention group members within a practice.
MHS was an opportunity for the industry to learn more about caring for Medicare beneficiaries in the FFS program, Doyle says. Determining how best to work with patients from different patient populations takes time, and the MHS interim results show that learning process.
“This was the first large-scale effort to implement care management/disease management in the fee-for-service setting,” Doyle says. “It took the managed care industry many, many years to develop the infrastructure to support care management. We understand that it will take time to do the same in the fee-for-service environment and know there would be challenges along the way.”
Chronic illness self-management puts patient in charge
The one who controls an individual’s healthcare is not the doctor, the health plan, or population health company. It’s the individual patient.
That sentiment is catching on as the disease management (DM)/population health industry tries to create a system that goes beyond nurse call center programs. Population health has already expanded into technology, remote patient monitoring, and automated call centers as a way to supplement the costly call center–based system. The next step in the process is putting the patient in control, says Kevin J. Widenhaus, PhD, director of behavior science and data analytics at HealthMedia, Inc., an Ann Arbor, MI–based healthcare company that specializes in combining technology and behavioral science to emulate a health coach via the Web.
The progression is needed because the current healthcare system does not prevent illness and disease, Wildenhaus says, adding that chronic illness self-management programs move away from member education to empowerment and away from information to intervention.
Progressive population health organizations understand the need to empower the patient with support from the healthcare support system, Wildenhaus says, adding that the future of population health is in emphasizing prevention and wellness, such as incorporating behavioral health programs as core components of the program. With chronic illness expected to inflict 164 million people in the United States in 2025, call center–based programs won’t be an economically viable option, he says.
“The only way you can provide population health is to take this approach that the patient has to be at the center. It’s the only viable way to do it financially and economically from a scalability perspective and a behavior change perspective,” says Wildenhaus, who spoke about the topic at the November 2008 DMAA: The Care Continuum Alliance conference.
Studies show that phone coaches only improve patients’ behavior for a short time and put them into a passive rather than active role in their health, Wildenhaus says. Traditional DM interventions intensify after a hospitalization or other healthcare interaction via outbound telephonic interventions. There is little intervention for those who are healthy or at risk of a particular disease.
Through the chronic illness self-management model, behavior change interventions continue across the care continuum via a Web-based program, Wildenhaus says. (See Figure 1 on p. 9 and Figure 2 on p. 10.)
HealthMedia’s Care for Your Health program is a Web-based chronic illness self-management program that emulates a health coaching session—without a coach—to deliver an individually tailored condition-specific management plan.
Wildenhaus says the following are ways to motivate people in self-management programs:
- Find out what is important to the individual at the baseline
- Build intrinsic motivation
- Build the individual’s confidence to help him or her overcome barriers, such as time management and feeling overwhelmed
- Tie the self-management program to the person’s values
“We really focus on the fact that many times, those external factors are what will bring people into a program or an initial step to take responsibility,” Wildenhaus says. “But in order to sustain or maintain it, you need to move people into an intrinsic driver.”
The self-management program asks people what internally and externally motivates them. The program teaches the difference between the two motivators and helps drive them to intrinsic value by tying it to values and give them the skills and techniques to overcome barriers.
Eight steps to self-management
Wildenhaus says there are eight elements to a successful chronic illness self-management program. Participants should:
- Manage medical, emotional, and personal issues
- Gain motivation and self-confidence associated with managing conditions
- Manage sleep, pain, fatigue, and depression
- Improve relations and communication with doctors and pharmacists
- Comply with treatment and medical advice
- Build social support
- Create goals and develop planning skills
- Accept personal responsibility for managing their condition or conditions and understand their role on the treatment team
It’s important to personalize the program so the individual is given information that interests him or her, Wildenhaus says. It’s one thing to place health information on a Web site. It’s quite another to make it something that people can integrate into their lives.
For example, rather than add dozens of links to the member’s home page about various types of disorders and health tips, customize the menu to include only items of interest, such as specifics about particular ailments, illnesses, and lifestyle choices. A 30-year-old mother of a small child is most likely more interested in early childhood and new mom programs, so populate the member’s page with links to those programs. Alternatively, a 75-year-old woman with high blood pressure is more apt to read articles about hypertension and elderly ailments, such as osteoporosis and falls.
“That’s why we get such a high response rate in terms of people feeling the program was tailored to them, reading the information completely, sharing with others, etc.,” says Wildenhaus. “We take out all the stuff that is not helpful to them or not necessary, and we boil down the essential thing in a tailored way that they would most need, want, and benefit from.”
This can be accomplished by gathering information from an individual’s health risk assessment and claims records and asking the individual to input information such as weight, body mass index, medications, and doctor visits. Through that information, Care for Your Health creates a member-specific Web page. Employers can also help by giving incentives for employees to visit their health Web pages.
Highmark Preventive Health Services, a Pittsburgh-based health insurer, implemented HealthMedia’s Care for Your Health program as part of its DM offerings, which includes the Blues on Call Condition Management, a 24/7 nurse phone line that helps members with chronic disease manage their conditions.
Connecting to the individual patient is an important way to spark change. There is already a slew of health-related information online, and most people know about healthy food and lifestyle options. The problem is that they still don’t follow the advice, which is where creating an individualized message comes into play, says Wendy R. Vida, RD, clinical preventionist at Highmark.
Many people don’t understand how healthy choices affect them or foster behavior change, but the Care for Your Health site has tailored information for the individual with specific calls to action, Vida says. Highmark’s members have increasingly turned to the Web to acce
- As Medicare Advantage Cuts Loom, Disagreement Over Program's Stability
- 3 Management Lessons from a Supermarket Debacle
- Medicare Advantage Carriers See 'No Choice' But to Accept Cuts
- Physicians to Appeal 'Docs v. Glocks' Ruling in FL
- CA Fines 8 Hospitals for Medical Errors
- Centralizing the Revenue Cycle Protects the Bottom Line
- Revenue Cycles Get a Boost from Simple JPEG Files
- IOM Identifies GME Problems, Calls for Finance Changes
- Employers Weigh Risks, Benefits of Private Exchanges
- Doctors Feel Pressure to Accept Risk-based Reimbursement