“Growing interest in ways to link a pathologist’s productivity to his/her compensation makes this a widely-discussed topic within many pathology group practices.” –Dennis Padget
CEO SUMMARY: Part Three continues THE DARK REPORT’S series on measuring pathologist productivity. In this installment, pathology practice consultant Dennis Padget identifies different approaches to appropriately link pathologist productivity with compensation. After four decades of service to the pathology profession, Padget, of Simpsonville, Kentucky-based DLPadget Enterprises, Inc., recently retired. This interview was conducted by THE DARK REPORT’S Editor-In-Chief, Robert L. Michel.
PART THREE OF A SERIES
EDITOR: Let’s resume our earlier discussion of workload-based performance compensation systems for pathologists. (See TDR, October 11, 2004.) Before plunging into this topic, would you review for our readers the most prevalent compensation models in use today, regardless of their basis. That would give us context about the role that performance-based systems can play. What is the most prevalent compensation model and which is the least prevalent?
PADGET: Thanks for inviting me back. I must preface my remarks by stating that I’m not a physician compensation and benefits consultant. These were not areas in which I focused my attention when working with pathology clients in the past. So the sense of proportion I’ll give you is from my gut, not from reliable study. I could be way off on the numbers, but my description of the various models will be on target.
PADGET: My sense is that, by far, the most prevalent compensation model used by pathology groups today is highly democratic: The group simply divides total income equally among all its shareholders or partners. During the year, all pathologists get the same base salary. Whatever undistributed income remains at year’s end is divided equally as a bonus.
EDITOR: What compensation model would come next in frequency of use?
PADGET: Likely the next most common model starts with a staggered base salary. Essentially, the most senior members of the group have a materially higher base salary than the junior members. Any undistributed group income left over at year’s end is given out as a bonus. The bonus might be distributed equally among all the shareholders or partners. Alternatively, it might be distributed based on some weighting factor, like base salary.
EDITOR: What compensation model is least often seen in pathology?
PADGET: Workload performance, sometimes referred to as the “eat-what-you-kill” approach. In my experience, this is the least-used compensation model. Some independent laboratories operate from this approach, but it is seldom used by hospital- based pathology group practices.
EDITOR: Can you speculate as to why the fundamental models you’ve described might be encountered in the proportion you cite? For example, why might the equal distribution model be the most prevalent?
PADGET: There seems to be a strong correlation between practice size and the compensation model used by a particular group. The equal distribution method fits well with small-to medium-size hospital-based practices. In this environment, all the pathologists tend to do a little of everything. That makes it relatively easy to divvy up the work equally. Patient cases, marketing, practice administration, lab oversight, and similar tasks are evenly shared among all pathologists in the group. So it makes sense that total practice income is shared equally as well.
EDITOR: I recall reading somewhere that small- to medium-sized groups dominate the pathology profession. If they are the ones most likely to use the equal distribution model, it stands to reason it would dominate too.
PADGET: Right. According to a survey by the College of American Pathologists (CAP) in 2002, 57% of respondents practiced in a group of one to six members. Further, 74% belonged to a group of one to ten members.
EDITOR: What type of setting encourages the compensation model based upon varying salaries for senior and junior pathologists?
PADGET: The “salary range” model is often used by large hospital-based practices. When you think about the usual characteristics of such a practice, this correlation makes sense as well. There is typically a wide range in the age and years of practice of the members. Many years separate the most senior from the most junior member. Skill and degree of sub specialization vary significantly. There is often a pronounced difference between pathologists in preference for work versus leisure. These, and other factors, can readily be accommodated by a salary range compensation system, to the point that most everyone in the group agrees the outcome is fair and reasonable.
EDITOR: That’s interesting. What are your thoughts about the pure workload compensation model?
PADGET: In my opinion, the pure work- load performance compensation model is ideally suited for the independent lab environment. In this setting, pathologists are focused on reading slides and churning out reports, almost to the exclusion of everything else. They don’t get many frozen sections, and “curbside” consults with surgeons are rare. They are seldom saddled with administrative duties, and because of specialization, they are less likely to be looking at tissue one minute, a Pap smear the next, and a peripheral blood smear the third. Compared to the inpatient hospital setting, there’s not a huge variation in case complexity.
EDITOR: So the work is focused and fairly homogenous, which makes it like an assembly line. Those are factors commonly felt to be necessary to make a workload performance compensation system function well in the manufacturing and other business sectors. Is that what makes it apropos for use by independent laboratories?
PADGET: Yes, you’ve got it just right! Pathologists want to get their work done early in the day, and then leave. Moreover, the lab corporation has a legitimate business need to incite high productivity. Thus, paying pathologists on a piecework basis in this type of laboratory setting makes eminent sense.
EDITOR: That brings us to an interesting contradiction. The majority of pathology groups are not using performance- based compensation systems to pay their physicians. Yet there’s plenty of talk in the profession about the benefits of linking productivity to income distribution. How do you explain this apparent contradiction?
PADGET: Workload performance compensation is a sexy concept! But its promise is likely overrated. This compensation approach is not right for every pathology group. I’ll explain, first from the perspective of the academic pathology environment, and then from the private practice environment.
PADGET: If I were the chair of an academic medical center pathology group, I’d be very skeptical of trying to use productivity directly in my faculty compensation formula. The system would have to be very, very complex, because the faculty duties are so diverse.
EDITOR: That’s true.
PADGET: For example, what common denominator is there that equates the work of a transfusion medicine physician with that of a neuropathologist, or the head of the autopsy service, or the physician director of microbiology? Even if you could devise such a linkage, would the incentive encourage the response you want, or something quite different? Look at it this way: You install a system that pays your surgical pathologists on a piecework basis to stimulate productivity; but what then happens to resident education? Will the quality and quantity of training diminish as attentions become focused on the compensation “carrot” you’ve put out there?
EDITOR: Academic pathology departments also have a different mission and budget process. That influences the options for crafting a feasible package of total pathologist compensation.
PADGET: True. There are faculty incentive compensation programs that work well in an academic environment, but I don’t think pure workload performance is one of them. A program that uses negotiated individual goals is an example. The chair sets aside a pool of dollars to fund faculty bonuses for the coming year, and then meets with each member of the faculty to mutually agree on a set of individual goals. Each member’s bonus is linked to those goals and is paid in proportion to attainment.
EDITOR: But the goals have to be objective and measurable. And some could be workload-based. Correct?
Every pathology group must carefully weigh the pros and cons before plunging into an arrangement where compensation is heavily linked to each pathologist’s personal productivity.
PADGET: Yes. Workload-oriented items might well be included, but those goals will be tailored to each individual’s responsibilities. It helps the chair better control for unintended consequences this way. This type of compensation approach provides many of the advantages of an incentive system, but without all the risks of the pure productivity model.
EDITOR: Good. Let’s turn to the private practice environment for a moment. The big difference here is not having to worry about resident education. So private practitioners have more to gain and fewer risks with pay-for-performance, right?
PADGET: The critical question for private practice groups to ask is “what is gained if each pathologist’s compensation is based on his/her personal productivity?” Like everything else, there’s a cost to developing, installing and maintaining the incentive system. There is also the risk it won’t work as expected—that it will have a negative outcome, not a positive one. Every pathology group must carefully weigh the pros and cons before plunging into an arrangement where compensation is heavily linked to each pathologist’s personal productivity.
EDITOR: That’s sound advice. People don’t always remember there is a cost to craft a compensation system that functions smoothly. But you are also saying there’s nothing inherently superior about a pay-for-performance arrangement, compared to something like the equal income distribution model.
PADGET: Think about a small-to medium-sized group practicing at a com- munity hospital. It’s close-knit. Everybody likes and respects everybody else. Work is evenly shared. Each pathologist pulls his or her own weight. Income is equally divided and nobody’s unhappy. What does such a group have to gain by converting to a performance-based compensation system? If the truthful answer is “not a lot,” then the risk of backfire clearly outweighs any change benefit. You will be better off to stay with the existing compensation arrangement.
EDITOR: You definitely have concerns about linking money to pathologist productivity.
PADGET: It may sound like I’m “down” on performance-based compensation systems, but I’m really not. They definitely have their place. In certain situations, they have a lot to offer some groups. I simply like to emphasize the need to carefully weigh the costs, benefits, risks, and rewards before any group heads in that direction. These systems are not a surefire panacea to every problem- real or imagined. And I’ve seen them literally tear a group apart. It doesn’t happen often, but that risk is what makes me respect their destructive powers as much as their beneficial properties.
EDITOR: Well said. But with declining payment rates for professional services, pathologist productivity is going to stay a high profile issue for the profession. For example, younger “go getter” pathologists are mixed in with pathologists who “paid their dues” over many decades and now want to slow down. Or some groups are burdened with a pathologist who’s slacked off, but still draws a full share of income. Wouldn’t an “eat-what-you-kill” plan help in these situations?
PADGET: Maybe, or maybe not. And by the way, “young” no longer automatically equates to “go-getter.” Today’s young physicians often value leisure more than money. But back to the point. I must again emphasize that work-performance compensation isn’t a panacea. There are proven ways to address these types of situations without going that route.
EDITOR: Please explain.
PADGET: For example, remember that having equal shares in a corporation doesn’t mean each share holder has to be paid the same compensation. That’s handled in the bylaws and the employment contracts.
EDITOR: So unequal pay for “rainmakers” and workaholics can be accommodated by using a traditional staggered salary program.
PADGET: Correct. At the other end of the spectrum, it’s quite common to pay a lesser base salary to pathologist-shareholders while gradually reducing their workload in anticipation of near-term retirement. A group’s accountant can assist in the details. But the message is that a group isn’t forced into a work-performance compensation arrangement to cover the scenarios we’ve discussed. It might be a good solution, so long as the benefits look like they will outweigh the costs. But effective, reliable alternatives exist.
EDITOR: Can we come back to the problem of the under-producing pathologist? That situation seems to plague a growing number of pathology groups.
PADGET: The mention of “slackers” raises another very important point: in business as in medicine, always focus on the problem, not the symptom. The proper way to handle a pathologist who is not performing up to standard—whether in hours worked, accuracy of diagnoses, or similar—is through disciplinary action. It’s definitely overkill to change a fundamental component of your infrastructure—your physician compensation system in this case—just to minimize the damage that person is causing.
In certain situations, they [performance-based compensation systems] have a lot to offer some groups. I simply like to emphasize the need to carefully weigh the costs, benefits, risks, and rewards…Dennis Padget
EDITOR: However, most pathologists shy away from confrontation if the pathology group has a problem with one of its people.
PADGET: Maybe so, but it doesn’t change the essential fact that the group has an under-producing partner, and this, not the compensation system, is the problem. I’ve always recommended to my clients that they address the potential of this problem at the time a new physician joins the group. Get solid legal advice to craft the group’s shareholder agreement and physician employment contract so these potential issues are solved before they even happen.
EDITOR: All this background is very important, and it’s been useful spending the time talking about it. But let’s get down to some nitty-gritty. What development and implementation steps should a group take if it’s determined that a productivity- based compensation system is the right way to go? Are there three or four critical “do’s and don’ts” to more-or-less assure a successful program?
PADGET: Over my years in the profession, I’ve detected four keys to a successful work-based compensation system. I’ll describe them in the context of the most common practice setting—the hospital. That also happens to be the set- ting where the biggest obstacles are encountered.
EDITOR: What’s the first key?
PADGET: First, all stakeholders in the group must buy into the idea that the system will compensate fairly for work performed, but it won’t necessarily distribute income in proportion to that generated by each individual physician. This is an extremely important principle that recognizes the unity and totality of the practice. It says you’re not going to get 30% of the practice income just because you personally generate 30% of the revenue; some of your associates have to fulfill vital practice functions that don’t generate income, or that don’t generate income in proportion to the work effort.
EDITOR: All stakeholders deserve to be fairly compensated for the tasks they are asked to perform by the practice, notwithstanding the vagaries of the market and the policies of third-party payers for valuing those tasks.
PADGET: That’s right. Due to CPT coding rules and insurer payment policies, major surgical resection cases typically don’t yield as much income on a time-adjusted basis—like income divided by minutes—as do dermatology, GI, and urology biopsy cases. Autopsy cases may not generate any income for the group. And your hospital may not pay the practice anywhere near the value of the time that has to be spent on lab directorship and oversight duties. But the group can’t just say “we won’t do major surgical cases, autopsies, or Part A work anymore, because they’re not very profitable.” The hospital has contracted for the entire package of physician services—not just those you think are profitable—and you won’t have the contract for very long if you start cutting back. So the compensation system has to pay each physician for the work units generated, not for the income that he or she personally generates.
EDITOR: Got it. What’s the second key?
PADGET: Second, you have to pick a measurement unit that’s fairly homogenous and reasonably reflective of work effort across a broad range of specimens and disciplines. You won’t use case, because there’s tremendous variability in the number of minutes required for any one case versus another. You won’t use specimen, because a simple nevus will only require one or two slides, but a liver FNA or a prostate TUR may require 8-12 slides. You won’t use Medicare RVU, because they involve too much “averaging. ” For example, a simple nevus and a prostate TUR are both 88305 services. Another problem with RVUs is that they aren’t really all that accurate from one level to the next, like minutes for an 88307 specimen versus minutes for an 88309 specimen.
EDITOR: So what would be a reasonable common denominator?
PADGET: All things considered, you’ll likely select slide as the unit of measurement for your performance-based compensation system. This unit is sensible because: (1) it’s common to both histology and cytology, including Pap tests; (2) even peripheral blood and body fluid smear reviews involve slides; (3) it takes add-on work, like frozen sections, immediate studies, and special stains, into account by a direct means; (4) it correlates well with case and specimen complexity; and (5) it’s already captured or can easily be captured by most lab information systems. It’s also relatively simple to establish an equivalency scale for hands-on procedures that aren’t strictly slide-based. For example, you can assign a protein electrophoresis interpretation an equivalent value of one or two slides, an irregular antibody reaction study a value of four to eight slides, and so on. In this way you should be able to measure basically all patient care procedures in terms of slide count.
EDITOR: That’s reasonable. What’s your third key to success?
PADGET: Third, you will want to establish a “slide-creep” monitoring system. You can think of this as a quality control check, much as you’d do with an automated blood chemistry analyzer; that is, you’d periodically run a check to make sure the calibration is still within accept- able tolerances. New “hands-on” clinical test interpretations will be added from time-to-time; clinical protocols will change; changes in surgical technique will alter the fundamental nature of some of the specimens you receive; and so on. Your productivity measurement sys- tem—based on slides—needs to accommodate these changes.
EDITOR: Might some people try to “game” the system? How do you check for that?
PADGET: Another major function of the “slide-creep” monitoring system is to check for unjustified patterns and trends. We’d all like to believe that nobody in our group would ever try to “game” the system, but it’s better to periodically test for this than to allow an issue to fester to the point it becomes a major problem. The much more likely scenario you’ve got to look for and control is where one or two docs are overly cautious in their handling of certain specimens or are not following an accepted clinical protocol for some reason. Straightforward examples would be where one pathologist is unsure of himself when it comes to lymph nodes, so he orders six H&E slides per node rather than the standard two; or where group policy says you order an Alcian blue stain on esophageal biopsies only when Barrett’s syndrome is suspected by the referring physician, but one pathologist routinely orders that stain “just to be sure.”
EDITOR: So the objective of the monitoring system is to make certain you are not rewarding somebody for inefficient work methods or standards, however they may have come about. Is that right?
PADGET: Yes, very good! Depending on the sophistication of your lab information system, you may be able to perform this check by looking for outliers in a printout of slide count by specimen type by physician. In the worst case, you can always select a random sample of each physician’s cases and look through them to see if anything unusual sticks out.
EDITOR: We’re at your last key to success. What is it?
PADGET: The fourth key to a successful work-performance compensation system is also the most difficult. It focuses on all the duties and activities that, while absolutely necessary, don’t directly generate billed revenue. The two biggest categories of such work are hospital lab oversight, commonly called hospital “Part A” duties, and practice administration. This consists of everything from meetings with attorneys and accountants to negotiation of hospital and managed care contracts to “beating the bushes” for more business.
EDITOR: I often hear this area is very hard to address from a productivity measurement perspective. Have you come up with a solution?
PADGET: No. I must admit upfront that I don’t have a surefire way to work these duties and activities into a pay-for-performance compensation system. Nor have I heard or seen where anyone else has either. Common sense and group member consensus-building are very important and you likely won’t find two arrangements quite the same.
EDITOR: So what are some of the alternatives for handling Part A and related duties in a performance compensation system?
PADGET: Take one end of the spectrum— a relatively small, close-knit group. In this case, it might be reasonable and fair to simply ignore non-revenue generating activities altogether. This would be okay for a practice that’s run in a highly democratic manner, in which all the work quite literally is shared equally among all the members. In essence, this treats non-revenue activities as “overhead” and says that each slide that’s captured for income distribution bears the same amount of “overhead” as every other slide, regardless of which physician handles the slide or where it came from.
EDITOR: What’s another method?
PADGET: Another fairly simple scenario is where one physician leader handles all practice administration duties plus the high-level lab medical director functions. All other non-revenue activities are shared equally among the other members of the group. Maybe the physician leader spends 70% of his or her time on functions other than direct patient care. What you can do here is reach consensus on how much 70% of the physician leader’s time is worth. He or she is then paid that amount “off the top.” All other non-revenue activities are treated as “overhead” as in the first scenario.
EDITOR: Is there another possibility?
PADGET: Yet another approach is to reach consensus on an hourly rate for non-revenue generating work. Then any given physician’s compensation for a month is the sum of slides times sliderate plus hours times hourly rate.
EDITOR: But this area is open to group discussion, debate, and compromise. What works for one group might not work for another group. Right?
PADGET: Correct. There is no single way, no “right way”, to handle this. I think the only “wrong way” to handle this area is to ignore it in a situation where the work clearly falls most heavily on one or two pathologists. It doesn’t take long before the inequity becomes a major political issue within the group and that’s not a good thing to have happen. Beyond that, my general advice is: keep any system or formula as simple as possible.
We’d all like to believe that nobody in our group would ever try to “game” the system, but it’s better to periodically test for this than to allow an issue to fester to the point where it becomes a major problem.—-Dennis Padget
EDITOR: Okay, could you pull it all together for us? How does the system work from month-to-month?
PADGET: Conceptually, the math is straightforward. You start with next year’s budgeted net income before share holder physician compensation. From that figure you subtract the compensation that will be paid to the doctors for their non-revenue generating duties. What’s left is budgeted shareholder-physician compensation payable on a work-performance basis. Divide that number by the budgeted number of slides for the coming year to get your compensation per slide rate.
EDITOR: I’m with you so far. What’s next? How do you figure how much to pay any one doctor for the month?
PADGET: Each month each shareholder-physician is compensated first for his or her non-revenue hours, or however else that set of duties is to be paid. Then you multiply the pathologist’s slide count for the month by the compensation per slide rate to determine his or her production-based pay. That’s about it.
EDITOR: Mr. Padget, thank you very much for sharing your ideas and suggestions with our readers.
PADGET: You’re most welcome!