Who Can I Trust? Ensuring Physician Competence

Abstract

Gwyneth Vives, a scientist at Los Alamos National Laboratory in New Mexico, suffered a complication and bled to death 3 hours after giving birth to a healthy boy in 2001. It was 4 days before Christmas. Vives suffered a vaginal tear and other lacerations during the delivery that caused profuse bleeding. Her obstetrician, Pamela Johnson, was sued for failure to order a blood transfusion for Vives as well as abandonment since she had turned over repair of the vaginal tear to a midwife. Two other patients also sued Johnson. Jean Challacombe alleged that Johnson tore her bowel and uterus while doing a dilation and curettage the same day Vives died. Tanya Lewis accused Johnson of doing an unnecessary hysterectomy.

Gwyneth Vives , a scientist at Los Alamos National Laboratory in New Mexico, suffered a complication and bled to death 3 hours after giving birth to a healthy boy in 2001. It was 4 days before Christmas. Vives suffered a vaginal tear and other lacerations during the delivery that caused profuse bleeding. Her obstetrician, Pamela Johnson, was sued for failure to order a blood transfusion for Vives as well as abandonment since she had turned over repair of the vaginal tear to a midwife. Two other patients also sued Johnson. Jean Challacombe alleged that Johnson tore her bowel and uterus while doing a dilation and curettage the same day Vives died. Tanya Lewis accused Johnson of doing an unnecessary hysterectomy.

Johnson had been forced to leave a previous job at Duke University Medical Center in North Carolina because of a “high surgical complication rate” and the “worst QA (quality assurance) file of anyone at Duke.” At least three patients had filed claims against Johnson for malpractice . Later, Johnson lied to get her New Mexico license, saying she had never lost hospital privileges, according to an order of the New Mexico Medical Board [1].

“It’s not bad people, it’s bad systems,” we said. But it has been a hard sell. When something bad happens, the natural reaction is to blame, to point the finger at the person who made the mistake, the bad doctor. We now know that this is both wrong and ineffective. Most harm , most errors —probably 95% or more—do, in fact, result from bad systems that lead good people to do bad things. That concept has been the main driver of patient safety: to get people to think of errors and harm as the result of faulty systems, not faulty people.

But there are some “faulty” people—doctors whose incompetence or negligence harms and kills patients. “That’s not a systems problem,” people would say. Ah, but it is. Our doctors are educated and trained by a system, certified by a system, monitored by a system, and disciplined by a system. What are those systems? And do they identify doctors when they begin to fail, assess them, and do something about it—before they hurt someone? A prevention system, or at least an early warning system. A reasonable question. Indeed, a vital question.

The System We Have

The system we have for producing a competent physician is composed of several interdependent systems. We have a rigorous educational system for medicine. Everyone knows that medical education is very difficult, intense, detailed, and challenging. Medical school is hard to get into, the bar is high, and graduates are well-equipped with scientific knowledge when they emerge. This is followed by 3–5 years of residency training and additional years of subspecialty fellowship training, essentially an in-hospital graded experience organized by specialty and culminating in examination and certification of competence by the specialty board (“Board certification”) .

The system for ensuring the continuing competence of the practicing physician also has several parts. The main responsibility falls to the individual specialty boards, who, in conjunction with their association, the American Board of Medical Specialties (ABMS) , attempt to ensure continuing competence for the 85% of physicians who are certified by repeated assessments of their diplomates (the word for those certified) through maintenance of certification programs.

State licensing boards exercise responsibility for continuing competence of physicians through periodic relicensing. All but two rely largely on physicians certifying that they have completed a required number of continuing education courses and truthfully answering relicensing application questions about such things as malpractice claims, other civil lawsuits, criminal charges, illness and substance use, and even whether they have paid their taxes.

At the hospital or practice plan level, the system for ensuring continuing competence is credentialing , a process that determines whether a physician has admitting privileges to the hospital (or practice in a group) for their patients. Privileges are conferred annually or biannually by a committee of physician peers based on the recommendation of the department chair.

What’s the Problem?

The problem is that these systems are not coordinated, and they don’t work very well. Despite the many layers of responsibility and the array of mechanisms for ensuring safe and competent care, too many physicians fall short, and too many patients are harmed. Let’s look at the facts.

Direct measures of physician performance are hard to come by. There is no nationally standardized system for routine measurement of outcomes of physicians’ treatments. An indirect measure of incompetence is malpractice claims, but only claims that result in a payment to the patient are recorded, about a fifth of claims [2]. In 2019, 8378 payments were made for claims against physicians, down significantly from 16,116 in 2001 [3]. (Perhaps as the result of improved disclosure policies?—see Chap. 19.)

Another indirect measure is disciplinary actions by state medical boards . In 2017, 4081 physicians were disciplined by state medical board , including 1147 reprimanded (i.e., censured), 1343 restricted, and 264 who had their licenses revoked [4].

Malpractice claims and disciplinary actions by state boards capture only the proverbial tip of the iceberg. Most negligence is not reported, and few patients sue (see Chap. 1). For each of these cases, there are dozens that are not reported and many more instances of substandard care that results in patient harm .

More information is available about behavioral problems . Studies of disruptive behavior are disturbing. These include angry outbursts, verbal threats, shouting, swearing, degrading and demeaning comments, and threats of physical force, as well as shaming and sexual harassment [5].

Surveys of nurses show that more than 90% report experiencing such abuse [6], many of them repeatedly. Abuse of medical students is also common. Annual surveys of graduating medical students by the Association of American Medical Colleges (AAMC) show that 12–20% report abuse [7], although other data suggest it is much more common [8, 9]. In one survey of students from twenty-four different medical schools, 64% reported at least one incident of mistreatment by faculty, 76% by residents [10].

Residents in training are also victims. Half of 1791 residents in 1 survey reported being subjected to bullying, belittling, and humiliation [11]. A meta-analysis of 52 studies of residents showed that the prevalence of intimidation, harassment and discrimination was 64% [12].

Patients are also targets of abuse . Surveys show that 13–27% of patients report problems with doctor communication [13]. Patient interviews show the percentage is closer to 50% [14].

Why Doctors Fail

Why? Why do physicians who have successfully completed medical school and a rigorous residency training fail to maintain their competence or develop patterns of disruptive and unsafe behavior that compromise their ability to give high-quality, competent care? There are many reasons. Some succumb to the urge to establish a large practice to fulfill monetary needs or underlying feelings of inadequacy and then find its demands more than they can cope with. Others become overconfident and become unable to acknowledge shortcomings. Some are lazy and just don’t keep up.

But for most, the causes are more mundane. Like everyone else, physicians have mental and physical health issues. Major depressive disorders occur in 16% of the public. The extent among physicians is unknown, but higher suicide rates—40% higher than the public for male physicians and 100% higher for female physicians—suggest depression is also probably more common [15].

The extent of physical illness among physicians is also unknown, but a reasonable estimate is that at least 10% of doctors must restrict their practices for several months or more at some time in their 40-year careers because of disabling illness [15]. Nor are physicians exempt from cognitive decline as they age, although we have no data. Approximately 10–12% of physicians will develop a substance abuse problem at some time in their careers. About half of these are alcohol dependence, the rest, opioids and other drugs [16].

Stress is another factor. Physicians are subjected to unique stresses that can lead to dysfunctional behavior. Overwork, sleep deprivation , decreasing reimbursement, and pressure to see more patients are common. In recent years, the fraction of physicians exhibiting burnout has skyrocketed [17]. Young physicians worry about achieving a work/family balance and paying off their educational debts, which in 2018 averaged $196,520 at graduation. Stress leads to isolation and maladaptive coping strategies, such as alcohol or drug abuse .

Putting this all together, the conclusion is stunning. As John Fromson and I wrote in Problem Doctors: Is There a Systems-Level Solution?: “When all conditions are considered, at least one third of all physicians will experience, at some time in their career, a period during which they have a condition that impairs their ability to practice medicine safely” [15].

The question is: What do we do about it?

Who Is Responsible for Ensuring Physician Competence and Safety?

Who is responsible for making sure that physicians are competent and safe? The answer is disarmingly simple: physicians. Society has given physicians an implicit contract: it grants them incredible powers to cross otherwise sacrosanct boundaries—to learn our most intimate thoughts and invade our bodies and our psyches—in return for the pledge that the profession will use its knowledge and skills for the good of society. It grants the profession substantial autonomy to determine its own educational standards and the right of self-regulation.

The essence of medicine’s contract with society is professionalism , the commitment of the physician to place the interests of the patient above their own, to maintain their skills, and to ensure that their colleagues do so as well [18].

One way physicians have met this obligation is through specialty societies, the AMA , and state medical societies, which from their origins have considered improving the quality of practice of their members their first priority—their purpose, really. The American College of Surgeons (ACS) , for example, was the first to set standards for hospitals, leading ultimately to the formation of the Joint Commission .

Specialty society annual meetings are largely devoted to learning, both from formal instruction and from research presentations. Larger societies, such as the American College of Physicians , American Association of Family Practitioners, and American Society of Anesthesiologists, have extensive extramural continuing education programs and online resources. State medical societies also sponsor educational programs. These programs assist physicians in accruing specific hours of medical education that are required for relicensure by their state boards of medicine.

Physicians have also met their obligation by developing professional organizations that set standards and exercise oversight. The primary responsibility for ensuring physician competence in the USA rests with two national organizations: the American Board of Medical Specialties (ABMS) , whose specialty boards examine and certify practicing physicians, and the Accreditation Council for Graduate Medical Education (ACGME) , which sets standards and oversees physician residency education.

American Board of Medical Specialties

Specialty board certification is the essential badge of quality for physicians. For decades, becoming certified required only that the physician pass a rigorous written and oral examination at the conclusion of residency. You were certified for life. That began to change in 1969 when the newly formed Board of Family Medicine required that diplomates be reexamined every 10 years to maintain certified status. The process of recertification gradually spread to other specialties.

By the late 1990s, just as the patient safety movement was beginning to gain momentum, the leadership of the ABMS realized they needed to do much more to ensure their diplomates were competent and to assure the public that was so.

In a rather remarkable joint effort, ABMS and ACGME came together in 1999 to explicitly define physician competence . They described six domains of clinical competency that physicians would be expected to achieve and maintain [19] (Box 20.1). The six competencies were adopted by the individual specialty boards as the basis for assessments of physicians. The ACGME adopted them as the framework for progressive training in a specialty.

Box 20.1 Six Domains of Competency

• Practice-based learning and improvement: show an ability to investigate and evaluate patient care practices, appraise and assimilate scientific evidence, and improve the practice of medicine.

• Patient care and procedural skills: provide care that is compassionate, appropriate, and effective treatment for health problems and to promote health.

• Systems-based practice : demonstrate awareness of and responsibility to the larger context and systems of health care. Be able to call on system resources to provide optimal care (e.g., coordinating care across sites or serving as the primary case manager when care involves multiple specialties, professions. or sites).

• Medical knowledge: demonstrate knowledge about established and evolving biomedical, clinical, and cognate sciences and their application in patient care.

• Interpersonal and communication skills: demonstrate skills that result in effective information exchange and teaming with patients, their families, and professional associates (e.g., fostering a therapeutic relationship that is ethically sound, using effective listening skills with nonverbal and verbal communication, working as both a team member and at times as a leader).

• Professionalism : demonstrate a commitment to carrying out professional responsibilities, adherence to ethical principles, and sensitivity to diverse patient populations.

Adapted from Ref. [19]

In addition, the periodic recertification examinations would be replaced by continuing maintenance of certification (MOC) , which required the physician to demonstrate a commitment to lifelong learning, self-evaluation, and improving their practice and to prove it through periodic assessments.

Each specialty board devised its own MOC program, tailoring the six competencies to its individual needs and defining how to meet the requirements. The certifying board would then “continually” determine whether or not the physician is in compliance with its MOC requirements. Finally, there was an answer to the patient’s concern, “I know he was competent when he was certified, but is he competent now?”

Boards differed greatly in how they assessed compliance . For most, physicians were required to periodically document that they had maintained the core six competencies . A four-part assessment was designed to test their medical knowledge, clinical competence , communication skills, and quality of care. Approaches have included patient registries, audits, peer review, and comparison to national benchmarks. Another is to give credit for participating in hospital quality improvement projects.

Physicians pushed back. Naturally skeptical, they had to be convinced that the process was relevant to their practices and would improve quality of care at a time when they felt overworked and underpaid. However, by 2012 about half of all certified specialists had complied [20].

Older physicians who had been exempted from the 10-year relicensing requirement when it began also sat this one out. As of 2012, of the 66,689 diplomates of the American Board of Internal Medicine who held only the old time-unlimited certificates, only 1% chose to become recertified through MOC [20].

The concept of the six competencies was truly brilliant. Its explicit definitions made it possible to measure competence for the first time. The ACGME incorporated them in standards for residency training . Medical schools adopted them to structure curricula, and the Joint Commission made them requirements for hospital evaluation of their physicians. CMS gave physicians a bonus on their Medicare reimbursement if they participated in its Physician Quality Reporting System and MOC .

Accreditation Council for Graduate Medical Education

The other professional organization responsible for ensuring competence of physicians is the Accreditation Council for Graduate Medical Education (ACGME) . Not unlike the Joint Commission , the ACGME accreditation process had for years focused on structural aspects of training programs, such as qualifications of the program director and number of teaching cases, plus a few outcomes, especially the percentage of graduates who passed certifying examinations.

Following the development of the six competencies , ACGME in 2001 launched the Outcome Project, which requires residency training programs to configure curricula and evaluation processes in the framework of the six competencies .

In 2011, in conjunction with changes in duty hour limits, a major emphasis was begun to improve supervision and providing a safe and effective environment for care and learning: the Clinical Learning Environment Review (CLER) Program.

Through frequent site visits independent of the accreditation process, CLER focuses on the resident experience and progress in six areas: patient safety; health-care quality and reduction in health-care disparities; care transitions; supervision ; fatigue management, mitigation, and duty hours ; and professionalism [21]. “Milestones ” were developed that describe the skills, knowledge, and behaviors in the six areas that residents are expected to reach at each level as they progress through their training.

These ACGME programs are described in greater detail in Chap. 18.

The Joint Commission

The Joint Commission plays an important role in enabling and ensuring physician competence through its oversight role with hospitals. Not only does it require hospitals to have systems and programs that foster quality and safety, which of necessity involve physicians, hospitals are also required to have programs to oversee and enhance physician performance . See Chap. 12 for more details.

State Licensing Boards

An interesting aspect of self-regulation is state regulation. While that sounds like an oxymoron, it is not. State medical licensing boards have legal authority to hold physicians accountable for competent practice, but for many years they were composed entirely of physicians, who were deemed the only ones qualified to judge other doctors. In recent years lay members have been added in some states, but physicians still dominate.

State boards exercise their authority primarily through licensing. Initial licensing for American medical school graduates requires passing three examinations taken in medical school and the first year of residency. Subsequently, physicians are required to complete a certain number of hours of continuing education annually and pay a fee to renew their licenses. State boards can also require physicians to undergo evaluations to ascertain knowledge and skill and require educational remediation and/or rehabilitation of physicians who have physical, mental, or substance use disorders .

Oversight is typically passive. Rather than actively monitoring or auditing performance of doctors in practice, boards tend to function in a reactive mode, responding to malpractice suits, patient complaints, and the occasional problem physician referred by a health-care organization.

State boards can place physicians on probation, censure, ask them to sign a letter of agreement to change behavior, restrict practice, or remove their license to practice. But they are reluctant to do so. Physicians on boards are very sympathetic to their colleagues, in part because they are aware of their own vulnerability. Of the nearly 600,000 physicians in practice in 2017, only 4081 (0.7%) were disciplined by their state boards [4].

Boards are particularly reluctant to take away licenses because doing so is an existential threat to the physician, rendering them unable to practice. As a result, it rarely happens. Licenses were suspended or revoked for 904 physicians (0.15%) in 2017 [4].

Boards also have a long history of being very forgiving of those with psychoactive substance use disorders . (William Halsted, legendary surgeon at Johns Hopkins Hospital and founder of the residency system, was a known cocaine addict.) Physicians with a substance use problem that interferes with their ability to practice medicine are usually required to enter into a 3–5-year monitoring agreement that includes mandatory random urine testing, workplace monitoring by peers and supervisors, attendance at meetings like A.A. or N.A., and seeing a therapist or alcohol/drug counselor. If they fail to follow through, they may have their license to practice suspended until a significant period of being clean and sober is once again documented.

While these practices accord with current thinking that addiction is a disease and not a moral failing, the difference from other fields, of course, is that impaired physicians put patients at risk. Forgiveness can be carried to extremes. For example, a Virginia psychiatrist was in drug rehabilitation 9 times and relapsed at least 12 times during a 10-year period before the medical board took away her license [22]. In a five-year period, only 1400 physicians across the country were disciplined for substance abuse and reported to the National Practitioner Data Bank [23].

The National Practitioner Data Bank was established by Congress in 1986 to stop doctors from escaping troubled histories by having a central location where any sanctions or malpractice verdicts could be recorded. Names are not made public, but they are available to state licensing boards , hospitals, and other health-care entities, including federal agencies, who are required to consult NPDB prior to hiring.

Nevertheless, hospitals and boards have dragged their feet on complying, reluctant to tarnish a physician’s reputation or restrict their ability to practice. Various tactics are employed to circumvent the requirement to report physicians, most commonly the hospital rather than the physician paying the settlement in a malpractice case.

Nearly 54 percent of all hospitals have never reported a disciplinary action to the data bank. For example, in the Vives case mentioned above, no one told the data bank that Pamela Johnson had been forced to leave her job at Duke. Enforcement is no better: no fine or penalty has ever been levied according to the federal Department of Health and Human Services, which oversees the system.

Federation of State Medical Boards

The national voice for state boards is the Federation of State Medical Boards (FSMB) . All 71 state medical and osteopathic boards are members. The FSMB is the spokesperson for issues related to regulation and discipline. It proposes policy changes and facilitates collaborative efforts of state boards and other entities. With the National Board of Medical Examiners (NBME) , it sponsors the US Medical Licensing Examination that is required of all medical school graduates for medical licensure.

FSMB recommends standards, but it has no enforcement power. In 2004, it promoted a radically new policy—“State medical boards have a responsibility to the public to ensure the ongoing competence of physicians seeking re-licensure”—i.e., meaningful maintenance of licensure like the maintenance of certification programs that were being developed [24].

In 2010, FSMB expanded this to declare that as a condition of license renewal, physicians provide evidence of participation in a program of professional development and lifelong learning based on the six ABMS competencies [25].

It then took the step that quality and safety experts had long called for to align licensing and certification: participation in the MOC process of their specialty board would satisfy the standards for relicensure [26].

FSMB also plays a key role in the assessment and rehabilitation of problem doctors. While many state medical societies have monitoring programs for doctors with alcohol and substance use disorders , fewer address knowledge and skill deficits, personality disorders, technical and cognitive deficiencies, or disruptive behavior .

A joint program of FSMB and NBME , the Post-Licensure Assessment System (PLAS) , administers a standardized examination of clinical knowledge to physicians referred by state medical boards or by themselves. If results are unsatisfactory, the physician may undergo an additional assessment and then choose (or be required) to participate in a remediation program [27].

Alternatively, physicians may be evaluated by the Physician Assessment and Clinical Education (PACE) Program , founded in 1966 at UC San Diego School of Medicine long before FSMB took on this responsibility. The program assesses physicians referred by state boards as a condition of maintaining their licenses. It conducts a rigorous evaluation of a physician’s ability to safely practice medicine. They undergo an oral clinical examination, clinical observation, and physical and mental health screening. PACE offers remedial courses in anger management, communication, professional boundaries, prescribing, and medical record keeping [28].

A number of other programs have been developed in recent years for doctors and other professionals with problems. For all, the goal is to enable dyscompetent professionals to undergo remediation and training so that they can remain in practice.

Unfortunately, there are many barriers to physicians’ participation in these programs. Foremost are financial. In our fee-for-service health-care system, the physician is in a triple bind. Not only do they lose income while undergoing rehabilitation, they often have to pay a substantial fee for it, and if they are absent for more than a few weeks, their practice deteriorates as patients find other doctors. In a more rational system, their employer would maintain their salary and pay the costs of rehabilitation.

Another problem is that if residency retraining is needed, it is difficult to find programs that are willing and able to add the physician to their roster of residents, even for a short time. Similarly, their colleagues and hospital may be reluctant to take on responsibility or potential liability for supervising their practice. Doctors are uncomfortable supervising their peers.

Experience shows that performance problems can be solved or significantly ameliorated for the vast majority of physicians. Few need to be, nor should be, removed from practice. We know what to do. Making it happen is another matter. Rehabilitation and remediation are still very much a work in progress.

New York Cardiac Advisory Committee

Perhaps the most effective—and unique—instrument of state regulation is the New York Cardiac Advisory Committee . In the 1950s, long before there was national interest in improving quality or safety, the Health Department of New York convened a group of respected cardiac surgeons and cardiologists to oversee the newly developing field of cardiac surgery. The committee was an outgrowth of the state certificate of need program that regulated which hospitals could establish new programs. It had the power to limit the number of hospitals performing cardiac surgery. Its responsibility was to establish and maintain high-quality programs geographically distributed to meet the needs of the state’s population.

In 1989, responding to the concern that its comparisons of mortality among hospitals were not valid because they were not adjusted for risk, and recognizing the need for a data-based approach, the CAC established the Cardiac Surgical Reporting System (CSRS) to develop risk-adjusted measures and collect data on outcomes of coronary artery bypass graft (CABG) surgery. For the first time, the adjusted outcomes of all cardiac surgeons and all hospitals performing cardiac surgery were measured and reported publicly.

The initial findings showed wide variations in 30-day operative mortality with low-volume surgeons and low-volume hospitals faring the worst. High-outlier hospitals were put on probation. The responses were prompt. The survival of their programs at risk, most of them undertook a variety of actions to improve their programs: establishing full-time chiefs, replacing chiefs and poor-performing surgeons, adding cardiac anesthesiologists and nurse specialists, etc. The results were dramatic. Within 3 years, mortality dropped 41%, giving New York the lowest CABG mortality of any state, a status it has maintained [29].

The reports attracted intense media attention in the early years, causing concern about shaming and government interference. Mortality for hospitals and surgeons were reported in the newspapers. Was that appropriate? Certainly, the public has a right to know. This notion, now well-accepted and enshrined in Hospital Compare and other public data, was a radical idea at the time and hotly debated. There is little question, however, that the public release of the information was a key motivator for change.

As the CAC began to measure risk-adjusted outcomes of CABG surgery in 1989, they approached our team at RAND to do an appropriateness study of CABG , angioplasty, and coronary angiography. Our earlier work had shown high rates of inappropriate use of these procedures. The CAC and the state health department wanted to know if their efforts resulted in lower rates in New York. With their collaboration we were able to get funding for the study, which we carried out over the next 2 years.

The results confirmed the higher quality of cardiac surgical care in New York. The inappropriate rate for CABG was 2.4%, far lower than the 14% found in a previous study in several other states. Mortality was also low, 2.0%, far lower than the national average of 5.5%. The inappropriate use rate for angioplasty was also low: 4%. These added to the evidence that close oversight and the feedback of risk-adjusted data are powerful motivators for quality [30,31,32].

The CAC program has continued to be successful. It is a superb example of the power of intelligent, well-managed regulation to ensure quality and safety of health care. Unlike other states, in New York the health commissioner has the authority to require reporting, to carry out audits to verify data quality, and to establish the oversight committee and the power to shut programs down.

Involving the state’s leading cardiac surgeons and cardiologists in the advisory committee gave credibility to its decisions and acceptability to the cardiac surgical community. The focus on objective evidence provided a powerful incentive for poor performers to improve [29]. The program is a model of effective regulation.

The Civil Justice System—Malpractice Litigation

Finally, when all else fails, the legal system steps in. Doctors can be sued and forced to pay substantial compensation if their performance can be proven to be negligent . The legal definition of negligence is quite simple: failure to meet the standard of care . Proving that is another matter. In the end, relatively few patients are compensated. The Harvard Medical Practice Study found that fewer than 10% of patients harmed by negligent care ever sued [33]. National studies show that fewer than half of malpractice suits result in a payment to the patient [2, 34].

But negligent care is only responsible for a small fraction of serious medical injury . The vast majority of injured patients have no recourse to the legal system. Malpractice litigation also fails to achieve its other purported objective: deterring bad behavior in the future. There is no evidence this happens. Physicians see cases as one off, bad luck, and unjustified. They often don’t believe they have done anything wrong. The process of being sued is devastating for physicians, however, as is discussed in more detail in Chap. 19.

A serious defect of the current system is that malpractice settlements are usually sealed, prohibiting any party from making the information accessible. Not only does this cloak of secrecy prevent the medical team from learning from the event and fixing the faulty systems, it keeps vital information away from state boards and future patients.

Overall, malpractice litigation is an ineffective tool for ensuring or improving physician competence . Interestingly, fewer patients are suing. Malpractice payments dropped from 16,116 in 2001 to 8378 in 2019 [3]. It is tempting to attribute this to a reduction in patient injuries or to improved disclosure practices, but there is little evidence for either.

A far better legal approach would be enterprise liability , in which the institution, not the physician, is responsible for compensating patients for the costs of harm . Hospitals and health-care organizations would be sued instead. It makes sense. If, as we maintain, harm results from failed systems (including systems for ensuring physician competence ), then it is the party responsible for the systems—the organization—that should be held accountable for their failures. Indeed, if we were really serious about this, we would require hospitals to compensate patients for all costs of the harm we have caused, even when no error is identified: no-fault compensation —as was recommended by the Harvard Medical Practice Study 30 years ago.

Hospital Responsibility for Physician Performance

As in politics, all quality is local. Medical specialty boards set standards, examine, and certify; states license and discipline; but meaningful oversight of physician performance , what happens in everyday practice, takes place where care is delivered. For 80% of physicians, that is the hospital. For others it can be their large multispecialty group. But for practitioners in solo or small group practice, such as primary care, psychiatry, and dermatology, oversight is often quite lax.

Hospital oversight is through credentialing committees, groups of physicians appointed by the medical staff who annually or biannually decide on admitting privileges and what procedures a doctor may perform. It is awesome power, second only to state licensing. If they are unable to admit patients to a hospital, most physicians cannot practice. They are professionally dead. Every hospital has a credentialing committee. Medical specialty boards are the carrot, credentialing committees are the stick.

The process that most credentialing committees use for carrying out this responsibility is quite simple: they rely on the recommendation from the specialty department chair. Typically, this is a pro forma process unless the department chair recommends against it. Then it can get very messy.

So, where “the rubber hits the road,” where the action takes place to ensure physician competence , is the department. The department chair is ultimately responsible for assessing the competence of every member of the department. How do they do it?

Until very recently, assessment has been informal, especially in smaller private hospitals where the chair has little authority. The chair relied on personal knowledge about the physician and feedback from peers. Absent serious complaints from patients or staff about the physician’s conduct, approval was routine.

Few department chairs actually reviewed patient outcomes or conducted peer assessment of performance. Annual physical examinations are still not required. Random drug testing is rare and hotly resisted by many physicians as an affront to their professionalism . Cognitive testing is almost nonexistent.

The good news is that methods for monitoring clinical performance have improved greatly in recent years. To be objective, evaluation must be based on data: compliance with standard practices and outcomes, how well patients do. While measuring outcomes is easiest with surgical patients, many “medical” outcomes are now also collected routinely. Individual results can then be compared with national and local norms to identify outliers who need attention.

The Joint Commission now requires that physicians currently on staff have an annual Ongoing Professional Practice Evaluation (OPPE) . This is a summary of ongoing data collected for the purpose of assessing a practitioner’s clinical competence and professional behavior. Newly hired physicians and those already on staff found to have competency issues on their OPPE are required to have a Focused Professional Practice Evaluation in which the medical staff evaluates the privilege-specific competence of the practitioner [35].

Psychosocial aspects of physician competence —communication skills, interpersonal relations, and ability to collaborate—have long been considered unquantifiable. They have traditionally been assessed informally through conversations with peers and coworkers. Personality or interest tests and the like have been tried and found not to be reliable. But one method of evaluation does produce data that is reliable and has proven to be quite useful: multisource feedback, popularly called “360” evaluations .

Multisource Feedback

Multisource feedback (MSF) is a formalized method of obtaining feedback about an individual’s performance from those with whom they interact. Since the late 1990s, it has been used to assess physicians by Lockyer in the Physician Achievement Review (PAR) program in Alberta, Canada [36], but is now being increasingly used in US hospitals. The PACE program in California has used it for some time to evaluate physicians referred for problem behavior.

The process begins by having the physician and their peers, nurses, residents, and patients complete a questionnaire of 10–40 items that assess clinical behaviors, such as communication, collaboration, professionalism , interpersonal, and management skills. Typically, 7–15 individuals in each of these groups complete the questionnaire, rating the physician on a five-point scale. The results are tabulated by group, and mean scores are compared to the physician’s self-assessment for each item. The department chairman then reviews the data with the physician to identify areas for improvement. Studies have shown that MSF has high reliability , validity, and feasibility [37].

The impact of the 360 review can be very powerful. In a pilot study some years ago in one department in a Boston hospital, we found that, as in Lake Wobegon, all physicians rated themselves above average for almost all questions. Peers tended to agree, but resident and nurse ratings were sometimes quite a bit lower, especially regarding interpersonal relations. Feedback of this information to the physician was always a surprise and sometimes emotionally very disturbing. Several were reduced to tears. It was a powerful motivation for change.

MSF is increasingly being used in the USA. ABMS now recommends that specialty boards use MSF to assess professionalism and knowledge, and ACGME requires training programs to use multiple evaluators to provide objective performance evaluation of residents. The Pulse 360 Program creates and sells 360 feedback tools and training programs for health care. It is used in over 200 hospitals [38].

Support of Physicians with Problems

With the demands of MOC , methods for evaluation and support are improving. Specialty boards, especially the ABIM , have become more engaged with hospitals in providing continuing education, translating standards into practice, and collecting outcome data to measure performance. Blue Cross Blue Shield Association, CIGNA HealthCare, Humana, and Wellpoint have incorporated them in their quality recognition programs.

But serious behavioral problems are often managed poorly. Department chairs may lack the training and skills to deal with them. Many fear confrontation and avoid it if possible. Peers are reluctant to be involved, valuing their own independence and respecting that of others.

A major barrier is that disciplinary action will often be vigorously resisted by the offending physician, who may even sue the department or the hospital. This leads to bad publicity in the newspapers and requires a number of doctors to spend many hours in depositions or hearings—a messy business, indeed. No wonder doctors shy away from judging their peers.

How Should it Work? The Ideal System

There must be a better way to ensure physician competence and improve quality of care. There is. It is for the hospital (or practice) to perform a meaningful evaluation of every physician every year using a routine, formal, proactive system of monitoring with validated measures , followed by action to remedy shortcomings when they are discovered. Some years ago, John Fromson and I proposed that the system must have three characteristics [15]:

• First, it must be objective, i.e., assessment must be based on data: patient outcomes data and compliance with performance standards , not on subjective judgments of personality or motivation.

• Second, it must be fair. All physicians in the organization must be evaluated by the same system, not just suspect individuals.

• Third, it must be responsive. When problems are identified, they must be treated promptly. There is no point in evaluation if nothing comes of it. Most physicians with problems will only need feedback. They can and will self-correct. Others may need counseling. Some may require referral to an outside program for assessment. Retraining may be needed.

An effective system is proactive. It is based on the notion that subpar performance can be objectively defined, routine monitoring can detect problems early, and the responses to deficiencies will be prompt and constructive.

The point is not to identify “bad apples” and throw them out, but to detect deficiencies early and correct them before patients are harmed, to enable good doctors with minor problems to become better, and to help those with more serious problems to overcome them if possible.

In the ideal system, the department adopts explicit standards, requires compliance , monitors performance, and responds to deficiencies. The department chair reviews performance data with each physician annually, and together they work out a plan for improvement as needed. In some cases, this may require external testing and remediation.

A similar oversight process should be required of larger medical groups and employed physicians. The remaining small number of physicians in solo or small practice might then be required by licensing authorities to take advantage of some mechanism like PACE or CPEP in order to maintain licensure.

Fortunately, as we have seen, the ABMS and specialty boards have worked hard in recent years to develop national standards of competence and behavior and to integrate them into the process of continuing certification. Closer coordination of this oversight with local review and response would lead to greater accountability and improved performance.

Nonregulatory Approaches to Improving Competence

Independent of the impressive changes to improve accountability by the establishment organizations described above, a number of independent voluntary initiatives have taken place over the years to improve the process of physician assessment and improvement. Several deserve special mention.

National Surgical Quality Improvement Program

In 1986, responding to a series of newspaper articles about poor care in Veterans Health Administration (VHA) hospitals, Congress mandated that VHA report risk-adjusted surgical outcomes annually and compare them to national averages. There was a problem, though: there were no known national averages and no known risk adjustment models!

But the VA was uniquely suited to develop them for its population. The VHA is the largest health-care provider in the USA, serving several million veterans and performing surgery in 128 of its 159 Veterans Administration Medical Centers (VAMCs). At the behest of their surgical leadership , a research group at the Brockton/West Roxbury VA Medical Center in Massachusetts led by Shukri Khuri , Chief of Surgery, and Jennifer Daley , an experienced quality-of-care researcher, carried out the National VA Surgical Risk Study from 1991 to 1993. Using data collected from 117,000 major operations in 44 VAMCs, they developed risk adjustment models for 30-day mortality and morbidity rates for noncardiac surgery [39].

They then turned their attention to measurement of surgical outcomes. Surgery is uniquely suitable for measurement of outcomes since there is a clearly defined expected outcome for every operation. Using this validated model for risk adjustment, outcomes could now be measured with some confidence in their validity.

In 1994 the VHA established the National VA Surgical Quality Improvement Program (NSQIP) , a reporting and managerial structure for the continuous monitoring and enhancement of the quality of surgical care, under an executive committee led by Khuri and Daley [40].

Surgical clinical nurse reviewers (SCNRs) were trained in the accurate collection and timely transmission of risk adjustment data, consisting of 45 presurgical variables, 17 surgical variables, and 33 outcomes. Logistic regression analysis was used to calculate a predicted probability of 30-day mortality and complications. Risk-adjusted observed versus expected (O/E) outcome ratios were calculated for all types of procedures at the surgical service of each VAMC and overall.

Feedback of these procedure-specific O/E ratios is provided annually to the chief of surgery, director, and chief of staff of each VAMC, and the CMO of each Veterans Integrated Service Networks (VISN) , as well as results for all participating hospitals, by code. Hospital leaders know only the code for their hospital.

The executive committee produces an annual assessment of high and low outliers and communicates levels of concern about high outlier status to hospital and VISN , as well as praise and rewards to low outliers. Persistent high outliers are subject to internal and external reviews.

NSQIP also develops and disseminates self-assessment tools to providers and managers and, at the request of a VAMC, organizes consultative site visits to assess data quality and performance. NSQIP provides management (directors and CMOs of VISN ) with advice and expertise in conducting external reviews and site visits and disseminates best practices reported by low O/E hospitals.

The first assessment of results showed that during the period from 1991 to 1997 30-day mortality decreased from 3.1 to 2.8 and morbidity decreased from 17.4 to 10.3. By 2006, postoperative mortality had dropped by 47% and morbidity rates by 43% [41].

The program was well-accepted by the chiefs of surgery who valued the feedback and learned to find and improve deficiencies. From the beginning, NSQIP has been about quality improvement, not judgment. The emphasis is on systems not providers. No individual provider-specific data is transmitted to the central data base.

Several aspects of NSQIP accounted for its success. Most important was the fact that VHA had in place a universal computerized record system, VISTA , that made clinical and laboratory data available for risk analysis. It also had access to the operating room log in every VAMC, so all procedures were automatically and reliably identified.

Second, for data entry it relied exclusively on trained surgical clinical nurse reviewers (SCNRs) who were experienced in practice, data collection, and quality assurance. This gave high levels of credibility, reliability, and validity to the data. Third, inclusion of surgical leaders from the field in the design of the program and oversight led to support by VAMC senior surgeons, administrators, VISN directors, and CMOs.

The private sector took notice. Why not use NSQIP for non-VA surgical departments? Within months of the first report, in 1999, a pilot program was begun in three academic surgical centers, University of Michigan, Emory University, and the University of Kentucky, to determine if the risk adjustment models would work for the more heterogeneous private sector patient populations. They did. Comparison of findings in 2747 patients at these centers with contemporary results in 41,360 patients in the VHA showed no differences in risk-adjusted mortality between the non-VA and VA cohorts [42].

Following this success, the American College of Surgeons (ACS) in 2001 sponsored a pilot program funded by AHRQ in18 private sector hospitals that showed that NSQIP also led to reduced morbidity and mortality in private sector hospitals. In 2004, ACS began enrolling additional private sector hospitals into ACS NSQIP . Within a year, 41 hospitals had joined. By 2018, participants included 568 hospitals in the USA, 96 in Canada, and 38 overseas. Nine of the top 10 hospitals ranked as America’s Best Hospitals by U.S. News & World Report in 2018 participated in ACS NSQIP [41].

Meanwhile, NSQIP continues to work on improving. More specialty variables were incorporated; additional outcome measurements, such as functional status, quality of life, and patient satisfaction, were developed and incorporated; and structure and process measures were added [43].

Analysis of Patient Complaints

In the early 1990s, Gerald Hickson , Associate Dean for Clinical Affairs at Vanderbilt University Medical Center , and his colleagues found that analysis of written complaints by patients to the hospital was a useful tool for identifying physicians with interpersonal problems. About 2/3 of complaints were about a hospital or practice service or system issue; 1/3 were about a named physician.

While patients often complain about their doctors, it is unusual for them to make a formal complaint in writing [44]; most physicians receive none or only one or two over their entire professional career. But some have more. Hickson wondered if there was a relationship between the number of complaints and the likelihood of the physician being sued. (“Claims” in risk management parlance.)

Indeed, there was. In a six-year period, he found no claims for 81% of doctors who had only one or no complaints. The majority of those with 2–6 complaints also had no claims. But physicians with 4 or more complaints over this period were 16 times more likely to have 2 or more claims than physicians with no complaints. Those with 25 claims or more had a 95% chance of being sued [45].

Hickson realized that patient reports could serve as the basis of an “early warning system” to more rapidly identify and engage with physicians before harm occurred and suits began to accumulate. They could then be helped to overcome their deficiencies. He developed a tiered intervention program, the Promoting Professionalism Pyramid, that defined a process that started with a conversation with a colleague and escalated if needed to formal evaluation and required behavioral change.

Following the first complaint, a colleague would have a “cup of coffee conversation” in which the complaint is shared with the physician in a nonjudgmental way and they are asked to reflect on the event. Often the physician has not recognized the bad behavior and justifies it because of the situation. The colleague makes no judgment, merely delivering the news. But for many, that is all that is necessary: their behavior changes.

At the second level, when there have been additional reports that suggest a pattern of inappropriate behavior, an awareness intervention is called for. A respected colleague presents the data to the individual showing how their complaint history compares to that of their peers and gives them the opportunity to respond. Again, in most cases this is all that is needed to lead the physician to change behavior.

For those that do not respond to the awareness intervention, the response moves to the next level. The department chair steps in and makes it clear that the individual must change their behavior. Chairs are trained to work with the physician to define an improvement plan that may range from coaching and counseling to formal outside evaluation and retraining.

If the physician is unwilling to undergo assessment and take responsibility for improving, or if these measures fail, then disciplinary action is required, which can include revoking admitting privileges or reporting to the state medical board [46]. Fewer than 1% fall into this category.

Hickson also developed a comprehensive program at Vanderbilt to reduce disruptive behavior by teaching interpersonal skills and professionalism at all levels: medical students, residents, and physicians. Physician leaders also receive skills training for conducting interventions [46].

He also developed a Comprehensive Assessment Program for Professionals to provide medical and psychological evaluation and treatment planning. Group classes were developed for disruptive behavior , prescribing problems and crossing sexual boundaries [47].

National Alliance for Physician Competence

This was one of the most unusual and exciting ventures I was ever part of, both for its goal, which was to set standards for good medical practice , and for those who participated, who were leaders of the national groups that could make it happen—in education, regulation, professional societies, and others. It was also one of the most frustrating.

The Alliance was organized by James Thompson , President and CEO of the FSMB , an ENT physician and former Dean at Wake Forest School of Medicine. Moved by the IOM reports, To Err is Human and Crossing the Quality Chasm , Thompson recognized when he took over FSMB that state medical licensing boards needed better methods for determining physician competence , both for licensing and for disciplinary actions. ACGME and ABMS had defined the six competencies, and the ABMS was moving to maintenance of certification . Shouldn’t state boards do likewise?

Thompson encouraged the Federation to issue a statement on the need for maintenance of licensure, but much more was needed to make it a reality. He conferred with experts he knew as a former Dean: Donald Melnick , President and CEO of the National Board of Medical Examiners (NBME); James Hallock , CEO of the Educational Commission for Foreign Medical Graduates (ECFMG); and David Leach, CEO of ACGME . They supported his effort but felt that a comprehensive strategy linking licensure to education and specialty certification was needed. The time had come to begin a dialogue about the future of physician education and self-regulation.

On March 24, 2005, they brought together more than 60 leaders and representatives from organized medicine, academic medicine , hospitals, regulatory agencies, the insurance industry, accrediting organizations, payers, and the public in Fort Worth, Texas, for the first “Summit” on Physician Accountability for Physician Competence (PA4PC) (Table 20.1). The goals were to determine (1) how to define a competent physician, (2) how to measure competency, and (3) how medical organizations would assure the public that physicians are maintaining competence throughout the lifetime of their practice [26].

Table 20.1 Institutional members of the National Alliance for Physician Competence

With help from Innovation Labs, and financial support from the NBME , the meeting explored the context within which physicians would be expected to demonstrate accountability in the year 2020. What should the system look like? The group was energized and quickly found common ground on the big issues.

In subsequent meetings, other relevant stakeholders , such as patients and content experts like myself, were added to the group. Over the next 2 years, in a series of semi-annual meetings PA4PC drafted detailed definitions of competence and the content for a document, Good Medical Practice , that described the behaviors and values one should expect of a competent physician. A task force worked on simplifying physicians’ access to credentialing information for multiple purposes such as licensing and board certification . The group renamed itself the National Alliance on Physician Competence .

The good practice document was our central focus. It was based on the work of the General Medical Council in the UK but reframed in terms of the six domains of competency defined by ABMS and ACGME . There were great debates about terminology. Should the document say doctors “should” do such and such or “must” do it? Ultimately, both were rejected. This would be a statement of who we are and what we do—who we aspire to be—not because it is required, but because of our values and commitment to our professionalism . We would use simple declarative sentences: “We respect each patient’s dignity and individuality”; “We promptly modify our practice to incorporate evidence-based care”; “We apologize promptly to a patient when an error has occurred.”

As it came into focus, we realized that the document should begin with The Patient’s Perspective: a comprehensive statement of what patients have a right to expect from doctors regarding medical knowledge and skills, communication and interpersonal skills, shared decision-making, access and availability, and ethical integrity. This is the lens through which we see our role, our duty. The purpose of competence is to provide optimal patient care.

We finished the first draft, Version 0.1, of Good Medical Practice – USA, on August 15, 2007. It described the behaviors expected of all doctors who are permitted to practice medicine. The Patient’s Perspective was followed by Duties of the Doctor consisting of one chapter for each of the six domains of competency. It was incredibly detailed, 200 statements in all, providing guidance on every aspect of practice, especially those that are difficult, such as knowing one’s limits, giving bad news, dealing with problem colleagues, etc. Simple declarative statements of what good doctors do.

We called on medical educators and regulators to incorporate these principles in everything they do and challenged all physicians to take personal responsibility for making it happen.

The Alliance grappled with the relationship between maintenance of licensure and maintenance of certification and how to engage the practicing community and the public in the effort. To facilitate the licensing and certification processes, it developed a standardized, comprehensive “Trusted Agent/Portfolio System” that would enable physicians to retrieve all needed credentials from a single source.

The Alliance examined how a “continuum of competence ” could be established: a system that would start in medical school and continue through residency programs, licensure, specialty certification, hospital credentialing and privileging, and the accreditation of institutions. How would the use of Good Medical Practice and the Trusted Agent/Portfolio System impact long-term maintenance of competency throughout a physician’s career?

The last meeting of the National Alliance for Physician Competence Summit was held on July 7–9, 2008. The goal was to prepare to go public. Small groups synthesized and polished models to shift the paradigm for competence . These were then rolled into a single model of 14 components. Others focused on finalizing the renamed Guide to Good Medical Practice . Plans were made to “go live” with it in September, when Alliance participants would distribute the document. A draft Alliance website was created. A revised Alliance Participant Agreement was approved.

(a) Jennifer Daley, (b) Jerry Hickson, and (c) Jim Thompson. (All rights reserved)

Then it all fell apart. From the beginning, the AMA had been a reluctant participant. It traditionally opposed anyone telling doctors how to practice and was against giving state boards more power. It declared opposition to the Guide even before it was written, maintaining that medicine is full of gray areas that are too difficult to measure. It opposed the concept of maintenance of certification .

At the last two meetings, it sought to undermine the process of the meetings by sending a large number of delegates who raised objections in all the working sessions. Although most of these were rejected by the majority, they disrupted the collaborative process.

Finally, at the last session the AMA withdrew its support. And, much to my surprise, despite the fact that it was the convener, so did the FSMB . It was proving to be too much for the individual state licensing boards . They were reluctant to take on this level of responsibility, and they saw no way to obtain the resources that would be required. The ABMS did not fight for it. It was having enough trouble figuring out how to implement the six competencies . The Alliance was finished. Our “brief shining moment,” our Camelot, was over.

The Coalition for Physician Accountability

But Don Melnick, Jim Hallock, and Darrell Kirsch , CEO of the AAMC, were not going to let the concept die. The next year, they formed the Coalition for Physician Accountability to continue the discussion and further the cause. Its membership includes the stakeholders who have direct responsibility for assessment, accreditation , licensure, and certification along the continuum of medical education and practice.

The Coalition provides a forum for dialogue about ways to “promote professional accountability by improving the quality, efficiency, and continuity of the education, training and assessment of physicians” [48]. The Coalition meets twice yearly to analyze critical issues related to the regulation of physician education and practice and to develop consensus on actions to address them.

It functions through its member’s endorsement of consensus statements about a diverse group of topics: regulation, innovation in medical school curricula, graduate medical education accreditation , interprofessional education, medical student and physician burnout , use of health information technology , opioid epidemic mitigation, interstate licensure, and a framework for professional competence and lifelong learning. It developed a consensus letter that was sent to Congress regarding maintaining Medicare support of GME, and it sent a letter to the National Coordinator outlining the commitment of Coalition members to promoting the use of health information technology .

Conclusion

Ensuring physician competence is a complex and difficult business. Despite the huge amount of work done by many diverse parties, it is still very much a work in progress. Oversight bodies, the state licensing boards and, especially, the specialty boards, have made substantial improvements in how they function, but the results still fall far short of achieving their objectives.

Why? Why doesn’t the system work better? Why don’t ABMS and the specialty boards make it work better and require, audit, and enforce adherence to the impressive and innovative processes they developed for maintenance of certification based on the six competencies ?

Undoubtedly, there are many reasons, but I suggest that the fundamental reason, the “root cause” if you will, is that it is contrary to human nature for any group to police itself. We have not asked that of the other major industries where safety is critical: aviation and nuclear power. They are closely regulated by specific government agencies.

Do we need a federal agency to regulate quality and safety in health care? I have long believed we do [49]. The federal agencies regulating aviation and nuclear power are good models. The government exercises strict oversight of compliance with its rules, but those rules were developed in collaboration with the industry. Participation leads to buy-in and higher likelihood of compliance . (Recall the New York Cardiac Advisory Committee .) An agency developing regulations for doctors should collaborate with the specialty boards and state boards as well as representatives from professional societies and health-care organizations.

Hospitals should be held accountable for their physicians’ performance. They should participate in developing regulations that ensure they are accountable to the public, such as required reporting of adverse events . The Joint Commission should be a partner in this process and play an important role by carrying out the necessary annual or semi-annual audits.

We have made tremendous progress in recent years in defining competence and measuring it. What was formerly implicit and casual can now be defined in an explicit and formal manner. We now know how to enable physicians to realize their full potential and by so doing immensely improve the quality and safety of patient care. The time has come to make it happen.