Primary care is often viewed as a team-based activity. Yet many practices have experienced difficulty implementing high-performing teams1. The larger the team, the more time and energy are spent communicating among team members and the greater the probability of fumbled handoffs. Smaller teams may have advantages.

San Francisco General Hospital’s Family Health Center (FHC), a residency teaching clinic, has explored the use of small, two-person teams called “teamlets.”2,3 A teamlet consists of a clinician (physician, nurse practitioner or physician assistant) and health coach who work together consistently and collaboratively. Both teamlet members share responsibility for the health of their patient panel. The primary function of the health coach is to assist patients to gain the knowledge, skills, and confidence to self-manage their chronic conditions.

Specifically, health coaches:

  • ◊ help patients set agendas for the clinician visit

  • ◊ make sure patients understand what their clinician wants them to do

  • ◊ determine whether patients agree with their care plans

  • ◊ provide support to patients’ efforts in adopting healthy behaviors

  • ◊ assist patients to improve medication understanding and adherence

  • ◊ function as a cultural bridge, point of access, and support for their patients

The health coach model has gone through several iterations since first introduced to the FHC in 2006. In 2007–2008, all FHC medical assistants and non-professional health workers received 12 h of health coach training led by faculty physicians and were paired with first year family medicine residents for patients with diabetes. Some of these teamlets worked well. Others floundered because the coaches had little interest in their new job role. The following year, the number of coaches was reduced to three, and job roles were rearranged to give them protected time for health coaching without hiring new personnel. The quality of coaching improved greatly, but with few coaches working with about 25 physicians, the strict teamlet concept was abandoned.

Since FHC health coaching was being performed without stable teamlets, we decided in August 2008 to revive the teamlet health coaching model in a different form. In this pilot, one health coach was paired exclusively with one faculty clinician to see if the teamlet coaching model was helpful to the clinician and acceptable to patients. We describe this experience here.

THE FHC TEAMLET

Victoria Ngo, a community health worker who came to the FHC after 4 years of health coaching experience in an outpatient clinic setting, works in a teamlet with Dr. Hali Hammer. As FHC medical director, Dr. Hammer shoulders major administrative and teaching responsibilities, and sees her own continuity patients during two half-day clinic sessions each week. When the two began working together as a teamlet, Dr. Hammer had 226 patients on her continuity panel, 20% of whom declared their primary language to be Cantonese or Vietnamese. Victoria had previously completed a health coach training curriculum (www.ucsf.edu/cepc) and was paired with Dr. Hammer in part because of her fluency in these two languages in addition to English, as well as her health coaching experience. Many patients could identify with her because she comes from both the Vietnamese and Cantonese communities. Victoria has been working with Dr. Hammer during each clinic session since August 2008. Their relationship is one of mutual trust, without the ups and downs of many professional interactions. The main source of their satisfaction is working with a stable partner to help each other problem-solve their patients’ concerns rather than having to confront difficult patient issues alone.

In this description, the physician is Dr. Hammer, and the health coach is Victoria.

Each clinic session begins with a huddle lasting no longer than 10 min, during which the teamlet discusses the day’s patients. By going over the scheduled patients, the teamlet can anticipate how to best address the patients’ concerns, as well as enhance the clinical experience for the physician by sharing the burden of handling non-clinical patient issues. For example, the teamlet decides for which patients the health coach should help set the visit agenda, eliciting the patient’s concerns and notifying the patient of the physician’s issues. In the huddle, patients are identified for whom the health coach will review medications prior to the visit. During a visit, the health coach can take notes, make photocopies of documents and ensure that the visit runs smoothly. These intra-visit functions are paper-chart versions of the University of Utah Health System model of a medical assistant documenting the visit in the electronic medical record during the physician visit1. After the visit, the health coach “closes the loop,” making sure patients understand what the physician said; the coach can also help patients develop behavior-change action plans to manage their chronic conditions. Phone follow-up between visits may be as simple as checking to see if the patient is feeling better and is taking the prescribed medications, or as complicated as helping the patient obtain authorization for a wheelchair or assisting the family to coordinate care.

The health coach is the bridge between the patient and physician. For example, a patient may drop into the clinic unable to obtain a new prescription, needing help arranging home care and requesting a letter documenting her disability to apply for financial assistance. Previously, a clerk might take a message, send the patient home and pass the request to the physician, who would see the requests at the end of a long day. The health coach, in contrast, can solve most of these issues herself, requiring only a signature from the physician. Both the physician and patient are well-served by these health coaching functions.

In the FHC’s teamlet health coaching pilot, the health coach is introduced to each of the physician’s patients at the time of their visit. Patients are told about the role of the health coach, asked if they are comfortable having the health coach be present during the clinic visit, offered health coaching and encouraged to use the health coach’s voicemail if they need help with non-clinical problems (unable to get prescriptions filled, questions about specialty appointments and authorizations, issues with their self-management goals). The health coach has met 85% of the physician’s 226 active patients. Just over half—114—have accepted and engaged in health coaching. Four hundred forty-two encounters have taken place with the 114 patients, of which 251 were in-person and 191 telephone calls. While most between-visit phone calls last 5–10 min, some take over 20 min. Table 1 describes the nature of the health coaching provided.

Table 1 Types of Health Coaching Provided, August 2008–May 2010
Full size table

From the primary care physician’s perspective, each domain of health coaching is important, time-consuming and difficult to prioritize during a typical primary care visit. Given the need to attend to preventive health care, manage chronic illnesses and address acute complaints, there is not enough time during a primary care visit to complete the essential health coaching tasks. Nonetheless, in most primary care practices, no other team member works closely with the clinician and patient to ensure understanding of what transpired in the visit, facilitate patient follow-through with behavior change and medication management, and provide access and continuity between visits to patients who need more frequent contact than is possible in busy primary care practices. The health coach assumes these important functions, relieves the primary care clinician of the burden of attending to them, and thereby allows the clinician to attend to other patient concerns that are often inadequately addressed during a rushed visit. Working as a stable teamlet allows for continuity and trust to be built between clinician and coach, and allows patients to transfer their trust in the clinician to trust in the teamlet.

HEALTH COACHING VIGNETTES

We tested the teamlet health coaching model with an experienced physician and a relatively stable panel of patients, most of whom had been seeing their doctor for 10 years or more. We wanted to address some clinicians’ resistance to sharing their practice with another team member out of fear of disrupting the therapeutic alliance. Since this was a practice improvement project rather than a research study, we did not measure patient outcomes. Yet we observed that patients are receptive to working with another member of the health care team, especially if they are language-concordant with the health coach, have complicated medical or psychosocial issues, or are challenged by the need to self-manage their illnesses. A few patient portraits illustrate the myriad ways in which the health coach worked with the physician’s patients:

Ms. L, a 76-year-old Chinese woman with diabetes and hypertension, lives in the basement of her son’s house with her husband. Mr. Q, her husband, is an 80-year-old frail Chinese man suffering from Parkinson’s disease. Ms. L would like to see her doctor monthly, but the doctor does not have sufficient appointment slots. The coaching relationship, involving many phone calls, alleviated many of Ms. L’s fears, reducing her desire for physician visits. The time spent by the coach educating Ms. L about her and her husband’s medications allowed her physician to focus her visits with Ms. L on reviewing test results, medication changes and acute problems. For Mr. Q., who has a hard time coming to clinic, the health coach conducted a home visit, which helped the physician understand the family’s resources and challenges.

Ms. A is a 40-year-old Bangladeshi woman with uncontrolled diabetes and depression. She did not like to talk on the phone because discussing her diabetes aggravated her stress. The health coach had to figure out a way to help Ms. A handle her stress. Rather than working on much-needed behavior change using action plans, the coach decided to build trust by checking in with the patient once every few weeks. Ms. A grew more comfortable with her health coach and began to talk about her diabetes, her fears and her problems taking her medications. The health coach also worked with Ms. A’s overweight 10-year-old daughter on healthier eating choices.

Mr. T is a 64-year-old Vietnamese man who believed that his blood pressure was high only when he saw the doctor. The health coach worked to overcome Mr. T’s denial of his hypertension, tracking his blood pressures with the aid of a home blood pressure monitor. Gaining an understanding of his hypertension, Mr. T started taking medications as prescribed, reduced the salt in his diet and brought his blood pressure under control.

Ms. G is a 55-year-old African-American woman who worked with the health coach to address her excessive soda intake. She lost some weight and felt comfortable to begin exercising. A foot injury and her son’s incarceration derailed her commitment to improve her lifestyle, and she began regaining weight. Through regular phone calls with the health coach, Ms. G recommitted to starting regular exercise.

These vignettes show the breadth of the interactions the health coach has had with the teamlet’s patients, as well as the deep connections the coach has made with some of the physician’s most complicated patients. Most of the actively coached patients had frustrated their physician by demands for frequent visits, resistance to making lifestyle changes and worsening medical conditions. In many primary care practices, physicians struggle with these issues alone, often failing due to lack of time.

DISCUSSION

Implementing teamlets is not easy. The Family Health Center has three health coaches at 50–80% time who work with different clinicians—residents, faculty and nurse practitioners—who refer patients with poorly controlled diabetes, hypertension or lipids to the coach. In this way, a significant proportion of patients with poorly controlled illness have access to non-teamlet health coaching. Workflow is a challenge; for example, the physician cannot wait while the coach spends 10 min in a pre-visit setting the agenda and reviewing medications. Moreover, if the coach is busy with a 20-min post-visit discussing behavior change, the coach is not available for the next patient for whom the physician needs the coach. In addition, the timing of health coaching sessions must coincide with availability of exam rooms. The health coach needs to understand the physician’s pace so that coaching encounters do not overly encroach on limited space. Constant huddling is needed to anticipate and manage these workflow issues.

Without payment reform, the business case for health coaching is elusive but not impossible. In a fee-for-service system, physicians paid an average of $60 per visit could pay for a health coach at $40,000 salary by seeing three extra patients per day; this would be justified only if the coach saves considerable time for the physician. In a capitated or salaried system, physicians could care for somewhat larger patient panels and thereby increase income, if they worked in a teamlet model rather than alone. In such a system, health coaching skills can be successfully mastered by non-professional team members who are considerably less costly than mid-level practitioners, nurses or pharmacists. If coaching were demonstrated in a controlled trial to reduce unnecessary hospital admissions and emergency room visits, payers (Medicare, Medicaid and private insurers) might reimburse health coaches. Some organizations have succeeded in implementing health coaching without increasing personnel, making coaching financially sustainable.

Years ago, primary care practices consisted solely of a doctor and a nurse, often working together for many years. They trusted each other, they worked out a division of labor, and the patients trusted them both. This was the original teamlet. This traditional model survives today in small primary care offices, though in many cases the nurse has been replaced by a medical assistant.

Many primary care practices are more complex, with several part-time clinicians, medical assistants and nurses who do not always work with the same clinician due to variable schedules and competing demands. Teaching practices are particularly complicated with medical assistants and nurses working with different residents and faculty every day. Stable teamlets are the exception: medical assistants are seldom linked with a specific patient panel and are given tasks—perform this EKG, check that blood sugar and re-do the blood pressure—rather than sharing the responsibility and privilege of helping to make their panel of patients as healthy as possible.

Can medical assistants—who take vital signs, assist with procedures and manage patient flow—additionally perform health coaching? For practices with many patients requiring assistance to manage their chronic conditions, medical assistants would not have time for health coaching, requiring practices to create a new job category of health coach. For practices with fewer patients needing coaching, the medical assistant and coaching functions could be merged in one person.

Can the FHC’s teamlet coaching experience be generalized to other settings? As a safety net clinic, the FHC—with patients speaking many languages—is ideal for health coaching since the coaches are language concordant with their patients, thus serving as a cultural bridge. Given the increasing diversity of the US population, these issues affect many primary care practices. Yet even in smaller practices with middle-class English-speaking patients, the problems of physicians not having adequate time for behavior-change counseling or proper education and adherence counseling for medications are widespread. The fact that 50% of patients with hypertension4 and over 60% of those with diabetes are poorly controlled5 attests to the need for functions provided by health coaches. As a teaching clinic, the FHC has about 70 clinicians, mostly residents and faculty who see patients only one or two days per week. This personnel structure impedes teamlet formation, which is more easily implemented in practices with full-time clinicians, allowing a health coach and clinician to build a relationship and truly share a panel of patients.

CONCLUSION

For 4 years, we have attempted to resurrect the traditional teamlet concept at the Family Health Center. With complex schedules and challenging logistics, it has been difficult. While we did not gather specific measures to evaluate the pilot teamlet described here, we feel that the patients and physician benefited from the experience. We suggest that primary care sites—whether community health centers, private offices or teaching clinics—consider incorporating the teamlet concept into their care model.