Summary: Studies of interdisciplinary health care teams have demonstrated that use of these teams can lead to improvements in the quality of primary care, but their impact on total health care costs and utilization has not yet received sufficient attention. Still, available evidence suggests that these teams may help expand the nation's capacity to provide primary care services, which is much needed due to a shortage of physicians and other primary care providers. But doing so quickly will require the financial support of federal, state, and private payers, as well as an investment of time by health care providers.
By Sarah Klein
As the landmark health reform law goes into effect, bringing millions of uninsured Americans onto insurance rolls over the next five years, demand for primary care services will increase. So, too, will demand for more accessible, effective, and efficient models of primary care.
Rather than hiring more primary care physicians, many medical practices, health centers, and other primary care settings have been experimenting with innovative models of care that both extend the reach of primary care physicians and increase the quality of ambulatory services. Many of these models bring pharmacists, social workers, nurses, and nurse practitioners to primary care practices. With them comes a new set of skills that can improve care and lower costs for patients with depression, physical disabilities, and other conditions that have proven difficult to treat in primary care settings.
Using teams of multidisciplinary providers in a primary care setting "allows far more intensive intake and assessments than a physician alone could do," says Robert J. Master, M.D., president and CEO of the Commonwealth Care Alliance, a Boston-based nonprofit health plan and delivery network that has introduced nurse practitioners to 25 community-based medical practices in Massachusetts.[1] The nurse practitioners, whose salaries are generally lower than physicians, lead teams of providers caring for patients with physical disabilities and complex medical conditions. They have assumed responsibility for assessments of patients, ongoing management, and first call while the physicians manage inpatient care.
The composition and caseload of the nurse practitioners' teams vary significantly with the condition and acuity of patients (see table). Community health workers are added to some teams to help engage Medicaid-eligible patients who are distrusting of providers, while personal care attendants help homebound patients with the tasks of daily living. The nurse practitioners determine the level of care needed and authorize all required services, saving the time and cost of preauthorization procedures.
Patient population |
Typical composition of team |
Ratio of nurse practitioners to patients |
Patients with one or more physical disabilities |
Nurse practitioner, physician, physical therapist, social worker, and durable medical equipment coordinator |
1:40 |
Frail elderly patient (may be homebound) |
Nurse practitioner, physician, social worker, and personal care attendant |
1:45 |
Medicaid-eligible patients with complex chronic illness(es) and behavioral health or substance abuses issues |
Nurse practitioner, physician, social worker, and community health worker |
1:90 |
Source: Commonwealth Care Alliance, August 2010
The Commonwealth Care Alliance invested heavily in the model—spending approximately $4 million on 25 practices, many of which are located in low-income, safety net clinics. The investment, which covers the cost of hiring the nurse practitioners by the primary care practices and investing in infrastructure such as electronic medical records, is more than offset in reductions in hospitalizations for preventable conditions as well as delays in nursing home placements, Master says. The number of hospital days per year per Commonwealth Care Alliance member who is dually eligible for Medicare and Medicaid is 2.0—77 percent lower than the rate of hospital days (3.6) per dually eligible patient enrolled in the Medicare fee-for-service program.[2] Within the Commonwealth Care Alliance program, the percentage of nursing home–certifiable patients permanently placed in the nursing home per year is 8.5 percent, compared with the Massachusetts rate of 12 percent. "We believe the return on investment is substantial," Master says.
Extensive Redesign Required
Having an integrated delivery system and local insurers willing to invest in the redesign of primary care practices has enabled some providers to move forward with team-based models of care.
After receiving a grant from a local insurer in 2008, Fairview Health Services, a nonprofit, integrated delivery system with seven hospitals and more than 100 primary and specialty care clinics serving residents of Minneapolis–St. Paul, began experimenting with a team-based model of care in two (and then four) of the 41 primary care clinics affiliated with its Eagan, Minn., operations. The teams include health coaches, pharmacists, and social workers or psychologists, as well as nurses who help manage chronic conditions such as hypertension by following protocols set by physicians.
The program has redefined the roles of clinic employees. Receptionists, who arrive before the clinic opens and take calls from patients, brief the team about the unfolding day. Health coaches use motivational interviewing techniques to find ways to encourage patients to pursue overdue mammograms or quit smoking. And instead of seeing a succession of patients in one-on-one visits, physicians have begun to proactively manage a panel of patients, looking for high-risk ones in need of care. "It's expanding a clinician's ability to serve a population of patients," says David Moen, M.D., executive medical director of innovation and network development, of the pilot.
Because the pay of clinic staff is based on their performance on measures of cost, quality, and patient satisfaction, rather than productivity, more care has migrated to the Internet and telephone. And costs appear to be holding steady. "What we've seen with our early data with these four clinics, we've been able to track that there's been a flattening of the total cost-of-care curve," for the 20,000 patients attributed to those clinics, Moen says.
Fairview plans to extend the model to all 41 primary care clinics in Eagan by the end of 2010. To finance the expansion, it is renegotiating its compensation plan for its employed physicians as well as its contracts with insurers. The compensation model for physicians will focus on panel size, acuity, outcomes, patient engagement, and the total cost of care. The new contracts with insurers will enable the clinic to share in the savings that ensue from better-coordinated care.
Fairview has also been experimenting with including patients in its team-based approach. Fairview has created a Web-based learning community for the families of children newly diagnosed with diabetes, which is only accessible to clinicians, patients, and their parents. The forum is supported by diabetes educators, endocrinologists, and other families with diabetes. "They're closed communities, and in those communities you can also access your care team. It's fascinating to watch. You just put the patients and the clinicians in the room and they create the solution," Moen says. Fairview plans to expand this service for families of children with autism.
Models Reach Underserved Populations
A team-based model has similarly improved the care of depression in patients participating in the DIAMOND program, which stands for "Depression Improvement Across Minnesota, Offering a New Direction." The program enhances the screening and treatment of depression among patients in 82 primary care practices in Minnesota by enabling those practices to hire case managers and consulting psychiatrists to oversee the care and follow-up of patients who have screened positive for chronic or severe depression.
Primary care physicians introduce patients to the care managers, who maintain frequent contact with the patients, educating them about depression and checking for side effects of medication and other issues that might warrant follow-up care. The consulting psychiatrist reviews cases in which patients are not improving and may recommend modifications in their medication regimens to their primary care physician. In more complicated cases, a patient may be referred to a psychiatrist for care.
Of the 1,759 patients contacted after being in the program for at least six months, 45 percent are in remission. Another 16 percent have seen at least a 50 percent reduction in severity of their depression. These results are 10 times better than the improvement noted under the usual primary care treatment in Minnesota, according to the Institute for Clinical Systems Improvement (ICSI), which brought psychiatrists, employers, health plans, and patients together to develop the program.
The IMPACT ("Improving Mood: Promoting Access to Collaborative Program") program upon which the DIAMOND program is based reported 45 percent of patients had a 50 percent or greater reduction in symptoms of depression, compared with 19 percent of patients in the usual care group. [3]. Unlike IMPACT, which required nurses with psychiatric experience to perform the role of care manager, DIAMOND has hired medical assistants, social workers, licensed practical nurses, and dieticians to perform as care managers. Early results suggest that medical assistants have the best results and the highest caseloads. The medical assistants "are extremely familiar with how primary care clinics function" and seem to handle anything that comes their way, says Nancy Jaeckels, vice president of member relations and strategic initiatives for ICSI.
ICSI is ready to spread the model to more clinics in Minnesota, but must first win the support of insurers, who now make an unspecified monthly payment to clinics for each patient enrolled in the program. The insurers want to see evidence of the program's cost-effectiveness first, Jaeckels says. If the DIAMOND program follows the timetable of the IMPACT program on which it is based, such data will not be available until 2011, its third year. The IMPACT program saved $3,363 per patient on average during four years, but the savings did not accrue until years three and four of follow-up. Part of the reason is the first-year cost of the intervention, which was $522.[4]. Still, early evidence suggests the DIAMOND program is having a short-term, positive impact.
Another serious issue is that providers participating in the DIAMOND program are not reimbursed by all payers, including Medicare and Medicaid. "Most of them are eating the cost," of providing care management services to those patients in those programs, Jaeckels says. Some are pursuing grants to cover the costs, an approach that is not sustainable in the long term.
Innovation First, Payment Later
Nonetheless, many who pursue new models of care often do so in hopes that insurers will recognize their value and change their coverage accordingly. Hilde Berdine, Pharm.D., assistant professor of pharmacy practice at Duquesne University in Pittsburgh, Pa., serves as a pharmacist in a local, two-physician primary care practice. The physicians refer patients with high blood pressure, high cholesterol, and diabetes to her to improve adherence to medication regimens and reduce side effects of medication. Her services are rarely reimbursed because many pharmacy benefit programs do not recognize her as a provider, though she has sought such credentialing. Instead, she provides the service to give her students a model of the future of pharmacy services. "I am not unique in that respect," says Berdine. "There are pharmacists doing this all over the country, but it is in the spirit of trying to develop practice [models] that we think are going to take hold in the future," as their impact on quality and costs becomes apparent.
Gaining an entré into primary care practices also requires reassuring physicians that their own livelihoods or quality of care are not at risk, says Brian Isetts, Ph.D., professor in the department of pharmaceutical care and health services at the University of Minnesota. Isetts helped establish a consulting pharmacist program for Fairview Health Services in Minnesota 10 years ago. When speaking to physicians about the prospect of such programs, he often begins his presentations by pointing out that pharmacists who do medication management are not seeking independent prescribing privileges, nor trying to implement therapeutic substitutions on behalf of insurance companies, or changing doses without physician input or consent. Once he clarifies that, physicians often welcome pharmacists' assistance and begin sending their more challenging patients with diabetes, heart failure, and other conditions their way. Fairview estimates that its medication therapy management program has avoided 48,000 medication therapy problems and 17,000 events, including hospitalizations, emergency department visits, and clinic visits.
Many physicians welcome the efficiency team-based models bring. At Geisinger Health System in Danville, Penn., nurse case managers employed by Geisinger's health plan were added to primary care practices to oversee care for high-risk Medicare enrollees. Sharing oversight of these patients among health plan employees enables nurses in the practice to provide the same sort of case management services for non-Medicare patients.
Adding Robots to Teams
Geisinger has also automated many staff functions, which further frees up employees for direct patient care. Automated systems identify and call patients about important screening exams. One of two staffers is available to answer questions that may arise, but the orders for recommended tests are automated by the system. "What this has done is allow us to increase our preventive services," says Thomas R. Graf, M.D., chairman of community practice service for Geisinger. The number of patients who received all recommended care increased by 250 percent since the program began three years ago. The system, which monitors tests for 210,000 patients, and the reporting of all results through an electronic medical record has enabled physicians and other health care professionals to concentrate on complex medical decisions and patient relationships. Physicians find that their work is more interesting and challenging, patient care is better managed, and patients are happier, Graf says.
The program, which will have been spread to 38 primary care practices by October, has already had a dramatic impact on quality and cost measures. The rates of hospitalizations per 1,000 patients between January and June of 2009 in medical homes fell 22.8 percent, compared with a control group of similar Medicare patients. The rate of readmissions during the same period fell by 23 percent. Data from a two-year period ending in December 2008 for 11 sites show cost savings of 4.5 percent relative to the comparison group after accounting for the additional start-up and operational costs of the technology. Graf stresses that even physicians with less automated systems can get results. "Most of the stuff we do, two doctors, a nurse, and a paper and pencil could do," he says.
Having a partner helps. Independent physicians may find one in hospitals that are keen to lower readmission rates. Insurers also may help. Physicians in the Geisinger system received a financial incentive to participate in the program. Medical practices received start-up money of $5,000 to $10,000 per month in the first year to hire staff, install phone lines, or otherwise improve infrastructure. Physicians also received $18,000 the first year; the second year, they received the same amount, but payments required them to achieve certain performance levels on cost and quality measures. There is also a shared savings program between the health plan and the providers. (The total savings for the first 11 sites was $2.5 million.)
While money may provide a powerful incentive for many physicians, developing teams can be challenging. Communication among team members who are trained in different disciplines is crucial to effective team work, says Ronald Stock, M.D., medical director of PeaceHealth, a nonprofit Catholic health system with six hospitals serving residents of Washington State, Alaska, and Oregon. Some of its hospital-based outpatient clinics are staffed by teams of geriatricians, nurse practitioners, nurses, social workers, dieticians, and pharmacists, who focus on frail elderly patients with multiple chronic health conditions.
Developing, nurturing, and sustaining a team is an ongoing effort. "Not only do we come from different disciplines, but we all as individuals have communication styles," Stock says. Physician leadership must be encouraged and training for all staff is required. PeaceHealth uses a variety of means to improve and test team cohesion. "When we developed our interdisciplinary team, we actually had a consultant come in and help us learn how to communicate," Stock says. PeaceHealth also uses a survey to assess key aspects of teamwork: whether the team is operating with clearly defined goals and an understanding of the means to accomplish them, whether team members have clearly defined roles and expectations of one another, and whether they handle conflict maturely. Patient care conferences, which are conducted weekly and involve entire teams, provide a venue to practice communicating, he says.
While the transition to a new model of care is difficult and some systems find that staff initially resist change, many say once employees experience the new model, they view the old ones differently. "They all, to a person, from what we could tell, think their jobs are 10 times better and their biggest fear is that we would go back," Moen says.
Notes
(1) No relation to The Commonwealth Fund
(2) Data supplied by Commonwealth Care Alliance for patient populations with a similar risk score.
(3) J. Unützer, W. Katon, C. M. Callahan et al., Collaborative Care Management of Late-Life Depression in the Primary Care Setting: A Randomized Controlled Trial, Journal of the American Medical Association, December 11, 2002 288(22):2836–45.
(4) J. Unützer. W. J. Katon, M. Fan et al., Longer-Term Cost Effects of Collaborative Care for Late-Life Depression, American Journal of Managed Care, February 2008 14(2):95–100.