Approaching the Issue of the Aging Physician Population

As our society's population ages, so do physicians. This has led to increasing numbers of older physicians in practice and new discussions in the health care community about physician competence and the maintenance of skills in older age. Among the topics of growing interest among regulators and other policymakers in this new environment: Should older physicians be the subject of some form of age-based competency screening?

In a 2011 conference titled “Practicing Medicine Longer: The impact of aging on physician clinical performance and quality of care,” the Coalition for Physician Enhancement (CPE) joined with the University of California, San Diego, Physician Assessment and Clinical Education (PACE) Program to review the current body of knowledge on aging and physician competence, and ascertain views from attendees on a variety of topics — including the question of age-based screening.

International experts in medicine, law and public policy gathered to discuss the multi-faceted impact of aging on physician clinical performance and quality of care during the two-day event. Attendees were surveyed in order to understand their opinions about age-based physician screening — whether or not it should be implemented, and if so, what age, interval and screening methods would be optimal.

The profession of medicine holds itself to high ideals of caring and competency, the first tenet being primum non nocere, “First, do no harm.” While other high-stakes professions—such as aviation—have proactive policies to ensure safety and quality, medicine has relied on self-regulation, in which incident-triggered evaluation is the norm. Policies for addressing the potential for health or age-related impairment of clinical practice are rare. The purpose of this paper is to share the salient data presented at the conference, present the results of the conference survey regarding age-based screening, and propose a construct for how an age-based screening agenda could be carried forward.

Physician demographic trends: AMA Physician Masterfile data demonstrates an increase in the aging physician population.1 In 1985, the number of active physicians in practice was 476,683, with 9.4 percent age 65 or older. In 2005, physicians in active practice numbered 672,531, with 11.7 percent age 65 or older and a mean age of 50. In 2011, physicians in active practice numbered 697,340, with 15.1 percent age 65 or older and with a mean age of 52.5. In addition to the concern about aging physicians, many sources predict an overall shortage of physicians in the future,2 particularly in primary care, to provide medical services for our aging population and the additional patients who will receive care with the implementation of the Affordable Care Act. Moreover, from the physician's perspective, the recent economic slow-down in the United States has impacted the retirement plans of many aging physicians, leading them to feel the need to practice medicine longer.

Healthy aging and the benefits of aging: Successful aging is an area of current investigation that includes the concepts of wisdom, social cognition and optimal decision making —areas beyond general physical and mental health.3,4 Basic neuroscience research provides evidence that aging brains remain plastic, and compensation, neurogenesis and synaptogenesis can continue to occur in enriched environments.5 From a large population-based study, predictors of successful aging include optimism, resilience, self-efficacy, low perceived stress, low level of depressive symptoms, exercise at any level, writing, computer use, and regular socialization.3 While there are physical and cognitive declines associated with aging, positive psychological attributes such as optimism, resilience, compassion, and wisdom do not decline, but stay stable or even increase with age.

Physicians are a unique group of professionals with a complex combination of assets, including intelligence, drive, conscientiousness, stamina; and liabilities, including stress, limited time, lack of diversified activities, and lack of planning for retirement activities/life. While there is some evidence of neurocognitive deficits in aging physicians with competence problems,6 data also demonstrates that older physicians' ability to cope with stress in improved ways increases and that adaptation occurs with the development of new skills.7,8 

Neuropsychological changes in aging physicians: A body of literature links aging to cognitive changes, but the association between these potential changes and occupational performance in physicians is a more challenging one to demonstrate. In one systematic review called “Relationship between Clinical Experience and Quality of Healthcare”9 which reviewed 62 published studies, more than half of the studies found declining clinical performance outcomes with increased age; only one study showed improvements in all outcomes with greater age. There is a well-supported theory that many older physicians draw from prior experience and rely more heavily on non-analytic, crystallized cognition. An example of this is the use of pattern recognition in a clinical presentation, which may be accurate most of the time, but can lead to early and incorrect cognitive closure some of the time. Fluid cognitive abilities, which include novel, spatial manipulation and mental speed, tend to peak in the third decade and decline more steeply in the seventh and eighth decades. Data suggests that older physicians tend to do less well when dealing with novel, complex patient situations: The tendency to rely on crystallized memory and pattern recognition led to incorrect diagnosis in 40 percent of complex presentations in one study, for example.10 In a neuropsychological evaluation study of physicians being disciplined by the Medical Board of California, Perry and Crean found relative deficits on tests of sequential processing, attention, logical analysis, eye-hand coordination, and verbal and non-verbal learning.11 These findings revealed that in this cohort of physicians, there is lower than expected performance on tests of intellectual and neuropsychological functioning. It was proposed that doctors can accommodate the cognitive changes associated with aging and mitigate against unsafe practice by shifting their work away from procedural work, allocating more time to each patient, using memory aids, seeking advice from colleagues, and seeking second opinions.12 

Factors (including age) that can impact physician clinical performance: When considering age as a possible factor affecting clinical performance, it is imperative to consider the other known risk factors, which can negatively impact clinical competence. These include:

• Poor performance in medical school

• Solo practice

• Lack of hospital privileges

• Lack of ABMS Board Certification

• Practicing outside the scope of training

• High clinical volume

• Fatigue/stress/burnout

• Health issues, both physical and mental

• Organizational, systems problems13 

In contrast, in a Rand study,14 factors which were associated with better clinical outcomes and performance were identified and included female gender, board certification in area of practice, and having attended a U.S. medical school.

In a study of 460,000 patients undergoing one of eight types of major operations, Waljee et al. found mortality outcomes were increased for older surgeons (>60 years) for three of the eight types of operations: pancreatectomy, coronary artery bypass graft (CABG), and carotid surgery. Additionally, older surgeons with the lowest volume of cases had the highest mortality rates.15 In another study, physicians with greater years in practice length were more likely to provide inpatient care with longer length-of-stay and higher risk of patient mortality.16 

Based on the collective data regarding aging and physician performance, the following suggestions were proposed at the conference:

• Aging results in a wide spectrum of physiological changes which may affect clinical competence. Amongst the most important are the reductions in dexterity and visual-spatial acuity, short-term memory, problem-solving, and ability to adopt new ideas and to re-examine old ideas.

• Aging is but one of several risk factors which may impact clinical competence: The degree and rate of decrement caused by aging do not occur in a linear fashion and the impact upon the clinical competence and performance of any one physician is highly variable.

• Programs exist which can and do assess medical knowledge, historical aspects of patient care and simulations of patient care and interpersonal skills and communication, but generally these assessments take place after the occurrence of an untoward outcome. Physicians are not regularly and routinely assessed.

• Assessment could include evaluation of mental and physical health, review of actual performance of clinical care — either diagnostic or procedural, documentation that learning and behavior change as a result of participation in CME has taken place, and review of quality improvement efforts.

• We should not establish mandatory retirement for physicians based on age alone for many reasons, including the inability to definitively conclude that age, in and of itself, is a risk factor for incompetence or dyscompetence. Establishing mandatory retirement based on age alone would further negatively impact the physician shortage in the United States and would lead to the loss of the wisdom and experience of many capable physicians.

Quebec: The age distribution of physicians in Quebec has also changed rapidly in the past few years, with physicians over 65 years of age reaching new high levels and increasing numbers of physicians now practicing in Quebec past 70 to 75 years of age. In a peer-reviewed study in Quebec from 2001–2010, 1,618 physicians were contacted two to three months in advance of an onsite visit in which their practice would be reviewed.17 Based on a thorough review of documentation, prescribing habits, data from the province's billing administration and clinical practice outcomes, each physician was placed in one of five categories:

• Level 0: No action, satisfaction letter

• Level 1: Recommendations

• Level 2: Recommendations and control visit follow-up

• Level 3: Refresher course or retraining or limitation (retirement was a frequent option with this result)

• Level 4: Cancellation of licensure

Physicians over the age of 70 had three times higher rate of cancellation (31 percent, N=516) compared to the group less than 70 years old (10 percent, N=1407). Of note, the group of physicians age 65 to 69 showed only slightly higher rate of cancellation (13 percent, N=338) but had nearly double the rate of Level 3 recommendation than for the physician group less than 65 years old (18 percent vs. 10 percent). Continuing Professional Development (CPD) was reviewed as an independent factor: poor quality or insufficient CPD was strongly associated with poorer outcomes in the peer review study. Additionally, physicians aged 65 to 97 had a higher percentage of inadequate CPD than the younger physician groups (30 percent vs. 10 percent). Physicians with recommendations were provided clinical training programs or tutorials, and the physicians in the over-70 group had less favorable outcomes even with remediation (22 percent success vs. 45 percent for the 65-to-69 year-old group).

Ontario: The College of Physicians and Surgeons of Ontario (CPSO) Peer Assessment Program offers a rich data set of outcomes based on peer review. The program was initiated in 1977 as a way to identify incompetent physicians. From 1977 to 1980, pilot data was collected; then, in the following five-year period from 1981 to 1986, 920 physicians were formally assessed.18 The assessment included four components: 1. Practice questionnaire, 2. A peer site visit to learn about the practice and inspect the physical office, 3. A chart review of 20–30 randomly selected charts, and 4. A chart-stimulated discussion with the physician. Of the 920 physicians in the initial five-year period, 11 percent had grossly deficient records or unsatisfactory care or both, with 10 percent having medical record-keeping deficiencies and five percent with actual clinical performance and patient care problems. This breakdown is consistent with subsequent outcomes from the CPSO program.

By age, the Peer Assessment Program in Ontario found that 22 percent of physicians in the group over 75 years old had gross deficiencies in their practice, 16 percent in the 50-to-74 year-old group had deficiencies, and nine percent of physicians under the age of 49 had deficiencies. When the age categories were split differently, 55-and-older physicians had poorer performance than physicians under age 55, and surprisingly, there was close to no difference in physicians' performance outcomes between the 55-to-69 year-old group and the group over 70 years old.18 

Another focus of Ontario findings was on the relationship between record keeping and patient care. The findings show that good records are associated almost entirely with good care, whereas the existence of poor record keeping can go either way regarding quality of care.

During the 2011 CPE/PACE conference, a special survey was conducted to gather data on the topic of age-based competency screening for physicians. The results of that survey are summarized here.

The authors obtained Institutional Review Board approval to obtain human subjects data on the opinions of the conference attendees. Two workshops were offered to all course participants: “Individual versus societal right: Point-Counterpoint” and “Creating an assessment battery for the aging physician.” The workshops were offered after the didactic portion of the program. Participants answered questions via an audience response system (ARS) (Turning Technologies, LLP, 2008 Turning Point software) during the course of the workshop. Data was collated and analyzed using the same software.

The total number of attendees for the conference was 110. Of the total, 35 percent were physicians, seven percent were PhDs, 25 percent were Juris Doctors (both judges and attorneys) and 33 percent were self-described as “other.” Thirty-eight percent had a college or university affiliation, 48 percent had a state, national or other government affiliation, six percent were from private law firms and five percent were from other organizations. Forty-nine percent of the participants were from California, 15 percent from another state in the United States and 36 percent from Canada.

A total of 71 individuals responded to the ARS questions. Not all participants answered all questions. Demographic information, including profession, professional affiliation, gender, age, location and specialty for medical practitioners was obtained. Fifty-two percent of responders were physicians or other healthcare professionals, 13 percent were state medical board members, 26 percent were from the legal profession (judge, attorney), six percent were research scientists, one percent were administrator and one percent self-identified as “other.” Of practitioners, 46 percent were primary care and 54 percent were specialists (see Table 1).

Participants were asked, “Do you believe that there is a need for age-based physician screening?” (Answers were multiple choice: yes, no or maybe.) Sixty-two percent of individuals responded “yes,” 18 percent responded “maybe” and 20 percent responded “no” (Figure 1). When the responses were segregated by the age of the participant (Figure 2), the “yes” response was much greater for those respondents age 60 and younger versus respondents over 60 years old. Regarding participant gender (Figure 3), the number of “yes” responses for men and women were nearly identical. For responses segregated by physician vs. non-physician (Figure 4), responses were nearly equal as well. For responses segregated by United States vs. Canadian resident (Figure 5), Canadians were more likely to respond “yes” than Americans (70 percent vs. 58 percent). Finally, related to physician specialty (Figure 6), there was a slightly higher percentage of “yes” responses from primary care physicians vs. specialists (67 percent vs. 62 percent).

Regarding the components of a possible assessment battery for aging physicians, participants felt strongly about including assessments of physical and mental health and a cognitive screen. Forty-five percent of respondents believed that age-based screening should begin at age 70, 27 percent at age 65, 16 percent at age 60, eight percent at age 75 and one percent each at 80 and 85 years old.

The majority of respondents agreed that there is a need for age-based screening of physicians. Notably, there were no significant differences across gender and profession, and only a modest difference between primary care physicians and specialists — with primary care physicians more strongly endorsing age-based screening. In the United States it is estimated that 35 percent of physicians are engaged in primary care and 65 percent practice in other specialties, which is similar to the physician cohort of participants in the survey.

The greatest difference in opinion was related to physician-participants age 61 or older, who were less in favor of screening for aging physicians than their younger colleagues. While specific reasons were not queried, the temporal relationship of older physicians' age to the proposed screening age may have affected their responses. Other notable results from the workshop surveys were that Canadians were more in favor of screening the aging physician than their U.S. counterparts. Perhaps the fact that age-based screening already exists in some Canadian provinces (Quebec and Ontario) influenced their responses. Additionally, responses from participants were gathered after data were presented in the morning didactic sessions, which may have influenced responses either negatively or positively. Another limitation to the findings of the survey was the predominance of participants being from California compared with a broader representation from other regions of the U.S. On the other hand, a strength of the cohort was its bi-national representation.

The information presented during the program, while not definitive, provides compelling evidence that aging is one of several independent risk factors for substandard clinical performance. It is incorrect to state that aging has only negative outcomes with regard to physician performance. Clinical ability does improve with experience; however, the neuropsychological assessment and clinical outcomes literature indicates that with aging, certain abilities begin to diminish. High patient-complexity, performing certain major operations and multitasking confer higher patient risk with older physicians.

The authors believe that there should not be a required retirement age for physicians. Undesirable outcomes from a required retirement age are negative impacts to the physician workforce numbers, poorer access to care and unnecessary loss of experienced and productive physicians. The evidence, however, does point to a need for evaluation of mental and physical health at appropriate junctures throughout a physician's lifecycle. A call for a process beyond self-regulation is warranted. In one study, 96 percent of physicians agreed that impairment should be reported, but only 55 percent who had encountered such colleagues actually reported their concern to anyone.19 

Screening for medical conditions (e.g., cancer, diabetes, hyperlipidemia) in the United States is a well-established practice. An analogous model for screening physician competence based on identified risk factors, such as age, should be considered. Like any screening test, the benefits, costs and risks must be established, but with minimal to no data existing for physician screening, professionalism should guide us initially. Preliminary data from Canada may help to craft policy in the United States and as some U.S. hospitals have instituted age-based screening, an evidence-based answer may be achievable in the near future.

In conclusion, the authors recommend that national discourse continue with regard to age-based screening. The authors recommend studying screening commencing at age 70 with an assessment battery consisting of evaluations of physical health (including vision, hearing and dexterity for proceduralists), mental health and a cognitive screen. In addition, other modalities such as peer review, Ongoing Professional Practice Evaluation (OPPE), Focused Professional Practice Evaluation (FPPE)20 and Maintenance of Certification for ABMS-certified physicians (or a similar process for non-ABMS-certified physicians) should be considered.

Physicians must take the lead in addressing this important issue. Medical professionalism calls for self-evaluation. In reviewing the current body of knowledge about the impact of aging on physician practice and reflecting upon the rationale for screening for conditions that are associated with adverse health outcomes, we are led to the conclusion that age-based screening for competency is an important safety measure. We applaud the hospitals and medical systems that have initiated age-based screening. The medical profession should act now, lest others dictate the direction of this important issue. We believe it should embrace a model that encourages physicians to think carefully about the timing of their retirement from practice — doing so in a way that supports a graceful, non-traumatic exit at the right time.

The Coalition for Physician Enhancement (CPE) is a consortium of professionals with expertise in quality assurance, medical education, and the assessment, licensing, and accreditation of referred physicians seeking higher levels of performance in patient care with members from the United States and Canada.

The University of California, San Diego, Physician Assessment and Clinical Education (PACE) Program is a physician assessment and remediation program that is dedicated to the education of physicians and other health care professionals; the detection, evaluation, and remediation of deficiencies in medical practice; and assisting the medical profession in its quest to deliver the highest quality of health care to the citizens of the United States.

Presenters at the conference included:

• Peter Boal

  Associate Director of the University of California, San Diego, PACE Program

• Dilip Jeste, MD

  Distinguished Professor of Psychiatry and Neurosciences at University of California, San Diego; Director of the Stein Institute on Aging

• William Perry, PhD

  Professor and Associate Director of the Neuropsychiatry and Behavioral Medicine Service at University of California, San Diego

• Stephen Miller, MD

  Clinical Professor of Plastic Surgery and Faculty in the PACE Program

• André Jacques, MD

  Director of the Practice Enhancement Division, College of Physicians of Quebec

• William McCauley, MD

  Past-President of the Coalition for Physician Enhancement; Medical Advisor, College of Physicians and Surgeons of Ontario

• David E.J. Bazzo, MD

  Clinical Professor, Department of Family & Preventive Medicine, University of California, San Diego; Course Director, U.C. San Diego PACE Program; Program Committee Chair, CPE

• Jill Silverman, MSPH

  President and CEO of the Institute for Medical Quality

• Elizabeth Wenghofer, BSc, MSc, PhD

  Interim Research Director, Centre for Rural and Northern Health Research Laurentian University; Chair, CPE Research Committee

• William Norcross, MD

  Clinical Professor, Department of Family and Preventive Medicine, University of California, San Diego; Director, U.C. San Diego PACE Program

The authors would also like to acknowledge the participation of the conference attendees and their willingness to share their thoughts and opinions on this important topic.