Population Health Insider, April 2009
Medicare demonstration project
Study shows difficulties in coordinating care
Saving money and reducing hospitalizations in the chronic care population is difficult, but more in-person interaction could play a role in bringing about cost savings and improved outcomes, according to a recent study.
Researchers looked into the Medicare CoordinatedCare Demonstration (MCCD) project, which tested whether care coordination programs reduced hospitalizations, cut Medicare expenditures, and improved quality of care for chronically ill Medicare beneficiaries.
The review of 15 randomized trials showed that onlytwo were successful in reducing hospitalizations and one achieved sustainable cost savings.
Among the highlights of the study: Care coordination programs without a strong transitional care component probably won’t bring Medicare savings, and substantial in-person contact targeting moderate to severe patients can be cost-neutral and improve some aspects of care, according to Effects of Care Coordination on Hospitalization, Quality of Care, and Health Care Expenditures Among Medicare Beneficiaries, which appeared in the February 11 Journal of the American Medical Association.
“Our results suggest that care coordination, as practiced by the programs participating in the demonstration from 2002 to 2006, holds little promise of reducing total Medicare expenditures for beneficiaries with chronic illnesses,” wrote the authors.
Randall Brown, PhD, director of health research at Mathematica Policy Research, Inc., in Princeton, NJ, and the study’s coauthor, says the results show that disease management (DM) programs with a call center–based foundation are not the right model for improving care for the chronically ill. This is because there isn’t enough collaboration between care coordinators, physicians, and patients or in-person contact for the classic DM model to be successful.
Brown says healthcare stakeholders are looking for a quick fix to the chronic illness problem, but reducing healthcare costs and improving patient outcomes is complicated and a call center–based approach doesn’t work.
“The encouraging thing is it can be done if you really look at the lessons we have learned over the past seven years of this study,” says Brown.
Jaan Sidorov, MD, MHSA, FAAP, an independent consultant who owns and operates Sidorov Health Solutions in Harrisburg, PA, and writes the “Disease Management Care Blog,” says the sobering results should not surprise the DM industry. Healthcare leaders continue to test care management approaches to see what can decrease costs and increase quality.
“For the right patient, it continues to make sense to rely on nurses that are available for in-person coaching, to use your predictive modeling algorithms to identify patients that are ‘high impact,’ to pursue medication adherence and persistence, to do everything possible to find and help patients that are recently discharged, and to build close relationships with the primary care physicians,” says Sidorov.
Demonstration tested high-risk beneficiaries
Through the MCCD, Medicare tested care coordination in the chronic disease population because half of the beneficiaries in 2002 were treated for five or more conditions, accounting for 75% of Medicare spending. Most of the costs were the result of hospital admissions and readmissions.
Studies have pointed to ways to effectively reach Medicare patients, such as patient-centered care, improving physician-patient communication, increasing patient adherence to recommended medications and self-care regimens, and making care more evidence-based. However, according to the authors of the MCCD study, “few relatively small, single-center trials of interventions that included some or all of these components have successfully lowered hospitalizations for Medicare patients.”
The 15 trials included fee-for-service Medicare beneficiaries, most of whom had congestive heart failure, coronary artery disease, and diabetes. (See Figure 1 on p. 3.)
Medicare paid providers a per month per member fee, and the program analyzed hospitalizations, costs, and quality-of-care outcomes.
Patients also completed a survey seven to 12 months after enrollment, which provided additional quality-of-care measures.
The interventions varied and included five commercial DM companies, three community hospitals, and three academic medical centers. All of the programs featured care coordinators (nurses) who provided patient education and monitoring, mostly via telephone, in hopes of improving adherence and communication with physicians. Nurses contacted patients twice per month on average. Among the interventions:
- All but one program educated patients on improving adherence to medication, diet, exercise, and self-care regimens
- Four programs focused on increasing physicians’ adherence to evidence- or guideline-based care
- 14 sought to improve care coordination
- 12 taught patients to communicate with their physicians more effectively
- 10 programs provided timely information on the majority of patient hospitalizations or ED visits
- Although the interventions varied in design, researchers found that the majority were simply not successful. Researchers discovered that:
- Two of the 15 programs showed significant reductions in hospitalizations compared to the control group
- None of the programs generated net savings
- Three programs had monthly Medicare expenditures less than the control group by 9%–14%
- No programs had favorable effects on the adherence measures, and only a few had a favorable effect on the quality-of-care indicators examined
“Despite these underwhelming results for care coordination interventions in general, the favorable findings for Mercy [Medical Center] and [Health Quality Partners (HQP)] suggest that the potential exists for care coordination interventions to be cost-neutral and to improve patients’ well-being,” the authors wrote.
Although the majority of interventions were not successful, the study’s authors did provide clues to why HQP and Mercy stood out:
- The programs had more in-person interaction. The two interventions averaged nearly one in-person contact per month per patient. “Relatively frequent in-person contacts may be necessary to develop the level of trust that patients and their families need to consider the care coordinator an integral part of their care network and to confide in the coordinator and rely on his or her judgment,” the authors wrote.
- The successful programs did not target patients who were low risk for hospitalizations or so seriously ill that interventions would not ward off hospitalization.
- The two groups taught patients how to take their medications.
- The care coordinators of the two groups worked closely with local hospitals, included timely information exchanges of hospitalizations, and worked to reduce readmissions.
- Care coordinators in the two programs collaborated with physicians. For example, HQP allocated space at its site for care coordinators to meet with patients, and Mercy set up time at local clinics for patients to meet with care coordinators at the same time as their physician appointments.
The authors wrote that the results and other evidence from the literature suggest that the most effective care coordination interventions would be a model similar to the ones implemented by Mercy and HQP, coupled with a proven transitional care model to prevent hospital readmissions.
Given Mercy’s and HQP’s success, CMS has expanded those interventions for up to three more years, with the continuation depending on whether an interim evaluation—slated for 2010—shows “evidence of persistent favorable effects.”
The sobering results of the MCCD may show that the medical home concept is a better way to improve care coordination, lower costs, and improve quality, the study authors wrote. The medical home could create close links between the patient’s care coordinator and physicians and allow for more in-person contact between the patient and care coordinator, leading to more timely information on hospital admissions.
“The successful interventions also may offer more detailed lessons for medical homes about how best to educate and monitor patients, the types of patients for whom they are likely to be most effective, and how to help patients overcome barriers to better self-care,” the authors wrote.
The MCCD study can also teach those interested in the medical home about the need for strong educational intervention and medication compliance, Brown says. Although the medical home concept has potential, medical homes must be targeted for at-risk patients. “If you don’t target people who are pretty high risk of hospitalization, then you’re going to spend a lot of money that won’t save you any hospitalizations, and that’s where the money is,” says Brown.
Sidorov questions the study authors’ conclusion that the medical home could work because there isn’t proof to reach that conclusion in the MCCD results.
“I personally think the pay dirt version will be provider-led care management that incorporates latest generation (telephonic) disease management and smart insurance benefit design and the right mix of physician reimbursement strategies,” says Sidorov, adding that patient registries, predictive modeling, and decision support will also play a role.
Although most of the interventions failed to control costs or reduce hospitalizations in the MCCD project, care coordination can work if designed properly, says Brown.
“The main thing is not to throw the baby out with the bathwater and conclude that nothing worked. Some did work,” says Brown. “It’s just that it’s not easy and it takes a well-designed program that is properly targeted if the goal is to save medical cost dollars. I think that’s true for both the commercial and Medicare population. That’s not to say that wellness and prevention programs aren’t a good idea for younger populations, but you’re looking for possible savings maybe years down the road and reductions in absenteeism and not reductions in hospitalizations for the most part.”
HealthLeaders Media survey
Health plans, physicians disagree on healthcare
Health plan executives believe quality improvement initiatives, pay for performance (P4P), electronic medical records, and consumer-directed healthcare will help their organizations in the next three years, whereas spiraling costs, sour relationships with physicians, and government intervention and mandates will have the most negative effect, according to HealthLeaders Media’s first health plans survey.
The survey also found that:
- Health plans and physicians disagree on several issues
- Health plans support consumer-directed health plans (CDHP), although many question whether CDHPs empower members
- Insurers like disease management (DM) programs, although they are unsure how best to provide those programs
The survey was part of the larger HealthLeaders Media Industry Survey 2009 report, which is based on seven concurrent surveys sent to healthcare leaders in seven segments across the industry: CEO, finance, technology, physician, health plan, marketing, and quality leaders.
Although health plan respondents pointed to several positives they expect during the next three years, they were not as positive about reimbursement cuts, physician and nurse shortages, uncompensated care, and increased regulatory scrutiny. (See Figure 2 on p. 6.)
When looking at the larger managed care and health insurance industry, respondents believe costs, provider relations, and government intervention will have the most negative effect on the industry in the next three years. (See Figure 3 on p. 7.)
Physician/health plan disagreement
For the past several decades, health plans/managed care and physicians have been fighting over payments, billing, and how best to provide quality care.
When asked which of six choices would most improve the relationship between payers and physicians, more than half of health plan survey respondents suggested P4P programs that focus on certain health measures or speeding up the processing, fixing, and paying of claims. Less than 20% of managed care respondents thought increasing reimbursement rates would be effective. (See Figure 4 on p. 8.)
On the other side of the relationship, more than half of physician respondents said that the best way to improve relations was to increase reimbursements, whereas only 7% suggested the health plan survey’s top response—P4P.
The disparity between health plans and physicians goes beyond reimbursements and the best model for healthcare, says Emad Rizk, MD, president of McKesson Health Solutions in Broomfield, CO. The two sides don’t even agree on the definitions of programs, such as what defines P4P, DM, and the medical home concept. “There’s too much disparity about what each of these are. I don’t think you can find two pay-for-performance programs that are the same,” says Rizk.
The disconnect was also evident concerning the model offering the best hope for healthcare. Health plan survey respondents suggested CDHPs were the way to go, whereas government-mandated universal health insurance and government-funded universal coverage lagged behind.
Ian Duncan, FSA, MAAA, president of consulting firm Solucia, Inc., in Farmington, CT, says he believes health plan respondents oppose government-run healthcare because segments of healthcare are still feeling the sting from the last attempt at it in 1994.
“Given that politicians in Washington are interpreting election results to imply a mandate for them to take greater control over one-sixth of the economy, the opinions of those who are actually responsible for delivering healthcare and who have day-to-day experience of current problems is sobering,” says Duncan.
Respondents to the same question from other areas, such as physicians and marketing personnel, suggested government interventions were a better solution. In fact, more than half of physician respondents chose a government solution. Physicians question CDHPs because the sickest patients don’t profit from the benefit design, says Rizk.
“The question about CDHP is whether or not it works for sick individuals versus healthy individuals. This is where the providers have an issue, because providers have a tendency to see sick people,” says Rizk.
As patients age, they need potentially expensive preventive tests. If the health plan does not waive those preventive fees, CDHP members can spend much of their deductibles on tests.
“I think CDHP is a generational thing—it always will be. It works for an individual; it doesn’t work for a whole family as much,” says Rizk. “They are not as sophisticated in getting enough information so they can make rational decisions.”
Although health plan survey respondents supported CDHPs, they were also realistic about its effectiveness. Slightly more than half of health plan survey respondents said they believe CDHPs save health plans money and empower consumers, but the remaining respondents did not believe the plans empower consumers. (See Figure 5 below.)
Many health insurers, population health management companies, and assorted vendors have looked to member engagement as a key to consumerism. Getting members activated and involved in their health and healthcare will ultimately improve patient outcomes and lower costs, the theory goes.
More than 60% of respondents said their company’s member outreach is moderately or very effective, whereas nearly 10% of respondents acknowledged that their organization’s communication with members is not effective. (See Figure 6 on p. 9.)
As to how to engage members effectively, nearly 33% suggested using the best health coaching tools, and another 31% pointed to linking financial incentives for physicians to positive health decisions. (See Figure 7 on p. 9.)
On the negative side, more than 17% of health plan respondents don’t think there is any way a managed care organization can truly alter behavior.
Health plan respondents supported DM and population health programs but were unsure how to best provide those services.
Most respondents (41%) said DM saves money and improves outcomes in some cases, with the second most popular answer being that it saves money and improves health outcomes in all cases (38%). Less than 5% of managed care respondents said it doesn’t save money or improve health outcomes. (See Figure 8 at right.)
Additionally, half of health plan survey respondents plan to insource all DM and wellness programs, and one-quarter plan to outsource some DM and wellness offerings. (See Figure 9 on p. 10.)
Christobel Selecky, president and CEO of DM company LifeMasters Supported SelfCare in Irvine, CA, says health plans are looking to insource programs because they want to contain costs. The pendulum has swung to insourcing DM, but whether health plans can provide those programs effectively is in question.
“It remains to be seen whether health plans that insource disease management are going to be held at the same standards as demonstrating outcomes and whether they do it in a cost-effective way,” says Selecky.
The industry believes DM works but questions how to provide those services in the most cost-effective ways and which programs work best for which populations, she adds.
DM companies have found that their programs have the best return on investment when they’re focused on the sickest populations, but questions remain about the model, Rizk says.
One of the reasons healthcare leaders are questioning DM’s effectiveness is because of CMS’ failed Medicare Health Support project. CMS ended the project in 2008, citing unenthusiastic results. Despite that setback, about half of respondents in the health plan survey believe CMS will try another DM-based demonstration project. (See Figure 10 below.)
“It doesn’t surprise me that companies think that Medicare will take another run at DM,” says Duncan. “Politicians have indicated this. Whether the reimbursement that CMS offers will make it worthwhile for companies to participate is the real issue.”
Few respondents believe a new medical home model will feature DM companies collaborating with physicians. DMAA: The Care Continuum Alliance (formerly the Disease Management Association of America) has been promoting the medical home as a collaborative effort between physicians and DM, but most respondents to the health plan survey don’t see that happening.
More than 42% of respondents do not expect medical homes will have any effect on health plans and managed care in the next three years. Those unimpressed with the medical home were unmoved because of the lack of demonstrated findings or because they didn’t know enough about the topic. (See Figure 11 at right.)
Selecky says many in healthcare are unsure about the concept because the theory is not fleshed out yet. She understands what can be accomplished with the medical home but questions how small physician practices can implement those programs.
The uneasy relationship between health plans and physicians are punctuated throughout the survey results, which Selecky says doesn’t surprise her.
However, she adds, physicians can be a powerful ally for health plans and DM companies if the sides can come together.
“Physicians are really hypersuspicious of anything that they see as an effort by managed care to influence or manipulate them,” says Selecky. “It’s understandable why they feel that way.”
VBID may have a place in healthcare reform
Benefit design could work in Medicare
Policymakers looking for ways to cut healthcare costs and improve patient outcomes may have a friend in value-based insurance design (VBID), according to a recent white paper released by Avalere Health and the Center for Value-Based Insurance Design at the University of Michigan.
The healthcare reform debate in Washington, DC, the sputtering economy, and spiraling healthcare costs could prove to be the recipe that sparks government officials to test VBID in Medicare.
In fact, policymakers at Medicare and the Department of Veterans Affairs are looking at VBID. Michigan Sen. Debbie Stabenow and Rep. John Dingell (D-MI) held a briefing at the Capitol in February to discuss the concept.
In response to the growing tide, Avalere Health and the Center for Value-Based Insurance Design at the University of Michigan conducted the analysis to see whether Medicare could use the concept with its Medicare Part D beneficiaries.
A. Mark Fendrick, MD, codirector of the University of Michigan’s Center for Value-Based Insurance Design in Ann Arbor and one of the cofounders of the VBID concept, says the white paper answers legal questions surrounding discrimination and confidentiality.
Tanisha Carino, PhD, vice president at Avalere Health in Washington, DC, and coauthor of the paper, says the current Medicare structure does not take a long-term view of saving money on prescription drugs or downstream medical costs. This is especially true for prescription drug plans (PDP) that offer drug-only coverage.
Rather than raising copays and passing costs onto members, Fendrick says VBID is a “clinically nuanced” solution. “While there are major initiatives to cut, cut, cut, we have the ability to preserve the baby while throwing out the bathwater of waste and inefficiency,” says Fendrick.
VBID supporters in the commercial setting also hope that getting Medicare on the VBID train could serve as a catalyst in the private sector.
In this age of cost shifting, VBID takes a different approach than the high-deductible health plans that have become common. The concept behind VBID is to lower medication and service costs for clinically beneficial services for patients with chronic illnesses such as diabetes and hypertension.
For example, a diabetic patient doesn’t have to pay for insulin and other evidence-based diabetic drugs. Lowering or removing copays for these services and drugs removes cost barriers and makes patients more apt to comply with their medication, resulting in better outcomes. VBID addresses the objectives of cost containment and quality improvement by promoting fiscally responsible, clinically sensitive cost sharing, according to the white paper.
Marriott and Pitney Bowes, pioneers in VBID, eliminated cost sharing associated with diabetes medications and achieved positive cost and quality outcomes. However, early users have also faced barriers, such as employee backlash because some employees were charged more for their medications, while also needing the technology platform and data exchange necessary.
“Tailoring benefit design for unique needs of patients is what everybody wants,” explains Carino. “But when employers in the private sector have gone to implement this, they have faced the hurdle of the data infrastructure to identify the right people and track their progress.”
VBID advocates and policymakers think the concept could be a winner with the Medicare population. Twenty-three percent of Medicare’s 26 million beneficiaries have five or more chronic conditions, which accounts for nearly 70% of the program’s spending.
Medicare beneficiaries are more likely to have a chronic illness and take multiple medications than the commercial population. Costs can create a barrier to medication compliance for beneficiaries. On average, the Medicare Part D population takes five prescription drugs per day, and nearly 20% are not able to fill a prescription or delay filling a prescription because of cost, according to the white paper.
“You have the potential to have an even greater impact [in the Medicare population] because the sicker the beneficiary is and the more you can target a value-based insurance design, the better the outcomes are likely to be,” says Lisa Murphy, manager at Avalere Health and coauthor of the paper.
Part D could benefit from VBID
Medicare prescription drug benefits provide coverage of most outpatient drugs for Medicare beneficiaries.
Beneficiaries who choose to take part in Part D coverage can enroll in a PDP that offers drug-only coverage or select a Medicare Advantage prescription drug (MA-PD) plan that provides medical and drug benefits. As of January, 17.4 million enrollees were in PDPs and 8.8 million were in an MA-PD.
Beneficiaries in either plan are also enrolled in a medication therapy management program (MTMP) developed in cooperation with pharmacists and physicians to help patients improve medication compliance and reduce adverse events.
Another initiative that focuses on promoting better medication use for beneficiaries with chronic conditions is chronic special needs plans (C-SNP), which are for dual eligibles who are institutionalized and have severe or disabling chronic conditions.
The focus of the two programs is to promote better medication use for beneficiaries with chronic conditions, which researchers say mirror the goal of VBID.
To research whether VBID would work in Medicare Part D prescription drug benefit programs and C-SNPs, researchers explored four factors that could affect the speed and scope of implementation: the potential size of the Medicare Part D population that would be affected (using diabetes as a test case), CMS’ authority to change policy within existing statutes and regulations, requirements for implementation, and political support.
Three options are possible now
Researchers presented five options, three of which can be implemented immediately. (See Figure 12 below.) For the options that would not need regulatory changes, Murphy says VBID could get a boost from Congress and CMS if they create incentives to get insurers to try VBID.
“Even easier to implement than incentives, just positive statements from CMS encouraging plans to do this, saying that this is something CMS might look favorably on, I think would go a long way in easing concerns plans might be having about whether this would be approved,” says Murphy.
The researchers found that options 1 and 3 would affect the largest number of diabetes beneficiaries (6 million). (See Figure 13 on p. 14.)
Option 1, targeting specific drugs or drug classes, is an option CMS can implement immediately, and one Part D plan, UnitedHealthcare’s Senior Dimensions, already has a similar program. Although option 1 is doable, stakeholders and beneficiaries might object to CMS choosing which drug classes to reduce or eliminate copays. “Though it looks on the surface that it could be much easier, it’s probably politically much more charged,” says Murphy.
Option 2 is also available immediately and targets beneficiaries with high annual drug spending who may benefit from an intervention, but only three PDPs offer gap coverage for brand-name medications. Researchers suggested that policymakers may want to create new incentives to encourage more Plan D plans to test option 2.
The third option now available, option 5, targets C-SNPs and would not affect as many beneficiaries as the other two options (less than 268,000 beneficiaries). However, researchers suggested the small patient population could be the ideal first step in introducing VBID to the Part D population. CMS and policymakers could encourage C-SNPs to incorporate VBID into their benefit designs and collect data on adherence and outcomes for CMS and policymakers to review before expanding VBID to a larger Part D population.
Although options 1, 2, and 5 would not need any policy or operational modifications, options 3 and 4 would need changes. (See Figure 14 on p. 14.)
For example, reducing cost sharing for enrollees with chronic conditions or for those enrolled in MTMPs would require CMS to revise its nondiscrimination clause.
Despite that barrier, the researchers suggested that option 4 could benefit MTMP participation and enrollment and improve medication adherence. Expanding this option could also improve health outcomes and lower overall costs.
The final option, option 3, could potentially reach numerous Part D beneficiaries and save costs, but the barrier of new legislation or regulations makes this option less desirable than the others.
“There would have to be some kind of mechanism put in place to allow plans to do that,” says Murphy. “It would be easier for MA-PD to do because they would have medical claims data and they would be able to identify certain claims that correspond to the chronic condition you’re trying to target. On the PDP side, without medical claims, they would need to get that information from CMS for the fee-for-service population.”
If VBID has such potential in the Medicare population, why have insurers in Part D not tried the concept? The researchers gave three reasons:
- Incorporating VBID in Part D could create adverse selection, meaning sicker beneficiaries could be drawn to those plans
- Plans may believe that reducing cost sharing for some beneficiaries or drugs will cause plans to increase costs for other beneficiaries
- PDPs, which offer drug benefits only, do not have the financial incentives because they won’t benefit from the lower medical costs
Despite those concerns, the white paper shows VBID’s potential in the Medicare population. “VBI
- Providers Lag as Consumers Set Agenda
- ICD-10 Delay Alters Provider, Vendor Prep
- Esther Dyson Launches Population Health Challenge
- Crisis Spurs Healthcare Payment Reform in Arkansas
- Payment Reform Naysayers 'Better Wake Up'
- Look Beyond Nurse-Patient Ratios
- HIT Leaders Want Flexibility, Transparency from Next HHS Chief
- As Hospitalist Patient Loads Rise, So Do Hospital Costs
- Reduce Readmissions by Activating Patients to Do 'Self-Care'
- Advance Directives: Let's Make a Law