Many in the medical community continue to have serious concerns about the concept of Pay-For-Performance while the major drivers of this movement - big business, for profit HMO’s, CMS and various health management consultants - seem to be steam-rolling ahead.

On the following pages appear thoughts, concerns and questions about P4P compiled by practicing physicians.  Please take the time to look them over and feel free to e-mail them to other physicians.

P4P is a term that is being applied for incentive programs that provide monetary bonuses to participating physicians who achieve or attain specific quality and/or efficiency indicators or standards including patient satisfaction that are established by the program. The size of incentive payments being proposed is typically from 1% to 5% of a physician’s revenue.
 
IS P4P AN AFFRONT TO THE MEDICAL PROFESSION?

At the outset, while a great many physicians understand the need for improved consistency in the delivery of quality medical care across the country, many feel it is an insult to dangle a carrot before them as an incentive for providing “better quality care.” To these physicians the very essence of a medical professional is to strive for the best care for their patients.  Doctors merely want to be paid fairly for the work they do, not be “bribed” with extra money to do what they should be doing in the first place.

Now let’s examine P4P proposals in a little more detail

Principles:

• Guidelines should be evidenced-based, developed and approved, wherever possible, by national specialty organizations, NQF or similar national forums.
• Any guidelines utilized for P4P should be consistent across all programs.

• Many managed care plans are developing guidelines established by their own scientific advisory boards (SABs). But these self-developed guidelines are suspect because these SABs are composed of consultants employed by the managed care organizations themselves.

• If different managed care plans have different guidelines, which are actually the right ones?

• If there are 10 different guidelines from 10 different managed care companies, how can practices cope with this? Won’t this be confusing for practices? Won’t it be confusing for patients? When the patients switch plans, will they be subjected to different standards?

• If we are calling the guidelines “best practices” or “standards to measure quality of care,” rooted in “evidence-based medicine,” shouldn’t there really be a single standard?

• There must be general agreement and total standardization among all health insurers as well as CMS because doctors cannot be held to what may be different quality standards.

• There needs to be total agreement between CMS and all insurers before P4P is instituted.

• There are many different types of performance measures and measuring organizations: HEDIS developed by the NCQA, ORYX developed by JCAHO, QIO which is CMS’ quality improvement organization, NQF measures from the Agency for Healthcare Research and Quality (AHRQ), RAND’s Quality Assessment Tools System, as well as hundreds more devised by different managed care plans. Which one(s) will be used?

• There are already many small-scale P4P programs that have been instituted or proposed across the country. The most consistent aspect of these various incentive programs is the inconsistency from one program to next which can make it difficult for physicians to know which criteria and sets of treatment guidelines apply to patients covered by different P4P programs.

Principles:

• The value and effectiveness of P4P programs should be documented in pilot programs before adopted nationally.
• The cost vs. benefit ratio for P4P should be studied.
• Proposed P4P Programs must take into account the required burden of data collection and address this issue before instituting the plan.
• P4P programs must address the following issues:
• Physicians with small numbers of patients in the program
• Patient controlled elements that physicians cannot influence
• Adverse patient selection- need to severity adjust
• Value, scientifically-based, of patient satisfaction perception on the quality of care provided.
• Prevention of reducing access to care to high risk patients.

• P4P only measures the tip of the iceberg of the totality of services and care that physicians provide, and doctors will be rated in managed care directories (with stars or check-marks) for only a handful of conditions. This is certain to be misleading to the general public who will most certainly perceive that it reflects on the “quality of general medical care that doctors deliver.”

• Since P4P guidelines must be based on validated medical evidence, what is the evidence for P4P itself? What is the evidence for the validity, efficacy, usefulness, practicality, benefit, cost-effectiveness, value, etc. of P4P itself?

• Some P4P programs require certain technologies that participating physicians may have to purchase. This will only serve to increase labor and administrative costs and negatively impact on physicians seeking to participate in the program. Where is the evidence that this will not happen?

• Comparing clinical outcomes requires large numbers of equally matched patients, based on factors such as age, severity of disease, complication rates and co-morbid conditions. Do these measures actually exist? Where is the evidence that they are accurate when gleaned from billing and encounter data?

• Patient satisfaction measures are a problem as well. Patients may register dissatisfaction with the type or timely care received from a physician when the problem is actually attributable to conditions imposed by the patient’s managed care plan itself.

• Across most illness conditions, patient satisfaction research has been consistently hampered by serious measurement problems (“How valid and reliable are patient satisfaction data?”  See an analysis of 195 studies. Sitzia J: International Journal for Quality and Healthcare 1999; 11:319-328)

• For P4P principles to be credible, there must be evidence-based measures that are broadly understood and accepted.

• Most providers and plans must be able to actually improve quality by using the measures; otherwise, care may be improved for only a few beneficiaries or instead only the reporting that is merely improved. Where is the evidence about these effects?

• Incentives should not discourage providers from taking riskier or more complex patients. Where is the evidence that this will or will not occur?

• What is the potential for gaming the system?

• Information to measure the quality of the plan or provider should be collected in a standardized format without extensive burden on the parties involved. Where is the evidence that this will be so?

• There must be evidence-based research that all P4P principles are true and valid before they are instituted globally; otherwise, this will be a big experiment based on conjecture and a terrible and unforgiving experiment on the American people.

• Another major problem is that what we term “evidence-based studies” is really based on statistical significance, which might indeed be wrong 5% of the time merely on the basis of chance.

• There is another inherent flaw in so-called “evidence-based medicine” in that sometimes patients are an exception to the rule.  Individually, a particular patient’s needs should not be sacrificed on the alter of conformity to evidenced based studies that are based on large populations.

• ‘Bridges of Excellence’ has calculated that the time spent by physicians in culling data for medical records to qualify for the diabetes program will take about 15 minutes per chart.  Other than these evaluations, there has been very little available data to quantify the potential cost.

• What about the specialist who now accepts hard-to-control diabetics?  Will such patients be readily incorporated into the practice when P4P measures might negatively affect the specialist’s rating and reimbursement for taking care of difficult patients?

• Doctors who currently provide care to a disproportionate number of minority and low-income population are in danger of having their quality of care judged as inferior because of lower than average patient compliance among such populations.

• P4P does not appear to effectively deal with some of the major preventable causes of death and disability and other societal issues, e.g., smoking cessation, alcoholism, obesity, drug abuse, unwanted pregnancy, and physical inactivity.

• There is a problem with an increasing accumulation of medical data that seems to point to different treatment efficacy for different groups of people, different genders, or people of different ethnicity. For example, some data indicate that the response to beta blockers, diuretics, and ACE inhibitors differ in different groups and recent evidence shows that hydralazine is effective and indicated for treating CHF in African-Americans but not Caucasians.

• There is the problem of the uninsured, which represents 48 million US citizens, because the experience of treating these groups will not be captured and therefore might lead to further marginalization of an already disenfranchised group.

• There are significant limitations to what report cards really represent because the majority of encounters in primary care and the management of most chronic diseases receive very little P4P attention. Many of the major causes of sickness and death among minority and low‑income groups, such as strokes, AIDS, and unintentional injury are underrepresented in report cards, as well as diseases that affect only minority groups, such as sickle cell anemia.

• The clinical spectrum covered by report cards may be largely irrelevant to vulnerable patient groups. The very emphasis on disease, clinical procedures, and healthcare status may be misplaced. For example, for many individuals of low-income, key concerns center around the issues of access, physical proximity of services, the extent of financial barriers, and the availability of patients’ ease of communication.

• Focus groups conducted with 35- to 50-year-old individuals revealed widespread misunderstanding of health care report card data.  For example, Medicaid and uninsured patients believe that providers and health plans that have high rates of hospitalization for asthma were providing higher quality care because it suggested an ability to be hospitalized when needed. (In fact, a high asthma hospitalization-rate tends to demonstrate poor quality primary care that could have prevented the need for admission).

Principles

• Guidelines must be easily incorporated in a physician’s practice
• Guidelines must be kept up to date
• P4P Programs must be voluntary
• P4P Programs should be transparent with the patient aware of the incentives and how it may affect their care
• P4P Programs should include new dollars not just reapportionment of existing dollars
• P4P Programs should take into account efforts by physicians even if targets are not reached.
• Systems should not penalize doctors for increased utilization of services if provided to patients as part of a P4P Program. 

• Guidelines must be easily incorporated into practice.

• They should be easy to modify when standards are modified. 

• They should foster, not hinder, the patient-physician relationship.

• Fair and ethical P4P programs should offer voluntary participation and must not undermine the economic viability of physician practices.

• Fair and ethical P4P program must use accurate data and scientifically valid analytical methods.  Physicians must be allowed to review, comment, and appeal results, prior to the use of the criteria because of problematic reasons in any type of reporting.

• Fair and ethical P4P program must provide new funding for positive incentives and not penalize one group of hardworking and dedicated doctors to transfer money to other physicians.

• Any data collection should involve minimal use of physician time; otherwise, it will take away from the time for patient care when most practices are already pressed for time because of an increasing shortage of physicians.

• P4P should be dependent on what physicians can control, not what they can’t control.

• P4P incentives should reward physicians’ efforts, not results, which frequently depend not on the physicians but on their population of patients.

• How will provider directories, which highlight highly rated physicians, handle doctors who only have a small number of patients in a particular plan?  For example, it is absurd to produce any rating for certain conditions or measures of care when there are only 50 or even a 100 patients in a plan, from which there might only be two or three people who have diabetes, two or three people who have CHF, coronary disease, etc.  How can bonuses be appropriately determined with small numbers like this?

• P4P programs should avoid pitting physician’s adherence to program parameters against the best interests of their patients.  Could the program promote patient screening and de-selection?

• The P4P program should offer true bonuses in the form of new dollars and not merely reallocate resources among physicians.  Is the program likely to create winners and losers or is it designed to “raise all boats”?

• What are the dangers that P4P program parameters will be used to establish standards that can be used against physicians in other venues such as credentialing and professional liability actions?

• Paying performance bonuses by reducing payment elsewhere can lead to serious quality of care concerns.

• It was reported in the AMA News (July 4, 2005) that a panel advising Medicare has recommended that the government move some of the money saved from any future pay for performance system from the hospital’s side of the program to the physicians’ side.  If we are truly going to pay for performance, there needs to be real gain sharing by taking some of the money saved from part A into Part B.  Will this be the case?

• Another problem with the CMS approach is that any significant increase in physician utilization of services resulting from compliance with P4P puts the entire physician community at risk for exceeding annual spending targets set by the Medicare formula.  When this occurs, doctors’ Medicare payments are cut in the following year.  The reality is that most physician performance measures focus on providing more care to patients, not less.  Therefore, Pay For Performance and the physicians rating formula are inconsistent because they will punish the physician community for conscientious participation.

• Dr. McClellan, head of CMS, has stated, “We simply don’t have enough well‑developed evidence on how these reporting systems could actually work in practice to lead not only to more predictable payments but also to fewer complications and lower the overall cost of care.” (American Medical News, January 23, 2006).  Shouldn’t we have the evidence or at least make sure the “bugs” are worked out before we proceed?

• There must be a way for physicians to exclude patients from the numerator and denominator if patients do not attend office visits despite reminders, if they decline interventional treatment, if they have not tolerated a medication that is specified and indicated, if they are already receiving maximum dose of the medication and the results are below optimal, or if they have a condition, or an allergy for which the intervention is not clinically appropriate, or if they have a supervening condition that makes treatment inappropriate (for example, someone who has a lipidemia but is terminally ill from cancer).

• Present JCAHO measures for treating community-acquired pneumonia and CHF specify “any patient who has a history of tobacco use within the past 12 months must receive a brochure on smoking cessation.”  What is the evidence that giving a brochure works?  How well does it work?  Most physicians believe that this is really an inadequate recommendation or an inadequate standard and that doctors should instead be appropriately reimbursed for counseling patients about smoking cessation (whether or not the patient actually stops smoking, since it is an addiction with a very low cessation rate).

• An article in The New England Journal of Medicine (September 30, 2004, page 1451) indicated that there is acknowledgment of substantial increases in the administrative burden on physicians who participate in P4P.  The same article suggested that the introduction of such a program in the United Kingdom cannot, at this time, translate into an easy introduction in the United States, as the United Kingdom appears to be better tuned for P4P because more than 90% of physicians there are using computers, whereas in the United States only 17% of practices are computerized.

• Where is the evidence that P4P will represent an overall cost saving, that it will change physician behavior, change patient’s behavior or lead to better care?

• The New England Journal of Medicine (September 30, 2004) also pointed out that some methods of gaming the system might be hard to detect. For example, if 139/89 is deemed to be the target of controlled blood pressure, what is the likelihood of finding BP’s of 142/90 in patient charts? Will there be silly hairsplitting incentives and/or punishments to induce offices to “cheat just a little”?

• The present CMS voluntary program happens to be scientifically invalid.  It will show very skewed results because it represents a very self-selected and unrepresentative group of physicians who participate.

FINANCIAL AWARDS AND INCENTIVES

Principles:

• Financial rewards should be enough to drive behavior
• Financial rewards should be new dollars
• Physicians should not be penalized if targets not achieved

• As an effective incentive, P4P programs that were recently introduced in the United Kingdom for GPs offer bonuses of up to 20% of total present reimbursement. Are managed care companies and CMS willing to follow their example?  In the United States, CMS and managed care companies are talking about a 2% to 3% reward, in addition to cost shifting by taking money from “poorer” performing doctors to give it to “better performing doctors.”

• What will happen as more and more physicians become “better performers”?

•  Will there will be less and less money available to reward them and rewards will become smaller and smaller? What kind of incentives will be left?

•  Incentives for practices must be high enough to absorb the cost of chart review and administrative time required to respond to P4P data requirements. Hence, a 2% to 3% incentive reward is insufficient to compensate doctors for their cost. 

• Practices that do not get the 2% to 3% bonus are even further penalized. How are they going to be able to cope with this?  Isn’t it especially important to raise the performance of poorer performing practices?

• What research exists about adequate incentives for practices?

Principles:

• Need to study P4P Programs to see if patient behavior changes
• Need to study P4P Programs to see if outcomes improve
• Need to study P4P Programs to look for unintended consequences
• Reduced access to high risk or non-compliant  patients
• Adverse effect on care if guidelines lag behind standards of care
• Inability to adjust care to individual patient needs because they may not fit into guideline.

• Where are the data on whether patients will actually be swayed to choose a better performing doctor if they have been satisfied with their present doctor?

• What is the estimates of the fiscal costs to CMS and HMOs to pilot, develop, initiate, publicize, measure, manage, field questions and complaints from patients and doctors, as well as handle medical appeals, and generally administrate P4P programs? How much funding will therefore be taken away from actual patient care?  $5 billion? $25 billion? $50 billion?

• We must remember that there are over 600,000 practicing physicians, 250 million insured in America with 100 different measures and parameters.  These represent a prodigious amount of data.  It actually translates into 15,000,000,000 bits of information.  How much of this will really be useful?  And how much will be ‘garbage science’?

• What is the cost-benefit ratio of the P4P endeavor?  The return on investment?  The true value of such a program?  Does any of this evidence exist?  Shouldn’t this be known before starting such a program?

• What will happen if, for example, one particular brand of drug in a class receives a new indication?  Will the guidelines incorporate that particular drug only, or generalize it to the class?

• What are doctors to do when a newly published data in a major journal showing the benefits of a new therapy or new approach that conflicts with established guidelines?  Should physicians give patients the benefits of this approach or should they rigidly obey the guidelines so as to avoid receiving a negative rating in provider directories?  What about the financial penalty physicians will sustain by not obeying the outdated guidelines?  Isn’t this a disincentive for doctors to continue reading journals and incorporating the latest in research into their practices?  Or should they first check with their “HMO P4P” department to see if this new innovation is allowed?  Will there now be a new pre-authorization number that needs to be called “to get permission without penalty to violate an existing guideline”?

• How long will it take for HMOs or the federal government to incorporate new research findings?  How easily will ‘Guidelines’ be changed when new research shows new findings?  What kind of bureaucratic process will be necessary to change guidelines?

• Will P4P inadvertently bog down and stifle innovation, cutting edge medicine and the incorporation of the latest research and benefits of new treatments and new approaches? 

• Will major academic, specialty and tertiary care centers be rated “poorly” and financially penalized because they are likely to use cutting edge information thereby violating ”present” established guidelines in favor of what will likely be guidelines “of the future”.  Do we want our academic medical centers to suffer such consequences?

• Tertiary care centers often take the most difficult cases.  Will they be at risk of receiving a relatively poor “rating” in a public reporting system (in spite of available severity adjustment indices) because they will be compared to hospitals that handle much easier cases?

• What about standards of care measurements regarding mental or emotional illness? Or are these unimportant conditions?

• What hoops will doctors have to jump through to report patients with white coat hypertension, or the elderly with pipe stem arteries and systolic hypertension (e.g. 178/60) who develop postural changes or adverse symptoms when more medications are added?

• Poorly designed P4P programs with their inherent demand for paper work and electronic health records could undermine their own foundation by driving physicians out of business (National Pay-For-Performance Summit, Los Angeles, Feb 6-8 / 2006, as reported in AAFP News, March 2006).

• It is critical to note that the HHS proposed budget for 2007 calls for initiating P4P Programs, but prohibits these programs from increasing Taxpayer, Medicare or Beneficiary costs while providing no explicit support for EHRs.

• Doctors who invest the additional time and effort in treating challenging poorer populations and more vulnerable groups may not meet the guidelines and be further punished financially for their dedication.

• This might create incentives to avoid caring for various populations of patients such as the less educated, the obese, the noncompliant, the cigarette smokers, the persons who have difficulty with alcohol, the persons who are depressed or anxious, or people who are unable to take charge of their life to take care of themselves properly, etc.

• If rewards are set at reaching 80% of the target guidelines, do we fail to incentivize and recognize practices that go from 40% to 60%, which would represent a significant improvement in quality? Do they get no reward for the major effort that they have made and for the great success that they have accomplished?

• According to Meredith Rosenthal of the Harvard School of Public Health in testimony before the House Committee on Education and Workforce, the Pay for Performance style of reimbursement is still an untested experiment and is not designed to reap cost savings.

Principles

• Must be transparency in the rating systems.
• Physicians must be given the opportunity to review data and correct inaccuracies or explain appropriate variations.
• P4P Programs must respond to concerns of physicians regard data prior to actions being taken, including public reporting

• There must be transparency in the rating system.

• Rating information must be sent to doctors at least three months before final adjudication of the bonus or release of public ratings to afford doctors ample time to respond to inaccuracies.

• Doctors should be given at least 45 days from the receipt of this information to reply to managed care organizations about such inaccuracies.  (In this regard, it is noteworthy that it is not uncommon for managed care companies to send out mailings pre-dated two to three weeks leaving the false impression of having been received by the physician several weeks prior to actual receipt. This dishonest practice must stop.)

• Managed care organizations must reply to physician challenges within a 45-day period.

• An objective ombudsman must be established, chosen by an unbiased entity such as the State Medical Society or a peer review panel, fairly selected and properly balanced, that can conduct a fair hearing when there is dispute.  This must not impose a major administrative or financial burden to doctors.

PRINCIPLES OF DISCLOSURE OF RATING SYSTEM TO PATIENTS AND THE PUBLIC.

• If the physician rating system is made available to consumers, there must be full disclosure by CMS and/or HMOs about the actual criteria that doctors are rated on, the methodology used to arrive at the statistics, the statistical significance of these measures (if any), and the limitations and possible errors and inaccuracies of these measurements.

• Such information must be disclosed to patients in simple, understandable, layman’s language, suitable to the national 8th grade readership level.

• A toll-free telephone number (staffed by qualified personnel) must be listed and provided to patients who can call to ask questions if they do not fully understand the system.  Will call centers be overwhelmed?  Will they be able to address patients’ concerns appropriately, or will ‘voicemail’ and lack of live operator assistance cause more patient alienation?

• How truly meaningful or accurate might these provider ratings be if managed care companies really have an incentive to make their doctors look good in their participating provider directories?  Will health plans thing that if too few doctors look “excellent” or “very good”, it would be an incentive for patients to choose another plan where a greater proportion of the doctors are high performers?   Might employers as well decide to switch to another managed care plan with better performers?

• HMOs must not be indemnified from losses, penalties, and damages resulting from their errors, omissions, oversights, and other mistakes in accurately rewarding doctors financially or in their public rating of these physicians, since doctors may suffer significant financial losses, public and professional embarrassment, loss of referrals, and loss of patients as a result of rating inaccuracies.

• How does one compensate physicians for unfair damage to their reputations?

• How can we possibly erase, adjust or undo an inaccurate rating that has already been circulated in provider and patient manuals?

• Perhaps ratings should never appear in print, but only on-line, so errors can be easily

corrected?

• Since standards of care often change with time, the rating system will often be as much as a year or two behind the current standards of proper care.

• While some believe that the electronic health record will ease compliance with P4P while also reducing paperwork and costly medical errors, the eventual cost savings have been estimated to be no more than 3% (Los Angeles Times, 1/30/06).

• Some EHR systems might require even more time and effort from the user compared to the paper systems they replace.

• There are the very substantial uncompensated costs of EHR and related disruptive effects on the physician’s practice for some 12-18 months until a transition is made and the bugs are worked out.  And it is the physician who pays the price.

• Managed care plans originally took hold and emerged as a major force because of their representations to business, government, and the general public that they would be the vehicles to control health care costs by managing the total medical care of patients.

97  The evidence is that on the whole they have failed both to manage patient care and, ultimately, even to control costs.  Instead, health insurance premiums (the cost to consumers and employers) have skyrocketed 85% in the past five years (New York Times).

• Furthermore, an increasing share of the health insurance premium is diverted to administrative costs, shareholder profits, and gargantuan executive compensation (“As Patients, Doctors Feel the Pinch, Insurer’s CEO Make a Billion,” Wall Street Journal, 4/18/06).  An increasing major share goes to ‘Big Pharma’ as well, with their shareholder profits and astronomical executive compensation.

• With managed care organizations having failed at controlling costs, their latest mantra or raison d'être seems to be that they are the new judge and jury of quality and “value” in healthcare.  Why? Because if they didn’t come up with P4P, there would now seem to be little justification for their continued existence, having failed miserably in their task of cost control.

IN CONCLUSION

• There is a credible school of thought that although P4P might be a noble idea, the difficulties, potential problems, potential harm, and actual benefits are all in question at this time. 

• P4P is truly at a very embryonic stage of development.  Untested, it seems like a truly radical experiment to foist on the American public, doctors, hospitals, and the American health system at this time.

• P4P implementation should be delayed until such time as all the details are worked out, until all the appropriate studies are done, until there is, in effect, evidence-based P4P.

• To implement P4P on a large scale now is premature and reminiscent of the ill-conceived Clinton Health Plan, the problematic capitation schemes, and the failed MSO and PHO movements, all of which seemed like good ideas at the time, but soon became practical disasters.

WE CANNOT AFFORD ANY MORE DISASTERS!

For further information, it is suggested that you consult “pay for performance, the good, the bad, and the ugly” which is free to AMA members. Log on to www.ama‑assn.org/goal/pfp to view the video or download the AMA’s printable guidelines for care for performance.