How do you devise a capitation plan that works? With understanding, hard work and gumption, says the chief financial officer of Sharp Rees-Stealy Medical Group in San Diego.
Sharp Rees-Stealy Medical Group,
San Diego, Calif.
Developing a successful compensation system for physicians in a medical group means creating a method that can evolve with the group practice. Of the different schemes around today, Sharp Rees-Stealy Medical Group in San Diego decided to move from fee-for-service productivity-based compensation to capitation. It wasn't an easy transition.
The group has 285 physicians. After the first two years, each receives one share in the medical group. We have 10 locations around San Diego, and about 140,000 covered lives, of which roughly 20,000 are elderly.
When work began on the plan, there were two things that needed examining: how to influence behavior by using incentives, and how to monitor that behavior. With capitation, money is made if you deny care or just do less — a perverse incentive that was the opposite of what we were used to. So right away we needed to design the correct compensation incentives.
The second half of any capitation compensation plan is monitoring physician behavior and practice. Sharp Rees-Stealy developed a Physician Assessment Report (PAR) to monitor four key areas: quality of care, quality of service, cost-effectiveness and organizational participation. The original plan was to have the PAR account for 20 percent of compensation. It was determined that the tools were not in place to accurately monitor cost-effectiveness, so the percentage was dropped to 14 percent. The physicians have been monitored for the past three years and receive quarterly reports.
Capitation may bring on an abrupt change in physician behavior. Six months after the capitation plan was implemented at Sharp Rees-Stealy, the average number of patient visits per month per physician had dropped from 312 to 284. For some, the attitude was, "I'm not doing all this. I'm going home at five o'clock and I'm not seeing an extra patient." It took a while, but that attitude turned around and the number of patients seen began increasing. Now we're at about 340 visits per month. The physicians could see that if patients started leaving the practice, the whole group would be in trouble.
Know your roots
Before you start building a new compensation plan, take a hard look at your group and thoroughly understand your current formula. Sharp Rees-Stealy was founded in 1923 by two physicians. Right away we had conflict: One physician was a specialist; the other handled primary care. And that conflict remains today with managed care.
In 1985, there were 40 physicians practicing in a multispecialty group. It merged with Sharp Health Care System, which consisted of two hospitals and a nursing home. Within a year, we signed 20,000 Medicare risk patients. By 1993 we had nearly 130,000 Medicare and commercial HMO patients. The compensation model was trying to catch up with what was going on.
In January 1993, 80 percent of our patients came through 13 HMOs (one Medicare, 12 commercial). Eighty-three percent of our revenue was from capitation. We had very little fee-for-service business. So we knew that we had to make a change, because we could not continue to pay the physicians on productivity when in reality we were heavily into managed care.
In March 1993, the physicians attended a retreat that involved some real soul-searching about compensation. By June we had started working on what we called division pools, from which salaries would be allocated. We came up with two pools: Pool A for the fee-for-service business — a major occupational medicine program is our main source of fee-for-service income — and pool B, containing all our HMO capitation money.
In January 1994, we came up with a plan to use patient panel size, adjusted for age and sex, to determine how primary care would be paid. We calculated everyone's panel sizes at that time. The medical director and the board participated in the process as we determined individual capitation by division. We looked at the average income in our area, in our state and by specialty. We knew what our physicians were making, but we needed to be sure that all divisions were paid competitively.
We added up all the individual incomes in a division and used that as the amount needed to pay the division. Then we did that for all the divisions. Not surprisingly, the totals were greater than our income. We made some adjustments to come in line with our income.
We monitor new physicians, but their compensation is fixed and guaranteed until their panel size is sufficient to pay them more than the guarantee.
What about part-timers? We've had physicians come in full time and then say they want to work perhaps three days a week. The board is considering two levels of employment, either full-time or half-time. Half-timers would share a panel with another half-time physician.
Who gets how much?
We didn't want 100 percent of their compensation to be all out of the capitated pool because we wanted to give them some incentive for pool A, the fee-for-service patients. A problem you get with managed care is that there is little incentive to see fee-for-service patients. Their workload and compensation all come from their HMO patients. We decided to pay out fee-for-service income as a percentage of the dollars that they bill. Astoundingly, the best we have been able to do is 21 cents on the dollar. The incentive is just not there, and that continues to be a problem.
As we looked at compensation, probably the most important thing we did was to recognize up front that, especially in Southern California where managed care has penetrated every aspect of health care delivery, physicians were losing autonomy. We even have a call center that was doing patient scheduling. There was such a feeling of loss of control by the physicians that we wanted to give them control over how they distribute the money within their divisions.
We eventually decided to determine the amount of the salary pools, but not the allocation formula. While the plans had to be approved by the board, how the money was to be divided was up to the division. We stipulated that the ranges had to be fair or they wouldn't be able to hire anyone into the division. We have 29 specialties that probably have 29 different ways of splitting up the money. Three primary care groups split their pools based on panel size and some form of productivity.
For example, a group of five cardiologists divides the total by five and there is no argument about who has the most cases, or who spends the most time in the hospital. They don't care.
The five members of one division last year were paid around $1.5 million. Their capitation pool may have been only a million dollars. Any increase in salary comes from pool A; that's the incentive for them to boost their fee-for-service and hospital risk pools. The base stays the same year after year unless we have a decline in membership.
There is no right or wrong way to allocate compensation, so long as the participants are comfortable, and there are sufficient incentives for individuals and the group as a whole to benefit.
Where the money comes from
Sharp Rees-Stealy Medical Group runs on a foundation model, where the management services organization belongs to the Sharp Healthcare System. The physicians are a personal services corporation. The money all comes into the group and the challenge is how to divide it.
We do allow pass-through. A lot of physicians get involved at hospitals: A cardiologist will read EKGs for the hospitals, some neurologists will read EEGs and so on. We allow the money they earn to pass straight through to them at 100 percent.
Another income source flows from full-risk (as opposed to professional capitation) contracts with HMOs. There is a hospital risk pool. If our bed days stay down and everything else is in line, there is a significant reward. But the hospital risk pools are really an unknown. They have varied from $1 million to $5 million. Planning compensation using risk pools is dangerous.
The hospitals are all under the umbrella of Sharp Healthcare. The risk pools are calculated on an extremely complex point system, but basically the different levels of care at the hospital are worth different levels of points. An ICU day is 10 points, an acute care day is 8 and an emergency department visit is 6. The points are added up, and based on the hospital's historical cost structure, a cost per point is assigned. That is what is put into the risk pools. The system works but doesn't lend itself to monthly tracking to see how the physicians' group is doing. There can be a lot of surprises at the end of the year. The medical group can know during the year what the bed days are, and know if they went up or down. If too many of those were ICU days, the return from the pool will be less than anticipated.
The physicians all get panel reports, so they know the number of active patients, urgent care visits and referrals to our own specialists. Under capitation, the incentive is to have a large panel and send patients to urgent care or specialists or not see them at all. That has to be watched because if patients aren't being seen, they start complaining, or worse, leave the group. We developed the PAR — really a physician report card — and we publish it quarterly. The doctors know how they are performing compared to the others in their division and to the entire group.
The challenges don't stop once you've devised the capitation plan, put it in place and earned compliance from the physicians. Inevitably you have revenue shortfalls in the pools. It happens like this: You create the pools, and then a group needs another physician. You say, you can hire another physician but you have to do it with the same pool of dollars, which is impossible.
Then there is the issue of fee-for-service patients. They are definitely less attractive to physicians, but have the highest compensation rates around. They pay 80 percent, maybe 90 percent of charges. It's wonderful. So we pay the physicians as a percentage of their dollars and they get 30 cents on the dollar to see the occupational medicine patients. It has been a struggle to find and offer enough incentive for the physicians to agree to see the fee-for-service patients. But we can't just give up and say we're only managed care, because there are fee-for-service and PPO patients out there that we can serve.
Pay equity among divisions and departments is another tough nut. Family practice believes the ranking should be family practice, internal medicine and pediatrics; internal medicine believes it should be internal medicine, family practice and pediatrics, and pediatricians think they should be on top. We use our surveys to help us on this issue, but we still go round and round on where everybody fits in.
Too much emphasis on panel size can come back to haunt you as it did us when the physicians got huge panels and then couldn't meet the patients' needs. San Diego is an extremely competitive market. For a breakfast and a visor, senior citizens will switch plans. Employers will drop you and move to somebody else for a dollar a month. In this environment, when the physicians aren't seeing patients and the patients get disgruntled, it is dangerous. In 1996 our dropout rate was about 20 percent. We offset that to achieve a 2-percent net growth in enrollment, but if you can't put a finger in the dike and keep them from leaving, you will never get ahead.