Problem Addressed

• Inadequate support: Physicians often lack the time and resources to provide ongoing self-management support to individuals with chronic conditions, including diabetes, asthma, cardiovascular disease, and depression. While short-term educational interventions have been shown to help patients control chronic conditions such as diabetes, such gains often evaporate within a few months of the program's termination.¹
• Unrealized potential of community-based providers: Patients visit pharmacies more often than any other health care setting, placing community-based pharmacists in a prime position to provide ongoing self-management support to those with chronic illnesses.² Studies suggest that pharmacist-led interventions can generate improved outcomes for chronic conditions such as asthma,³ but relatively few pharmacists have been trained or have the incentive to provide such support. Other community-based providers, including clinic-based nurses, may also be in a position to support these patients.

Description of the Innovative Activity

• Financial incentives to participate: Employers typically create a meaningful financial incentive to participate, consisting of reduced or waived copayments for drugs and supplies related to the targeted condition(s), and/or reductions in the participant's share of the monthly insurance premium. While incentives vary across employers and conditions, most employers create at least a $5 to $10 difference in copayments for generic drugs (e.g., $0 for participants, $5 to $10 for nonparticipants) and a $20 or greater differential for preferred brand-name drugs (e.g., $0 to $20 for participants, $20 to $40 for nonparticipants). Monthly premium differentials also vary but can be as high as $50 to $75. (See the Adoption Considerations section for more information on legal issues surrounding incentives.)
• Marketing and enrollment, typically with support from health plan or vendor: The program typically operates as part of the health benefit offered by self-funded employers, many of which work with a health plan or vendor that provides administrative services that support the program. Plan materials describe the program, its potential benefits to the individual's health and well-being, and any financial incentives to participate. Those interested complete an enrollment and consent form that specifies the conditions for participation, including that ongoing eligibility for incentives depends on the employee's active participation in the program (e.g., attending scheduled sessions with the care manager).
• Initial educational sessions: In communities where health education centers already exist (typically hospital-based centers), the program explicitly covers—and requires participant attendance at—group classes led by certified educators with expertise in the condition. For example, the City of Asheville and Mission Hospitals require participants to attend several group sessions (typically lasting 60 to 90 minutes each) led by certified educators who teach self-management principles related to diabetes, asthma, hypertension, and dyslipidemia. In communities without such centers, care managers typically provide this education as part of their regular meetings with participants. (See below for more details on these sessions.)
• Assignment to care manager, usually a pharmacist: Employers assign a care manager to each participant, typically a pharmacist who works in the community or a hospital-based clinic. (One employer—Hickory Springs Manufacturing Company—uses nurses who staff onsite clinics and community-based pharmacists as care managers; the nurses and pharmacists refer enrollees to each other for issues outside their area of expertise.) Participants typically are assigned to care managers located close to their work or home; the enrollment process gives them a chance to express preferences about the location of their sessions.
• Regular meetings with care manager to support self-management: The core of the program involves periodic sessions with care managers who help the individual manage his or her condition(s). The care manager focuses on promoting self-management behaviors (e.g., appropriate diet, adequate physical activity, refraining from smoking) and on monitoring medications to ensure the individual adheres to the appropriate regimen. Details about these sessions are outlined below:
  • Meeting logistics: The care manager meets with each participant at least quarterly and as often as once a month. Right after enrollment, meetings tend to occur more frequently, allowing the care manager to identify and address pressing issues. As these issues stabilize, meeting frequency often reverts to once every 2 or 3 months. Meetings generally occur in the care manager's office, although in some cases sponsoring employers prefer the sessions be held at the worksite. For example, at Hickory Springs Manufacturing Company's main plant, in-house nurses conduct sessions at an onsite clinic, and some employers ask participating pharmacists to hold periodic "office hours" at the worksite. Most sessions involve face-to-face meetings, although some smaller employer sites (which do not have enough potential enrollees to justify training local care managers) have initiated telephone-based sessions with out-of-area care managers. For example, Hickory Springs uses this approach at some of its 60 smaller plants located outside Hickory, NC.
  • Initial session to assess current situation: At the initial session (which typically lasts an hour), the care manager establishes a relationship with the individual and gathers background information, including relevant medical history. For example, with the depression program, the care manager assesses the patient's current mental status, ongoing stressors, psychiatric history, social and family histories, and medical history, including allergies and comorbid conditions. As appropriate, the care manager conducts a brief physical assessment (e.g., height, weight, blood pressure) and asks about current providers and medications (including any problems with these medications), as well as any condition-related goals or action plans (e.g., an asthma action plan) that have been previously established. The care manager also begins to assess the individual's baseline knowledge of the condition and ability to self-manage it, and works with the individual to develop treatment goals and an action plan, if needed. 
  • Subsequent meetings to monitor progress, develop skills: During subsequent meetings (which typically last 30 minutes), the care manager works with the individual to review the current medication regimen and assess progress. Care managers also periodically take or review relevant measurements, such as blood pressure, lipid levels, and spirometry or peak flow meter results. For example, with the asthma program, the care manager assesses the patient's medication use, including usage of reliever and controller medication, and reviews the patient's inhaler technique, symptom frequency, and spirometry and/or peak flow meter results.
• Post-visit communication with physicians: As necessary, care managers communicate relevant findings and recommendations to the participant's regular physician, typically by fax after the session occurs. (Time-sensitive communications occur by telephone as needed.) Care managers send a report only when issues exist, as most physicians do not feel the need to correspond with care managers if everything is fine. Recommendations tend to be worded as "suggestions" to the patient to followup with the doctor (e.g., to discuss the merits of a medication change). If the care manager has a longstanding relationship with the physician, suggestions might be more specific and pointed, including recommending a particular drug and dosage. At subsequent meetings, care managers followup with the patient to see if the doctor followed the advice. If not, the care manager works with the patient to encourage him or her to talk with the physician about the issue at the next visit.
• Support from electronic system: In some cases, the care managers have access to electronic systems that can be used for documentation purposes and that provide updated medical and prescription data and highlight gaps in care based on national standards. In other situations, the care manager uses paper charts and may not have access to a complete medical and prescription claims history.
• Monitoring of (and feedback to) care managers: In those instances where electronic systems are used, program managers track indicators of care manager performance, including frequency of visits, whether the appropriate data are being collected at each visit, and how the participant's health indicators (e.g., blood pressure, weight) have changed over time. This information is used to monitor care manager performance, with relevant information fed back through a series of monthly and quarterly reports. These reports also include lists of individual patients who may need greater support, such as those whose blood pressure has consistently remained above established goals.

References/Related Articles

Contact the Innovator

Results

• Greater adherence to recommended care, self-management behaviors: The program improved adherence to recommended care processes and self-management behaviors related to asthma and diabetes, as outlined below:
  • Diabetes: For the City of Asheville and Mission Hospitals, participants exhibited substantial improvements in four guideline-recommended care processes and self-management behaviors, including having a hemoglobin A1c measurement in the past 6 months (with the percentage meeting this recommendation rising by 18 percent), having a foot exam in the last 6 months (43 percent), taking an ACE inhibitor (38 percent), and performing self-testing of blood sugar (92 percent).⁴ At Hickory Springs Manufacturing Company, the proportion of participants regularly testing their blood glucose levels increased from 79 percent at baseline to 97 percent three years later.⁵
  • Asthma: For the City of Asheville and Mission Hospitals, the percentage of enrollees having asthma treatment plans increased from 63 percent at baseline to 99 percent at followup.³
• Better disease control: The program has improved outcomes related to disease control across all four targeted conditions, as outlined below:
  • Diabetes: For the City of Asheville and Mission Hospitals, mean hemoglobin A1c levels decreased consistently over a 3.5-year period, with between 57.7 and 81.8 percent of patients demonstrating improvement versus baseline at each of seven measurements. Cholesterol levels also improved at each measurement compared to baseline, with modest declines in low-density lipoprotein (LDL) and increases in high-density lipoprotein levels.⁴ For Hickory Springs Manufacturing Company, the proportion of enrollees achieving blood glucose control (defined as a hemoglobin A1c level below 7 percent) increased from 38 percent at enrollment to 53 percent 3 years later. Improvements also occurred in the proportion achieving recommended levels of LDL cholesterol (46 to 57 percent), systolic blood pressure (55 to 72 percent), and diastolic blood pressure (60 to 71 percent).⁵
  • Asthma: Among City of Asheville and Mission Hospitals participants in the program at least a year, the percentage classified as having severe or moderate persistent asthma fell from 82 percent at baseline to 49 percent after 5 years. Spirometry results and symptom frequency also improved significantly over the 5-year period, while quality-of-life scores increased in every major category.³
  • Cardiovascular disease: Among City of Asheville and Mission Hospitals participants, average systolic and diastolic blood pressure declined by a statistically significant amount, with the proportion achieving their blood pressure goal increasing from 40.2 to 67.4 percent. Average LDL cholesterol levels fell from 127.2 to 108.3 mg/dL, with the percentage achieving their goal rising from 49.9 to 74.6 percent. Statistically significant declines also occurred in average total cholesterol and serum triglycerides.⁶
  • Depression: Among City of Asheville and Mission Hospitals participants in the program at least a year, depression severity scores fell by a statistically significant amount, from 11.5 to 5.3 (as measured by the Patient Health Questionnaire-9). Patients with severe depression at baseline experienced the largest improvements. Overall, more than two-thirds (68 percent) of participants exhibited a positive response to treatment.⁷
• Lower utilization and costs: The program reduced total health care costs across all four targeted conditions, as outlined below:
  • Diabetes: For the City of Asheville and Mission Hospitals, total average annual direct medical costs per enrollee decreased by $1,200 to $1,872, compared to baseline. Overall prescription costs increased, but inpatient and outpatient physician service costs fell by a greater amount.⁴ For Hickory Springs Manufacturing Company, average total health plan costs conservatively decreased by roughly $2,700 per person per year. Spending on prescriptions rose significantly, but this increase was offset by reductions in other medical expenses.⁵
  • Asthma: For City of Asheville and Mission Hospitals participants, annual emergency department (ED) visits and hospitalizations per 100 enrollees fell dramatically, from 22 before enrollment to 3 afterwards. The program yielded an estimated net savings in direct medical costs of $725 per enrollee per year (after accounting for program-related costs).³
  • Cardiovascular disease: For City of Asheville and Mission Hospitals participants, cardiovascular events (e.g., heart attacks, ED visits, hospitalizations) fell by 53 percent and cardiovascular-related medical costs fell by 46.5 percent over a 6-year period. Cardiovascular drug costs nearly tripled, but this increase was offset by a drop in the costs of cardiovascular events.⁶
  • Depression: For City of Asheville and Mission Hospitals participants, average annual health care costs rose significantly less than expected, from $7,935 per enrollee at baseline to $8,040 a year later (including the costs of the program). This figure is 11 percent below projected costs of $9,023 per enrollee.⁷
• Higher productivity: The program enhanced productivity among employees with diabetes and asthma, as outlined below:
  • Diabetes: Average annual sick days in a group of 37 City of Asheville employees fell consistently over a 5-year period, from an average of 12.6 days per employee at baseline to roughly half that level 5 years later. This decline translated into an estimated annual productivity benefit of $18,000 for these 37 employees (just under $500 per employee).⁴
  • Asthma: After enrolling in the program, City of Asheville and Mission Hospital employees reported missing a half day of work each year due to asthma, well below the 2.5-day average prior to enrollment. Similar declines occurred in reported hours lost each year due to "presenteeism" (time on the job when asthma interfered with the ability to work), which fell from 66.5 hours per participating employee before enrollment to 16.8 hours afterward. These reductions generated approximately $1,230 per participant per year in indirect cost-savings for the sponsoring employers.³
• Significant ROI: The four programs implemented by the City of Asheville and Mission Hospitals (for diabetes, asthma, cardiovascular risk factors, and depression) generated average benefits of approximately $4.00 for every dollar spent. At Hickory Springs Manufacturing Company, the diabetes program generated an estimated $8.48 in savings for every dollar spent, with savings based on comparisons to expected cost trends (which historically had risen 8 percent a year). The ROI drops to $4.89 per dollar spent if savings estimates conservatively assume that no increase in expenses would have occurred in the absence of the program.⁵

Evidence Rating (What is this?)

Context of the Innovation

Planning and Development Process

• "Selling" local physicians through respected peer: From the outset, the City of Asheville had the physician advisor to its health plan attend all planning meetings about the proposed program. Well-respected within the community, the advisor sent a letter to community-based physicians about the demonstration project, emphasizing its potential to help them and their patients. The letter did not oversell the program, instead urging physicians to give it a chance. Program leaders also gave a presentation at a regional physician association meeting, covering the same basic information included in the letter. Area physicians generally responded favorably, with little or no pushback.
• Recruiting and training pharmacists: Program leaders alerted community-based pharmacists about the opportunity to participate in the demonstration projects. Roughly 20 pharmacists volunteered, mostly from independent pharmacies. Representatives of the pharmacy schools at the University of North Carolina and Campbell University sponsored a 20- to 30-hour training session to review key concepts and the latest findings related to diabetes medication management and patient education.
• Formalizing program after successful demonstration: Approximately 3 months into the 6-month trial, the city benefits and risk manager decided to institutionalize the program after receiving positive feedback from participating employees, including several who came to his office to thank him personally for sponsoring the program. The city manager worked with participating pharmacists to compensate them for their time.
• Expansion to Mission Hospitals employees and dependents: Roughly 2 years after the demonstration project launched, leaders of Mission Hospitals (as the organization became known after the merger of Memorial Mission Hospital and St. Joseph Hospital) decided to adopt the diabetes program for the 15,000 enrollees of its self-funded health plan. These leaders were swayed by the substantial evidence from the City of Asheville that the program improved outcomes (e.g., blood pressure, blood glucose, cholesterol) and reduced costs.
• Advocating for third-party reimbursement: Several years after the demonstration project, leaders of the state pharmacy association successfully lobbied the state legislature to pass legislation that allowed pharmacists meeting certain requirements (including having a supervising physician) to qualify for reimbursement from third-party payers, similar to nurse practitioners and physician assistants. Under the legislation, qualified pharmacists become known as "clinical pharmacist practitioners"; the legislation did not require the State or any other payer to cover these services, but allowed them to do so at their discretion. Those lobbying used the published results from the diabetes demonstration project to convince legislators to pass the law. Several years later, pharmacy leaders throughout the country lobbied for the same thing on the national level, convincing Congress to include medication therapy management by certified pharmacists as a reimbursable service in the Medicare Modernization Act of 2003. These leaders again used published evidence from the Asheville Project to make their case.
• Expansion to other conditions: Based on the success of the diabetes program, the City of Asheville and Mission Hospitals expanded the concept to other conditions, beginning with asthma in January 1999. The City of Asheville launched the cardiovascular program in 2000, with Mission Hospitals following a year later. Both organizations launched the depression program in July 2006. As the program expanded, the pharmacy schools revamped the training program to incorporate the new conditions. Participating pharmacists who had previously attended the training session took periodic refresher courses via teleconference, while pharmacists new to the program completed the latest version of the full 20- to 30-hour curriculum.
• Spread to other organizations: The same basic approach used in the Asheville Project has been adapted for use by employers throughout North Carolina and the nation. See the Use by Other Organizations section for more details.

Resources Used and Skills Needed

• Staffing: As a general rule of thumb, a half-time care manager (devoting 20 hours a week to the program) can support a caseload of approximately 300 patients per year. The program also requires staff time to analyze data and generate reports for care managers and sponsoring employers; outside vendors often handle these responsibilities under contract with the sponsoring employer.
• Costs: Data on program costs are unavailable; costs vary significantly across conditions and sponsoring employers and generally depend on the number of visits with care managers, the number of prescribed medications, and the size of the employer incentives.

Funding Sources

Getting Started with This Innovation

• Estimate potential savings as self-insured employer: Self-insured employers will likely be the early adopters of this type of program, because they (rather than their health plan) reap the direct benefits of resulting cost reductions and productivity enhancements. As a first step, employers should analyze internal claims data and compare these data to national or regional benchmarks. Key metrics include the proportion of the covered population with specific chronic conditions, along with annual health care spending and productivity losses (e.g., absenteeism, presenteeism) for those with these conditions. This analysis helps in determining the merits of adopting the program and in choosing the best target condition(s).
• Elicit input from—and explain program to—area physicians: Some physicians might react negatively to the pharmacists' expanded role in managing the care of their patients. To minimize the potential for this type of response, elicit the input of area physicians when designing the program, and have a well-respected physician explain the program to them before the launch.
• Offer adequate incentives: The incentive must capture the attention of employees and dependents. At a minimum, create copayment differentials of $5 to $10 for generic medications and $20 for brand-name drugs. In addition to (or instead of) copayment modifications, consider creating meaningful ($50 to $75) differentials in monthly premiums, which may appeal to those who manage their condition(s) without medications. Experience to date suggests that premium differentials encourage broad participation, attracting 80 to 90 percent of eligible individuals to the program (vs. roughly one-third with copayment differentials alone).
• Consult with legal experts when structuring incentives: Various laws regulate how incentives (particularly premium differentials) must be offered so as not to discriminate against individual employees. As a general rule, every covered individual must have the opportunity to earn the incentive by meeting certain requirements, such as completing a health risk assessment, seeing a physician and getting certain tests at least once a year, and attending all care manager visits. Would-be adopters should consider having proposed incentives vetted by legal experts prior to implementation.
• Weigh merits of in-person versus phone-based counseling: Face-to-face sessions tend to be more effective than those done via telephone. However, for sites that do not have enough employees to justify training local care managers, telephone-based sessions likely make more sense. As a rule of thumb, an employer needs at least 1,000 covered lives in a community to justify in-person sessions.

Sustaining This Innovation

• Monitor individual engagement: Enrollees who repeatedly fail to adhere to program requirements (e.g., not attending appointments with care managers) should lose the financial incentives.
• Monitor and report on program impact: Sponsoring employers will likely remain committed to the program as long as it produces tangible benefits. Consequently, program leaders should regularly monitor and report in an objective manner on the program's impact. Evaluations should go beyond process measures (e.g., number of contacts) and patient satisfaction scores, focusing as well on tangible financial benefits being realized by the employer. Key metrics include the program's impact on health benefit costs and employee productivity, with data trended over time and compared to relevant benchmarks. ROI calculations also tend to resonate with company leaders.
• Monitor and provide feedback to care managers: Care managers will engage more actively with enrollees if they know their performance is being monitored and they receive constructive feedback on a regular basis. Ideally, program managers provide care managers with "report cards" that evaluate performance and provide specific suggestions for improvement, including lists of patients who may need extra support.
• Limit physician contact: Busy physicians will likely not be interested in reading about every encounter a patient has with his or her care manager. Consequently, care managers should contact physicians only when important information needs to be communicated, such as potential treatment changes to be considered.

Additional Considerations and Lessons

• Disclosure: As described in the Use by Other Organizations section below, American Health Care (for whom Dr. Bunting now works) helps employer clients implement and operate this type of program, and hence stands to benefit financially from its spread.

Use By Other Organizations

• Expansion throughout North Carolina: In early 2000, several other Asheville-area employers adopted the diabetes program pioneered by the City of Asheville and Mission Hospitals. Around the same time, American Health Care, based in California, became the pharmacy benefits manager for Mission Hospitals. During subsequent years, American Health Care leaders learned about the Asheville Project. Recognizing its potential, these leaders (pharmacists by training) created a model to commercialize the program, analyzing data from Mission Hospitals and other area employers to document its potential benefits. They approached a health benefits broker in Hickory (with whom they had a relationship), who in turn got them an audience with Hickory-based employers, including Hickory Springs Manufacturing Company. Enticed by the program's potential, Hickory Springs leaders modified the diabetes program and adopted it in 2005 for the 4,500 members of its self-funded health plan. As part of this effort, American Health Care staff developed the aforementioned electronic record and integrated it with established guidelines for diabetes care.
• Expansion throughout nation: The program has expanded nationally with the support of several organizations, including (but not limited to) American Health Care and the American Pharmacists Association. American Health Care works with roughly 60 employers in 30 states that have adopted the program. In 2005, the American Pharmacists Association Foundation secured grant funding from several large pharmaceutical companies to launch the Diabetes Ten-City Challenge, which led to implementation of the diabetes program by large employers (including municipalities) in 10 cities. The grant money funded the program for 1 year, after which many of the employers decided to pay for it on their own. The American Pharmacists Association Foundation subsequently supported adoption of the program for other conditions in these 10 cities. (More information on the Diabetes Ten-City Challenge can be found in a related profile, available at http://www.innovations.ahrq.gov/content.aspx?id=2602.) To promote further adoption, several national associations now offer training programs in local communities that allow pharmacists to become certified as care managers and hence qualify for third-party reimbursement.