In 2005, the Centers for Medicare and Medicaid Services announced a policy of rewarding physicians for meeting quality standards in clinical practice, a program known as “pay for performance.” The announced purposes are to ensure best practice patterns among all physicians, reduce unwarranted variations in care, select quality measures that correlate with better outcomes, and reward physicians for the quality of care rather than for providing a service. The Medicare program that just became effective Jan. 1, the Physician Voluntary Reporting Program, includes 36 measures of which 16 are considered the “core starter set” as well as a surgery worksheet that lists five measures.
The purpose of P4P is to pay physicians based on quality and efficiency instead of on a fee-for-service basis. Members of the CMS and the U.S. Congress, which is considering a variety of legislation related to P4P, have been very clear that they believe the physician community, in particular physician specialty organizations, should develop the quality measures for their specialty.
The American College of Surgeons has collaborated with surgical specialty societies through both the Surgical Care Improvement Project and the National Surgical Quality Improvement Program to develop proposals that are based on recognized best practices and avoid imposing undue administrative burdens on surgeons. These proposals utilize process measures related to the prevention of certain complications such as infection and postoperative pneumonia.
A consortium of surgical specialty societies developed the Quality Improvement Framework for Surgical Care, which represents a phased approach for implementing a pay-for-performance program and addresses the difficulty of developing surgically applicable quality of care measurements. The fundamental principles that underlie the proposed framework are:
- The primary goal of pay-for-performance programs must be improving health quality and safety.
- Physician participation in pay-for-performance programs must be voluntary.
- A nonpunitive audit system should be implemented to ensure the accuracy of the data.
- To account for differences across surgical specialties and in the federal government’s ability to collect and analyze meaningful data, any Medicare pay-for-performance program must be pilot tested across settings and specialties and phased in over an appropriate period of time.
- Practicing physicians and their professional organizations must be involved in the design of Medicare pay-for-performance measures and programs.
- Physician performance measures used in Medicare pay-for-performance programs must be evidence-based, broadly accepted, and clinically relevant.
- The metrics must be fair and balanced across specialties and developed using evidence-based work or consensus panels of expert physicians. The metrics must also be kept current to reflect changes in clinical practice.
- Physician performance data must be fully adjusted for case-mix composition including factors of sample size, age/sex distribution, severity of illness, number of comorbid conditions, and any other features that may influence the results.
- The program should foster the patient-physician relationship and must not discourage physicians from treating patients with significant health problems or complications out of fear that they will have a negative influence on quality scores and reimbursement.
- There must be a mechanism for exceptions to pay-for-performance compliance metrics for clinical research protocols and in situations where measures are in conflict with sound clinical judgment.
- Performance measures should be scored against both absolute values and relative improvement in values, as appropriate.
- Medicare must positively reward physician participation in pay-for-performance programs, including physician use of electronic health records and decision support tools.
- Pay-for performance programs must compensate physicians for any administrative burden for collecting and reporting data.
- Pay-for-performance programs must not be budget neutral within the Medicare physician payment system or be subject to artificial Medicare payment volume controls such as the sustainable growth rate mechanism.
- Pay-for-performance programs should not penalize physicians for factors beyond their control.
- Mechanisms must be established to allow performance awards for physician behaviors in hospital settings that produce cost savings outside the physician’s Medicare fee schedule including reducing length of stay, and avoiding readmissions, which substantially reduce hospital costs covered under Medicare Part A reimbursements.
- Physicians must have the opportunity to review and correct performance data, and those data must remain confidential and not subject to discovery in legal proceedings.
The consortium sent these criteria in July to Rep. Nancy Johnson, chair of the House Ways and Means Health Subcommittee, and she introduced the “Medicare Value-Based Purchasing for Physicians’ Services Act of 2005” (H.R. 3617) based on these measures. The legislation would restructure the Medicare physician reimbursement formula to link payment to quality incentives. The essential provisions of the bill include a phased-in, value-based purchasing program over several years starting with voluntary initial reporting beginning in 2007.
Guidelines as Quality Measures
Performance measures selected to evaluate physician performance in patient care should be based on published scientific evidence. Evidence of optimal outcomes and effectiveness frequently is summarized in practice guidelines that are created after a rigorous literature review and assessment of the quality of the evidence. According to the Agency for Healthcare Research and Quality, practice guidelines are formally defined as “systematically developed recommendations, strategies, or other information to assist healthcare decision-making in specific clinical circumstances.”
Guidelines are formulated based on evidence that results from scientific research, and the strength of the recommendations is based on the quality of the evidence available in the medical literature. Evidence is ranked as:
- class I: randomized, prospective clinical trials;
- class II: nonrandomized case series, case-control studies; and
- class III: expert opinion, case report.
By convention, recommendations supported by class I evidence are termed “standards,” while class II evidence recommendations are termed “guidelines” and class III recommendations are “options.” Most available evidence is class II or III because well designed and conducted prospective randomized clinical trials are rarely available.
Surgical care guidelines may include: verification of appropriate side and site preoperatively, antibiotic administration within one hour prior to the beginning in operating room time-out immediately preceding the surgical procedure, appropriate immediate postoperative documentation of the surgical procedure, appropriate postoperative pain management, and appropriate postoperative patient care.
What constitutes appropriate postoperative care may be difficult to determine since the types of procedures performed within a single specialty, like neurosurgery, are extremely variable. Generally applicable measures to most surgical specialties which are incorporated into the American College of Surgeons Surgical Care Improvement Program may include the monitoring of perioperative events such as wound infections, cardiac events, thrombophlebitis, and development of pneumonia.
Even though these are all reasonably determined perioperative care problems, the inclusion of these events in guidelines already has resulted in unintended consequences. An example can be found in cardiac surgery. Cardiac surgeons led the way in the development of guidelines, one of which involves the occurrence of perioperative stroke. Because high incidence of perioperative stroke could negatively impact the participation of cardiac surgeons in the Medicare program, there is an immediate incentive to avoid reporting or to underreport the incidence of perioperative stroke, rather than to report it for eventual improvement in the quality of care.
In neurosurgery, guidelines have been developed to standardize treatment in some areas where controversy or variation in care existed and where evidence supported it, such as in the areas of brain trauma and cervical spine injury. Guidelines in neurosurgery have never been used to prescribe specific norms of care that must be followed by everyone in all circumstances, but simply as pathways to assist physician decision-making. A prime example of neurosurgery guidelines are the lumbar fusion guidelines published in the May 2005 issue of the Journal of Neurosurgery: Spine. Nearly all of the recommendations noted in the guidelines for lumbar fusion are based on class III evidence, and therefore do not fit the true definition of guidelines.
Further, little is known about barriers to physicians’ adherence to clinical practice guidelines. Compliance with guidelines often is limited by a physician’s disagreement with the interpretation of reported clinical trials, and other factors such as potential patient discomfort, and adverse effects. In fact, as Boyd and colleagues report in the Journal of the American Medical Association, the strict application of guidelines can adversely affect patient care. Moreover, the degree of non-adherence to clinical guidelines seems to be independent of the strength of the evidence in support of specific interventions reported in previous trials, even when the evidence presented is considered class I.
Potential Liability of Guidelines for Neurosurgery
A powerful argument against adopting guidelines as the yardstick to measure physician performance is the potential misuse of guidelines by plaintiff attorneys who may introduce them in medical malpractice cases as equivalent to standard of care. Standard of care is a legal concept which is used by attorneys to judge medical performance by focusing specifically on deviations from the norm. A particular act or decision in medicine is considered standard of care when a physician of similar training would have acted in a comparable manner in the care of a particular patient given the same or similar clinical circumstances.
Some argue that the use of guidelines in medical malpractice actions is not necessarily for the purpose of confining the physician to their use as standards of care. In this sense the concept is that plaintiff attorneys will review guidelines when preparing to establish a medical malpractice action, will measure the physician’s performance according to the guidelines, and often will be led away from filing the malpractice action because the physician complied closely with the specific guidelines. In addition, some support the concept that a carefully written disclaimer as part of the guidelines will prevent the attorneys from using the guidelines as the recipe book of medical practice. However, an inquiry with a variety of attorneys conducted by Monica Wehby, MD, illustrates the concern that guidelines are used rather regularly by plaintiff attorneys in medical liability actions:
- In response to the question: “Have you ever seen guidelines used against a physician?” Attorneys from the three major malpractice defense firms in Oregon all responded “Absolutely,” and followed with a litany of recent cases. One stated he had seen dozens of instances in the past year alone.
- Alabama: “The plaintiff always wants what I refer to as the “Reader’s Digest Condensed Version of Medicine,” a set of written standards or guidelines that make medical practice a checklist that a jury of lay persons can apply with confidence. It’s easier to explain to a jury than clinical judgment. The use of clinical guidelines can sometimes tie the hands of the practitioner and is always going to be touted as the equal of standard of care. The jury will use and accept guidelines as the standard of care because it is something they understand. If you are considering promulgating practice guidelines please, please, please don’t.”
- Florida: “The guidelines are always referred to, but people with Florida drivers licenses (our jury pool) think guidelines are the rules, they were broken, and therefore the physician acted wrongly.”
- Texas: “I have had evidence of the physician’s violations of the AAP [American Academy of Pediatrics], ACOG [American College of Obstetrics and Gynecology], and ACEP [American College of Emergency Physicians] guidelines come in against my clients. It is always damaging to the defense and usually raises the settlement value.”
Guidelines are used by plaintiff attorneys like a building code. With a copy of pertinent guidelines in hand, the first thing they do is look for a deviation. They then compellingly present the deviation to the jury, and the burden of proof suddenly shifts to the defendant to demonstrate the validity of his or her actions against the norm established by the specialty’s guidelines. Even an extensive disclaimer attached to the guidelines, such as the ACOG utilizes, has little effect on the jury because the differences between a standard, a guideline, and an option are difficult to communicate. As far as the jury is concerned, you broke your own rule.
The guidelines currently in use in obstetrics and gynecology, anesthesia, emergency medicine, and pediatrics have been written by specialty societies. Perhaps a similar process could evolve in neurosurgery, but the recognized liability risks would accompany guidelines development absent the certainty of improving care.
Reaction to P4P Demands
Across the country, physicians remain resistant and resentful to the implementation of P4P measures. Some may prefer the potential 1 percent to 2 percent Medicare pay cut to the consequences of being bound by reporting performance measures. Some believe that as clinical information becomes available to alter their current standards of practice, it should be incorporated in the form of educational seminars tied to the maintenance of certification process. Many also believe that literature reviews should be published as a review of clinical articles or presented as clinical options, not as practice guidelines.
In the National Review, Dolinar and Leininger reported:
Not only is there little evidence that the government’s version of a “pay-for-performance” scheme would actually work to the benefit of patients, but there is also the likelihood it would do the opposite. Understand that “pay for performance” in Medicare would mean that Congress would pay doctors according to how well they’ve complied with government-defined medical guidelines. That would create another layer of bureaucracy between patients and doctors, and it would involve federal bureaucrats even more in patient care. If Congress wants to improve patient care in a cost-effective fashion, another layer of bureaucracy and red tape isn’t the way to go.
At the October meeting of the Council of State Neurosurgical Societies, a resolution to oppose development of guidelines based on anything other than class I evidence was enthusiastically debated at length. Proponents of guidelines development argued for a proactive approach rather than to allow performance measures to be imposed by outside entities. Supporters of the resolution voiced concerns about the usefulness of guidelines, their application in liability cases, and even the varying connotations of the word “guidelines.” In the end, a substitute resolution passed that asked the AANS and CNS to address the matter through the Quality Improvement Workgroup of the AANS/CNS Washington Committee.
The following month, pay for performance was a hot topic at the AMA House of Delegates meeting. The AMA delegates brought forth three resolutions, 902, 908 and 910, demanding the AMA to actively oppose any P4P initiatives that do not meet the AMA Principles and Guidelines for P4P:
- Resolution 902 asks that our American Medical Association insist that in all medical decisions, the best interests of the patients are the top priority in the doctor-patient relationship and physicians shall not be coerced, forced or required by any means to comply with clinical practice guidelines not felt by the physician and the patient to be in the individual patient’s best interests; that our AMA seek and support legislation and administrative code implementing mandatory reporting or implementation of process measures or guidelines only when the process measure or guideline has been clearly linked to an improvement of outcomes based on class I evidence and consensus position statements of specialty and other medical societies who are qualified to review the measures; and that our AMA seek and support legislation and administrative code protecting a patient’s right to access medical care that is not subject to third party decision-making outside of the doctor-patient relationship and supporting a physician’s right to advise patients based on the patient’s best interests.
- Resolution 908 asks that our American Medical Association strongly oppose any pay-for-performance (PFP) programs or pilot PFP programs that are not fully in compliance with AMA principles and guidelines; that our AMA develop public education materials to teach patients and other stakeholders about the potential risks and liabilities of PFP programs, especially those that are not fully compliant with AMA policies, principles, and guidelines; and that our AMA provide a report back to the House of Delegates at its 2006 Annual Meeting.
- Resolution 910 asks that our American Medical Association oppose the creation of the Medicare “Value-Based Purchasing Program” described in S. 1356, as well as the Deficit Reduction Omnibus Reconciliation Act of 2005 (in the 109th U.S. Congress) as required by AMA policy H-450.947 as the program is significantly out of compliance with the AMA’s Principles and Guidelines on Pay-for-Performance; and that our AMA direct all staff and leadership to aggressively pursue opposition to the Medicare “Value Based Purchasing Program” using its entire means including a significant lobbying, political and public relations campaign.
The testimony on these resolutions was impassioned and diverse. A general theme that emerged, however, was that the AMA should not compromise its well-developed and comprehensive Pay-for-Performance Principles and Guidelines as a condition to receiving a physician payment update. The Reference Committee agreed with several comments that any pay-for-performance model must not be based on cost containment, but rather on ensuring quality of care, fostering the patient-physician relationship, offering voluntary participation, using accurate data and fair reporting, and providing fair and equitable program incentives.
The Board of Trustees Report 19 recommended that the AMA: (1) continue to communicate strong objections to the CMS’ Physician Voluntary Reporting Program; and (2) work with other federation organizations to express organized medicine’s strong concerns on the proposed implementation of the Physician Voluntary Reporting Program and to offer assistance to rectify deficiencies in the program. There was a clear and potent sense of frustration that the CMS, unwilling to provide an administrative fix to the flawed sustainable growth rate formula, would expect physicians to voluntarily take on a new reporting obligation.
As a result of the pressure from the AMA and organized surgical specialties, the U.S. House of Representatives passed the Budget Reconciliation Bill in December. The bill included several elements demanded by physicians and omitted pay-for performance provisions. The bill met significant opposition in the Senate, and after extended and contentious deliberation the Senate amended the bill, forcing a return of the legislation to the House for another vote. The legislation ultimately signed by the president on Feb. 8 retained omission of pay-for-performance provisions.
Another Balancing Act The value of improving the quality of care by guidelines development as a measure of clinical performance must be balanced against the quality and soundness of the data on which the guidelines are based and their clinical relevance as well as their ability to influence various factors such as: physician behavior; the potential abuse or misuse of the guidelines by plaintiff attorneys in their pursuit of medical malpractice actions; and the potential abuses that may result from some individuals who may want to game the system for their financial benefit.
The financial return that P4P offers to physicians represents less than 2 percent of what they are being reimbursed in an already significantly discounted payment plan. Far greater motivators for the vast majority of U.S. physicians are pride in their work and the desire to decrease patient suffering, to use their skills to meet specific challenges, and to maintain a sterling reputation in the community.
The superficial financial rewards of Medicare’s pay-for-performance program would likely create incentives to rig the system in several ways that may cause the quality of healthcare to decline, even if a handful of indicators seem to be improving. Quality assurance in medical practice must remain the purview of practicing physicians and their respective professional organizations and should not be mandated by a paternalistic government bureaucracy that likely will make of quality assurance an abysmal failure.
Fernando G. Diaz, MD, PhD, is chair of the Council of State Neurosurgical Societies, www.csnsonline.org.
For Further Information
|