Two years ago, our then-four-physician family practice decided to implement a new physician compensation plan for the upcoming fiscal year, only a few months away. We sent the plan to our accountant with instructions to implement it immediately. He read the plan and sent it back with a one-word memo: “HOW?” “How” became my project.

Developing the model

The compensation model we wanted to use was created by William R. Greenfield, MD, a physician in my practice [see “In Search of an Effective Physician Compensation Formula,” FPM, October 1998]. His goal was to develop a plan that would work in a group practice like ours, which has a mix of capitated, discounted fee-for-service, Medicare and private-pay patients, and that would be flexible enough to accommodate change and be applicable in various practice settings.

Dr. Greenfield established five specific criteria to guide his development of the model: It had to be consistent with the goals, philosophy and mission of the practice; economically fair to each physician; easy for the physicians to understand; easy for the practice to administer; and able to stimulate the physicians to participate in activities and behaviors beneficial to the group.

Because we had decided that the new compensation plan must be consistent with our mission statement and goals, we had to begin by writing a mission statement and identifying our goals. The process required evaluating and formalizing the procedures and philosophies that had been “assumed” during the 45-year history of the group. Other steps included creating the incentive pool, identifying the areas we wanted to incentivize and determining how each related to our mission and goals. It all looked easy on paper, but in reality these were difficult issues with which to contend.

To develop and write our new mission statement, we formed a committee representing each area of the practice (physicians, nurses, lab technicians, receptionists, record clerks and business clerks). Using the nominal group process, the committee composed the following statement:

The mission of our practice is to provide the highest quality care to all members of the family in a pleasant atmosphere by a helpful, caring and friendly staff. Our goal is to be the best health care provider in the county.

[For more information on the nominal group process and developing mission statements, see “Finding Diamonds in the Trenches With the Nominal Group Process,” FPM, May 1999, and “Effective Mission Statements,” Ask FPM, September 1998.] Once our mission statement was in place we were able to articulate our goals. Simply put, they are to provide the best patient care possible, while maintaining the fiscal success of the practice.

Establishing the incentive pool

Next we had to plan the transition from our old compensation system of equal shares and straight salary to our new model, which combined a fixed base salary with a quarterly incentive bonus. Dr. Greenfield suggested that we allocate the equivalent of 75 percent of the group's previous-year W-2 earnings for fixed salaries and place the remaining 25 percent in an incentive pool. He also proposed dividing the incentive pool into 10 areas, assigning each area a percentage of the whole based on its importance to the practice's mission and goals.

Dr. Greenfield's suggestion of reducing the previous year's salary by 25 percent seemed reasonable until we looked at the actual numbers. After a lengthy discussion, we agreed to reduce the base salary by 10 percent and eliminate each physician's $500 monthly expense account. This would provide us with a minimum annual incentive pool of $90,000, assuming no drop in our level of business.

Next we assessed the 10 areas of the incentive pool that Dr. Greenfield had identified. Our task was to decide how each one related to our mission and goals, whether each area should be included in our incentive pool and, if so, what percentage of the incentive pool to allocate to it (see “Areas of the incentive pool”). To accomplish this, our group discussed each of the following categories:

Seniority. Two physicians have left our small group in the past few years. In both cases, it seemed the physicians decided to leave rather than work out the problems they were having with our practice's administration. Recognizing that a stable staff of physicians adds continuity to the practice and helps us retain patients, we wanted to reward those physicians who stayed. As a result, we allocated 5 percent of the incentive pool to seniority.

Special qualifications. Although the four of us are board-certified family physicians, we don't all provide the same patient services. Yet we are well aware that our ability to offer certain patient services (e.g., obstetrics, nursing home visits, Federal Aviation Administration physicals and vasectomies) can increase cash flow and decrease referrals. We decided to allot 10 percent of the incentive pool to physicians providing services in the practice that would otherwise require referral.

Productivity. Simply put, “The patients must be seen.” Physicians who are available to see more patients contribute to patient satisfaction and benefit the practice. With this in mind, we decided to reward physicians with higher visit rates, which we defined as the number of patient encounters per physician per month, by allotting 10 percent of the incentive pool to productivity.

Panel size. When Dr. Greenfield developed this compensation model, he assumed that panel size reflected a physician's ability to attract and retain patients and originally suggested allocating 20 percent of the incentive pool to this area. Yet during implementation we discovered that some of our patients don't identify their designated primary care provider as “their doctor” and are willing to see any of the physicians in our group. Also, since it can take at least 18 months for new physicians to be included on a managed care plan's provider list, we realized that they will have small panels, regardless of how capable they are. Nevertheless, large panels are essential to the fiscal success of capitated practice, and we wanted to reward physicians who have them. Therefore, we assigned 5 percent of the incentive pool to panel size.

Utilization. During the past several years, our practice has added one part-time and one full-time employee to process referrals for our managed care patients. Confident that we could reduce overhead by making fewer referrals, we allotted 20 percent of the incentive pool to utilization. Twenty percent is admittedly a large slice of the pie, but we wanted to induce the physicians to make referrals when necessary for quality patient care rather than for the physicians' own convenience.

Compliance. Although we had often discussed the importance of complying with managed care companies' rules for coding and billing procedures, we still had room for improvement. In an effort to improve physician documentation skills so that the notes in patient charts support the codes submitted to insurers, we assigned 15 percent of the incentive pool to this area.

Patient satisfaction. During the six months prior to deciding to implement a new compensation plan, our practice experienced an increase in patient complaints, some directly related to physician behavior (e.g., tardiness, not returning phone calls or providing insufficient patient education). Since patient satisfaction is essential to the mission and goals of our practice, we assigned it 20 percent of the incentive pool.

Overhead. Our practice had an ongoing problem of physicians holding large numbers of charts in their offices for days, requiring the records clerks to spend significant time searching for them. In addition, some physicians had stopped or reduced the number of phone calls they made to patients, requiring us to hire an extra nurse. To reward physicians for helping to reduce our overhead and to encourage those who weren't to change their behavior, we assigned 15 percent of the incentive pool to this area.

Citizenship. Dr. Greenfield developed this area to reward physicians who took extra call, worked extra days and developed strong working relationships with staff and other physicians. With the exception of counting extra call and days worked, we found this to be an extremely subjective area. We decided that a group of four physicians was too small to avoid the emotional issues that could result from judging each other's commitment to citizenship and, consequently, we omitted it from our incentive pool.

Administration. This area rewards physicians who provide administrative services to the practice. Since we directly compensate the two physicians who act as our chief executive officer and chief financial officer, we omitted this category from our incentive pool.

Crunching the numbers

Once we'd identified the areas for the incentive pool, our final step toward implementing the new compensation plan was to calculate the percentage of each area that each physician would receive and then add these to determine the physician's total share of the incentive pool. This was not as difficult as we had assumed although it did require us to collect some data about the practice, which we did during a one-week period each month. We then used this data to determine averages for each quarter. Our goal was to keep the calculations simple, both for the staff, who were going to be keeping track of the data, and for the physicians, who were being evaluated.

Seniority (5 percent). Together, the four physicians in our group have spent a total of 76 years with the practice. To determine what percentage of the seniority pool each physician should receive we simply divided each physician's years with the practice by the total.

Dr. A – 30 yrs/76 yrs = 39 percent

Dr. B – 25 yrs/76 yrs = 33 percent

Dr. C – 19 yrs/76 yrs = 25 percent

Dr. D – 2 yrs/76 yrs = 3 percent

Special qualifications (10 percent). The physicians in our group provide 16 types of patient service in-house that would otherwise have to be referred. To determine how much of the incentive pool each physician should receive, we divided the number of special services each physician provides by the total.

Dr. A – 4 services/16 services = 25 percent

Dr. B – 5/16 = 31 percent

Dr. C – 6/16 = 38 percent

Dr. D – 1/16 = 6 percent

Productivity (10 percent). Using the patient sign-in sheets from the front office reception area, we determined that there were 5,091 office visits in the third quarter of 1999. Again, each physician receives the corresponding percentage from this area.

Dr. A – 1233/5091 = 24 percent

Dr. B – 1211/5091 = 24 percent

Dr. C – 1533/5091 = 30 percent

Dr. D – 1113/5091 = 22 percent

Panel size (5 percent). Using the capitation statements from our managed care companies, we found that during the third quarter of 1999, our practice had 4,708 capitated patients. To determine shares in the incentive pool, we simply divided each physician's panel size by the total number of assigned patients in the group.

Dr. A – 1167/4708 = 25 percent

Dr. B – 1235/4708 = 26 percent

Dr. C – 1292/4708 = 27 percent

Dr. D – 1014/4708 = 22 percent

Utilization (20 percent). Acquiring the data needed to determine utilization rates, quality, patient satisfaction and overhead required a little more legwork on our part. To reward physicians for limiting referrals to only those necessary for quality care, we first had to determine the average rate of referral for the group as well as for each physician. For one week per month per quarter, we tracked the number of referrals written (excluding emergency department visits) and the number of visits from patients whose insurance plans require written referrals for testing or consultation outside of our practice. In the sample week of the third quarter of 1999, 908 referrals were written, and the total number of patients who might have needed a referral was 1315, for an overall referral rate of 69 percent. Broken down by physician, the data look like this:

Dr. A — 164/222 = 74 percent

Dr. B — 290/342 = 85 percent

Dr. C — 212/336 = 63 percent

Dr. D — 242/415 = 58 percent

Once we had this initial data, we decided that physicians whose rates were below the average referral rate or no more than 5 percent above it would receive one point. Physicians whose referral rates were at least 6 percentage points below the average received two additional points, and physicians who were more than 10 percent below the average received another three points. A total of 10 points were awarded, and each physician received a share of the incentive pool equal to the percentage of points he earned.

Dr. A – 1 point = 10 percent

Dr. B – 0 points = 0 percent

Dr. C – 3 points = 30 percent

Dr. D – 6 points = 60 percent

Compliance (15 percent). We set an easily attainable goal for this category, rewarding physicians with compliance rates of 50 percent or greater. To measure compliance, our practice reviews 10 adult charts and five pediatric charts per physician per month. Adult charts are evaluated to determine whether patients are up-to-date on seven key services (e.g., Pap smears for women, PSA screening for men). Pediatric charts are evaluated for five key services (e.g., immunizations, use of growth charts).

We included only services scoring above 50 percent to determine how to split the pool. Since adult patients outnumber pediatric patients in our survey 2:1, we decided to give physicians two points for each adult service scoring above 50 percent and one point for each pediatric service. Among us, we scored above 50 percent in 10 of the 12 services measured, for a total of 30 points. The incentive pool was divided as follows:

Dr. A – 8/30 = 27 percent

Dr. B – 4/30 = 13 percent

Dr. C – 17/30 = 57 percent

Dr. D – 1/30 = 3 percent

Patient satisfaction (20 percent). We measured patient satisfaction by asking patients to complete a survey evaluating the physicians in nine areas. We found we were able to acquire sufficient data by having 45 patients per physician complete the survey each quarter. To determine compensation, we simply counted the number of areas where 80 percent or more of the respondents ranked the physician as “good” or “excellent” and divided it by the group's total. For example, in the third quarter of 1999, the group received a “good” or “excellent” in 27 out of a possible 36 areas (i.e., 9 areas x 4 physicians = 36). The incentive pool was divided as follows:

Dr. A – 5 areas/27 areas = 19 percent

Dr. B – 6/27 = 22 percent

Dr. C – 9/27 = 33 percent

Dr. D – 7/27 = 26 percent

Overhead (15 percent). Because we wanted to motivate physicians to return patient phone calls and patient charts to the files, we divided the money allotted for overhead to address both of these issues. As I mentioned previously, although each physician had his own nurse, several were relying heavily on the new nurse to return patient phone calls. We encouraged physicians to change their behavior by compensating them if the nurse made only 25 percent or fewer phone calls per physician per quarter. We asked the nurse to track phone calls for one week each month. She determined that she'd made 1,074 phone calls during the third quarter of 1999. Two of the four physicians ended up sharing equal parts of the “phone call” pool:

Dr. A – 0/1074 = 0 percent

Dr. B – 0/1074 = 0 percent

Dr. C – 504/1074 = 47 percent

Dr. D – 570/1074 = 53 percent

To reward physicians who returned charts to the files, we used a point system similar to that used for utilization. First, we tracked the number of charts each physician and his nurse had out at the start of the day for one week, determining a mean of 15 charts per physician. Physicians who had up to 20 percent more or less than the average number of charts out received one point. Physicians from 21 to 50 percent below the average received two additional points, and physicians with less than 50 percent of the average earned another three points. Each physician received a share of the incentive pool equivalent to the percentage of points he earned.

With our mission statement written, our goals clearly defined and our incentive pool established, implementing our compensation plan became simply a matter of doing the math each quarter to determine the incentive bonus. For the third quarter of 1999, the incentive pool was $20,000, slightly less than the expected $22,500. “Dividing the incentive pool” shows how that $20,000 was split among the physicians.

Lessons learned

Although everyone in our group had initially agreed that we needed a new compensation model, support lagged when we implemented the plan last year. Several physicians balked once they saw concrete numbers and began to realize how the plan would affect their income. We were caught unprepared. After an emotionally charged discussion, we decided to implement only parts of the plan. Some physician behavior changed immediately. The number of charts left in physicians' offices dropped to zero almost overnight. And instead of having the nurse call patients about lab results, several physicians began to fill out postcards to notify patients themselves.

We now recognize that dividing the incentive pool into eight areas may have been a bit overzealous on our part. In reality, when the incentive pool is divided into so many areas, no one area is weighted enough to encourage real behavior change. Consequently, we are currently revisiting this issue.

On the flip side, we were pleased to discover that the model requires us to use only minimal data. Moreover, the data was easy to collect and did not place an undue burden on our already busy staff. And, since the measures are tied to the goals and mission of the group, it is easy for the individual physicians to understand exactly how they are being evaluated. They can see how the conclusions were reached and, more importantly, how they can change the results by changing their behavior.

While we have not yet reached our goal of fully implementing this compensation plan, we are getting closer. And although we've added three new physicians and our payer mix now includes less capitation revenue than it did when we began this process, our formula is flexible enough to accommodate these changes too.