By Mark Smith
See also the probing Q&A posed by Richard
of the University of Massachusetts Medical Center to author Mark Smith.
We haven't reached the pointyet where physicians are being paid not to see patients, as farmers are paid by the government not to grow tomatoes.
It's clear, however, that payment systems for physicians must go beyond traditional production-based models for hospitals, HMOs, management services organizations, medical groups and other entities that employ physicians. What are the elements of a strategic physician incentive program in the age of managed medicine? A physician incentive program should include:
The intent of a physician incentive program is to remove the temptation for physicians to "overdoctor" while ensuring that they are nevertheless motivated to provide quality care within a framework of finite resources. There are a variety of strategic considerations within each of these elements. Old requisites such as fair reward for effort must coexist with new considerations such as skill in keeping a large panel of patients well and satisfied.
Physician productivity. Traditionally, physicians have been rewarded for aggressively generating revenue or for seeing a large volume of patients. This production-based model is counterproductive for physicians trying to compete in a capitated environment.
One strategic method for gauging and rewarding physician productivity is a point system based on billable patient services. One such system features adding up the total relative value units (RVUs) assigned by the Health Care Financing Administration to all billable patient services designated as "included services."
A peer review committee will need to rule on what are "included services," though the general principle is that services that are personally provided by the physician qualify, while those services provided by others are excluded. For example, office visits or procedures directly provided by the physician would be included, while laboratory tests, X-rays and therapy would not.
Under guidelines used by HCFA, a limited office visit (CPT code 99213) has been assigned a total relative value of 1.00 units. All other services have been assigned total relative value units below, equal to, or above 1.00. In this physician incentive program, one total RVU will be equal to one production unit for purposes of determining individual productivity (see "Example of a Physician Productivity Point System," page 47).
Physicians will be expected to generate a certain number of production units based on their seniority and their involvement with governance and quality assurance efforts. Productivity in this model does not consider payer source or payer amount, nor does it consider the level or type of ancillary services ordered, or the nature or amount of services provided to patients by other members of the group. Physicians therefore are encouraged to categorize patients based on their health care needs, rather than on the method (Medicare, insurance, HMO) of payment. An incentive to expend effort and work hard remains, while the motivation to order tests or generate specialist fees diminishes.
Compensation in this model can be adjusted upward for those physicians in the group who have a problematic case mix (frail, elderly patients, disabled persons, etc.) who may require more time during office visits with the physician. In addition, physicians may receive bonuses for exceeding the group's production norms or be penalized if their production is below the norm.
Outside income from existing contracts with nursing homes, hospitals, managed care plans and the like will be additional income to the physician beyond the standard group compensation formula. New contracts, and the compensation to physicians derived from them, must be approved by the board of the medical group.
Under the guidelines of this hypothetical physician incentive program, the production point system accounts for 40 percent to 60 percent of a physician's overall compensation.
Managed care efficiency. With the steady increase in capitated plans, appropriate use of resources by physicians is paramount. A common physician incentive program formula bases approximately 25 percent of total compensation on a physician's ability to provide quality care within budget. A utilization review committee must determine what constitutes reasonable per-patient costs for lab, X-ray and pharmacy, and must evaluate physicians based on their ability to provide care within these standards.
A group citizenship component. A physician incentive program is more than a formula for determining physician compensation. It represents a commitment to a new mindset, a new way of practicing medicine. Without the earnest participation of physicians, no such plan will work. A physician incentive program requires that physicians participate in medical group governance, utilization and peer review on an elected or rotating basis. Generally, physicians are expected to devote four to 10 hours a month on such activity. The goal is to educate physicians and instill in them a commitment to practicing quality care within the cost framework imposed by managed medicine. Those physicians who take a more active leadership role should be rewarded with additional compensation. Though formulas vary, physician incentive programs commonly base about 10 percent of overall compensation on group "citizenship."
While citizenship activity may be irksome to some physicians, it is preferable to having production point systems, utilization review and other factors imposed by administrators, managers or other "outsiders."
A patient satisfaction component. With the increased emphasis managed care organizations and employers place on "quality," patient satisfaction has become a leading issue for many physicians. Groups may either design their own patient satisfaction surveys to measure quality or turn to a wide range of consultants to do the job for them. These surveys are generally physician-specific and cover such areas of concern as waiting times, thoroughness of exams and courtesy. Most physician incentive programs base 10 percent of overall physician compensation on how well or poorly physicians score on patient satisfaction surveys. The goal is to focus physicians on the fact that patients have the option to choose other physicians and medical groups.
A group profitability component. An effective physician incentive program takes as its starting point the premise that physicians in a group must be committed to long-term collective profitability. Unlike the National Basketball Association and other professional sports leagues, which annually inflate players' salaries and pass on the cost to fans who seem perpetually willing to pay the freight, physicians must be compensated from a limited pool of funds. A physician incentive program therefore generally requires that an annual budget be established with a group profit-sharing arrangement. If the group produces an operating surplus in excess of projections, physicians will be collectively eligible for bonuses. If the budget is not met, physicians will receive reduced compensation. In either case, compensation is structured to favor the long-term security of the group, not the immediate aggrandizement of one or two of its physicians.
A physician incentive program with some or all of these elements creates a system of checks and balances that encourages cost-effective care without sacrificing quality or inhibiting physician effort (see "Five Elements in a Physician Incentive Program.") While such a system includes the kind of practice interference that physicians often resent, there are two consolations. First, a physician incentive program may be necessary for physicians' economic survival. Second, though such a program may put a physician's practice style "on trial," the physician is at least being judged by a jury of peers.
Mark Smith is vice president of Merritt, Hawkins & Associates, a national health care staffing firm based in Irving, Texas.
Even systems that are sound in theory can hit a snag in practice. That's why Managed Care asked Richard Lerner, M.D., to challenge author Mark Smith of Merritt, Hawkins and Associates with specific questions about how Smith's physician incentive programs would work. Lerner, who is in the Division of Primary Care in the Department of Medicine at the University of Massachusetts Medical Center in Worcester, found it a timely assignment. The division's group of 25 physicians, many with teaching and/or administrative duties taking up various proportions of their time, now faces some of the compensation issues Smith describes.
LERNER: Some doctors complain that their patients tend to be older or sicker than those of their colleagues. In your article you mention that that issue can be factored in. Do you have a specific numerical value, say, for an 80-year-old woman versus a 25-year-old man?
SMITH: Not in those terms. But the system of relative value units attempts to compensate physicians for the effort and additional time that needs to be spent with certain patients. The more geriatric the patient, or the more frail or ill the patient, the more time needed, and we suggest that it be coded accordingly. That itself takes care of part of this problem, to roughly the level of two standard deviations on each side of the norm. Beyond that, I've seen some groups adjust the RVU scale for specific physicians whose caseload, everyone agrees, is just more difficult.
LERNER: Some systems have tracked visit rates, that is, they count how many times an old person comes to see you versus a young person.
SMITH: Sure. In terms of the patient mix. But then I'll look at physicians who see 25 patients a day and others who see 45 patients a day with the same patient mixand with the same patient satisfaction. Sometimes it's not only an issue of effort, but of the results.
LERNER: Sure. Their efficiency.
SMITH: Yes, efficiency, skill level, and ability to manage a high volume of patients and make the patients still feel they're receiving quality care.
LERNER: By your system, if I'm seeing an older person all the time and Dr. X is seeing a younger person, presumably I'll be coding an "established patient extended," whereas maybe Dr. X is coding an "established patient limited." But there's probably some inaccuracy in that. One of the systems I've seen accounted for that problem. It weighted each age so that a younger man, I think, was weighted a 0.6 versus an older woman, who was close to a 2.0. But they didn't necessarily weight out the type of visit.
SMITH: What I've found is that you need to have a program that people can understand. When you stick with the basic RVU system, it's one that we already have in our heads to a certain extent. Billing codes are familiar, so we can discuss the system all having the same logical base, even though we may not agree on the results. But if I know I can see 35 patients a day because the average age of my patients is 40, but my friend Fred can only see 25 and his average age is 65, I don't want to see his patient mix. And I'm fine with him being paid the same amount as me, because I know he's working just as hard, as long as there's an annual review of case loads. In terms of what you mentioned about the accuracy of the billing code, the governor that has been established on this is utilization review.
LERNER: You have sort of an auditing system?
SMITH: Right. As you insert the UR process in this, much as you would in any other system, that's part of the checks and balances, intended to discourage physicians from taking advantage of the production side of the system, and to make them cognizant of providing cost-effective care.
LERNER: I've seen at least one system in which, as you got up in level of difficulty of the patient, the increment in the patient's value decreased. So when they went from a 99214 to a 99215 they didn't go from 1.52 to 2.34; their delta from the fourth to the fifth level was less, sort of saying, "It's not worth it to upbill." I'm not sure I philosophically agree with that. Your way of solving the problem of upbilling is to literally have someone going over some things.
SMITH: Right. In some parts of the country you've got at-risk Medicare HMOs, which are developing with great frequency. Without a compensation system that recognizes different levels of difficulty in patient visits, there's going to be a disincentive for physicians to see elderly patients, or at least to give them sufficient time.
You know, physicians I speak with get upset with managed care, not just because of the money, but because they feel someone else is judging the quality of care. So our system leaves that control in their hands, but says there's also going to be a watchdog. That watchdog function has become fairly comfortable to most physicians we encounteras long as it's performed by peers.
LERNER: It takes someone's time to do that.
SMITH: It does.
LERNER: What does an RVU point come out to in dollar value? I could see 24 "average return" patients, and then if I add on four extra, how much would that get in your system, compared to, say, working in a clinic, where I know I'm making $50 or $60 an hour?
SMITH: There are two ways to do it. The complicated way is to take the average of what you're reimbursed for your fee-for-service patients on each of these billing codes and divide these reimbursements by the assigned relative value units to develop a weighted average. For example, you might come up with $50 a point. Then multiply the result by the percentage left after overhead is deducted50 percent, let's sayand multiply by the percentage of the physician's income that reflects productionagain, 50 percent is a likely figure. That gives you $50 x .50 x .50 = $12.50 per relative value unit. The other way to look at this thing is to say, "OK, so production units make up 50 percent of my physician's base compensation of $120,000, so it should come to $60,000." So instead of breaking it down to the specific point, you kind of break it down to that whole unit of compensation.
LERNER: I see.
SMITH: So let's say the level you're looking for is 25 points a day. And as you look at this person's average over the month, over the quarter, over the year, and you think, "Based on this, he only reached 90 percent of his goal," Maybe you'll decide that your acceptable standard deviation is only plus or minus five percent. If the physician is between 95 and 105 percent, he gets his $60,000. If he's at 90, well, he only gets 90 percent of that number. It's a simpler calculation with probably the same net result. Of course, if the group itself isn't profitable, none of this matters.
LERNER: That was another question I had, because I know you mentioned group profitability as a factor. If it's not profitable, then how do you compensate those of us who worked harder versus those of us who worked less hard? Does everyone split the loss?
SMITH: That's a difficult one. Typically the organizations I work with have the fiscal backing to hit a bleep, to have a bad year, and to have it not necessarily affect the physiciansas long as the problem isn't too radically severe (which I have seen, and of course major amendments seem to be made very quickly in those cases). In that second year if you see that same trend, the only way I've seen to deal with it is through direct management of the problem physicians, the ones that aren't carrying the load. And I'm seeing a surprising number of physicians terminated for that these days.
Here's how certain basic services could be rated, and how combinations of those services might make up an appropriate daily total, in a physician incentive program typical of the kind suggested by author Mark Smith.
Total RVUs/ Production Units
|99213||Established patient, limited/intermediate||
1.00 (average visit)
|99212||Established patient, brief||
|99211||Established patient, minimal||
|99214||Established patient, extended||
|99215||Established patient, comprehensive||
|99201||New patient, brief||
|99202||New patient, brief||
|99203||New patient, limited/intermediate||
|99204||New patient, extended||
|99205||New patient, comprehensive||
|Daily office visit mix (potential examples)||
Production unit goal (set
|24 average return patients, or||
|12 average returns, 4 comprehensive new, or||
|8 average returns, 8 brief returns, 3 new comp., or||
|8 average returns, 6 brief returns, 4 extended returns, 2 new comprehensives||