Accepting the challenges of ageing and retirement in ourselves: The need for psychiatrists to adopt a consensus approach

Introduction

The Australasian psychiatric workforce is ageing. Data from the Royal Australian and New Zealand College of Psychiatrists (RANZCP) shows that 42.2% of approximately 2950 Fellows practising in Australia are aged 55 years and older, including 17.7% aged 65 years and older, suggesting that a number of psychiatrists are working into their 70s and 80s. The equivalent proportions for the 320 RANZCP Fellows working in New Zealand are 42% and 14.4%, respectively.

There are a number of implications of this greying of Australasian psychiatry, which has yet to be fully appreciated by the profession. In contrast, the College of Intensive Care Medicine of Australia and New Zealand recently initiated a process to develop guidelines for ageing intensivists and retirement (Skowronski and Peisah, 2012). The process was geared as much to promote workforce retention, encourage career management of late-career specialists and workplace adaptations, as it was to provide guidelines regarding impairment.

The need to address such issues is equally relevant and pressing for an ageing psychiatric workforce. While we do argue for a similar set of guidelines to be considered by the RANZCP, it is recognised that the specific cognitive and other skills required for the practice of psychiatry would vary from those applied by procedural specialists.

The benefits afforded by late-career psychiatrists

There are three reasons why each psychiatrist should continue working for as long as practicable.

The first is workforce planning and service provision. The current Australian ratio of approximately one psychiatrist per 9000 people is favourable when compared with the World Health Organization recommendation (Mental Health Workforce Advisory Committee, 2008). This bare statistic, however, hides a number of concerns, namely: a maldistribution of psychiatrists between certain states and urban/rural areas; the increasing proportion of female trainees, with future implications for the full-time equivalent workforce; and the ageing of RANZCP Fellows, with less than 30% of Fellows practising in Australia aged under 45 years.

The reaction of responsible bodies to the need to ensure a continued supply of specialists has been to focus on making psychiatry a more attractive career option for medical students and junior medical officers, and to increase the support afforded to trainees (Mental Health Workforce Advisory Committee, 2008). Yet, there is also a need to support late-career psychiatrists to continue working, as delayed retirement is another way of maintaining the workforce.

The second reason to encourage retention of psychiatrists is that late-career psychiatrists provide a valuable service to the profession. RANZCP Fellows, in particular younger and early-career psychiatrists, believe that senior psychiatrists have ‘wisdom’ to offer junior colleagues, in particular mentoring and supervision, life experience and acceptance of personal limits/fallibility (Draper et al., 1999a).

The third reason is that each individual medical practitioner has the right to continue working, as long as the primary ethical principle of doing no harm to the patient, profession and wider society is observed. Psychiatrists, of all medical specialists, should be most aware of the benefits of work to psychological well-being.

Indeed, there may be advantages to being an older practitioner. For instance, burnout seems lower with increasing age and years in practice (Peisah et al., 2009b). Older doctors reported greater personal accomplishment, the development of protective defences in therapeutic relationships, and increased confidence and familiarity with their professional role via the accumulation of experience. A survey of RANZCP Fellows reported that the benefits of ageing were an improved ability to communicate with patients of one’s own age and life stage, and a more balanced life perspective (Draper et al., 1999b). A specific sample of Fellows aged 55 years and older reported little in the way of problems due to age – nearly one-third reported there were no professional drawbacks to age, and only 10% and 12%, respectively, reported having difficulty maintaining their knowledge base or having poor memory (Draper et al., 1997).

Age, illness and impairment

The other side of the coin is the extent to which the capacity to practice medicine is affected by the cognitive, physical and psychological correlates of ageing.

Impairment has a specific, statutory definition in the Australian Health Practitioner Regulation National Law 2009. A medical practitioner is impaired if a physical or mental condition or disorder, including substance abuse or dependence, detrimentally affects, or may potentially detrimentally affect, the capacity to practice medicine. This is similar to the New Zealand definition of impairment which states that a doctor is not fit to practice if, due to a physical or mental condition including one caused by alcohol or drug abuse, they are not able to perform the functions required for the practice of medicine (Health Practitioners Competence Assurance Act 2003).

There is conflicting information about the number and proportion of impaired late-career practitioners. There was no increase in the number aged over 60 years under the care of the Health Committee of the then Medical Practitioners’ Board of Victoria (Adler and Constantinou, 2008), although the most recent UK data has shown that there are more complaints about older practitioners (General Medical Council, 2012). The accuracy of these data is limited by its reliance on formal notification, leading to unrepresentative sampling of impaired doctors.

Half of all notifications of older doctors to the Medical Practitioners’ Board of Victoria in the first study were for cognitive impairment. Similarly, a study of older doctors who were the subject of notification to the equivalent New South Wales (NSW) body reported that 54% suffered cognitive impairment, 29% substance abuse and 22% depression (Peisah and Wilhelm, 2007). Of particular concern, was that during the study period (2000–2006), five doctors practising with frank dementia were identified.

Age and performance

A systematic review of 62 studies reported that 73% of the studies showed an inverse association between performance for all or some of the outcomes assessed and years in practice (Choudhry et al., 2005). The authors concluded that doctors in practice for longer and older practitioners possessed less factual information, were less likely to adhere to appropriate standards of care and had poorer patient outcomes once factors such as patient co-morbidity or specialisation were controlled.

Only two studies of psychiatrists were included in this review, both using clinical vignettes. The first study of a small cohort of only 38 psychiatrists showed that diagnostic accuracy of anxiety and depressive disorders in the medically ill, compared with expert consensus guidelines, decreased as a function of years in practice (Epstein et al., 1996). The second, consisting of 278 psychiatrists, showed that age did not influence antidepressant prescribing, but psychiatrists in practice for fewer years were significantly more likely to diagnose depression in the medically ill (Epstein et al., 2001). It is unclear how such tests based on hypothetical reasoning correlate with clinical performance. Further, these findings may have been due to a cohort effect with less experienced psychiatrists more likely to be practising in managed care systems that required Diagnostic and Statistical Manual of Mental Disorders diagnoses, whereas more experienced psychiatrists may have been reticent to diagnose major depression according to rigid diagnostic criteria, especially in the context of medical illness.

Notwithstanding the limitations of these two particular studies, there are a number of factors which might contribute to poorer performance in older doctors, including natural diminution in cognition with ageing, neurodegenerative disease, as well as other non-cognitive processes. In a study of poorly performing practitioners referred to a Canadian health programme, a significant number (particularly among older doctors) had cognitive impairment sufficient to explain their poor performance (Turnbull et al., 2000). However, the negative correlation between age and performance was even greater when those performing most poorly on neuropsychological assessment were excluded, suggesting a significant contribution of non-cognitive factors, such as sensory decline, to impairment.

Age and cognition

With regards to natural diminution of cognition associated with ageing, there is no reason to suggest that doctors are immune. Fluid intelligence – the capacity to think adaptively and apply critical or analytical reasoning – memory and speed of information processing all decline with age (Christensen, 2001). Yet, there is age-associated stability, and possibly improvement, in crystallised intelligence, a measure of accumulated knowledge and wisdom that is dependent on education and experience. For instance, compared to younger peers, older and more experienced medical practitioners perform as well with regard to medical knowledge that has been static since their residency, but demonstrate poorer knowledge of new information (Day et al., 1988). Diagnostic accuracy is greater in more experienced practitioners when only contextual information is presented (Hobus et al., 1987), but declines when contradictory information about a case is presented (Cunnington et al., 1997).

These findings suggest that the capacity for non-analytical thinking is retained by older practitioners, but consideration of more complex problems or alternate hypotheses is confounded by impaired analytical abilities, and that there is a reduction in the ability to assimilate new information into clinical practice. The relevance of these findings to the practice of psychiatry, which speculatively is highly reliant on analytical thinking and consideration of complex problems using fluid intelligence, is unknown.

Another important caveat to any conclusions about ‘normal ageing’ is the increase in distribution of cognitive performance scores, with greater inter-individual heterogeneity associated with age (Christensen, 2001). This issue is crucial when we are developing policies for ageing practitioners. For example, a neuropsychological study of practitioners across the age spectrum showed that while the average total cognitive score declined with age, the standard deviation increased with age, suggesting potential heterogeneity of performance (Weintraub et al., 1994). Almost 15% of practitioners aged 75 years and older performed in the average range of those aged less than 35 years.

In addition to recognising this large variation in cognitive ageing among medical practitioners, it is important to parcel out the ‘task-specific’ aspects of each speciality in understanding the relevance of cognitive changes. Social cognition, simplified as the capacity for empathy and sympathy, is the most recently studied domain of cognition and is likely to be relevant to psychiatry. Theory of mind asserts that people are able to imagine that they and others have minds that guide behaviour, and that others may have different perspectives (Rakoczy et al., 2012). Preliminary studies suggested that there may be lifelong improvements in social understanding. More recent studies have shown that older adults were less proficient in the ability to infer and ascribe complex intentional attitudes, and to recognise and ascribe emotional states, although age-related differences were largely explained by declines in processing speed and executive function (Rakoczy et al., 2012).

While social cognition has not been extensively studied across age groups, and the use of video screens to study recognition of facial expression again may not correlate well with clinical practice, we propose that the putative skills of the psychotherapist in particular requires intact neurocognition and social cognition. For instance, therapeutic alliance has been linked to personal attributes of the therapist, including flexibility, warmth, interest, and to therapist techniques, such as exploration, reflection, facilitating the expression of affect, accurate interpretation of and attending to the patient’s experiences (Ackerman and Hilsenroth, 2003). Perceiving prosody and facial expression requires intact temporoparietal function; frontal executive function is required for abstract thinking and verbal abstract reasoning; and exploration and reflection require intact working memory. Empirical research of these hypotheses regarding the relevance of social cognition to performance in ageing practitioners is essential.

Finally, with regards to the contribution of frank neurodegenerative processes to impairment, medical practitioners have a reduction in the theoretical risk of developing dementia. Factors such as higher levels of education and occupational complexity lower the risk of dementia by 45–50% (Valenzuela and Sachdev, 2006). Yet, medical practitioners are certainly not immune to dementia, as borne out by the aforementioned study of the NSW Medical Board (Peisah and Wilhelm, 2007). Moreover, dementia is a very late presentation of cognitive impairment. More subtle cognitive changes, encapsulated by the diagnosis of mild cognitive impairment, known to affect higher-level complex cognitive functions, may be more relevant.

Physical illness, psychological disorder and cognitive impairment

There is a strong association between physical health, psychological problems and cognitive impairment.

One-third of doctors who attend psychiatric care experience concomitant chronic physical illness (Kay et al., 2004). Yet, medical practitioners do not necessarily manage their own physical health and illness well, often failing to follow preventative health guidelines or consult an independent general practitioner (GP). Physical illnesses associated with vascular pathology such as hypertension, dyslipidaemia, diabetes mellitus and obesity all increase the risk of cerebrovascular disease and therefore cognitive impairment and depression (Hickie et al., 2001).

Although high levels of depression have been reported in medical practitioners, studies have not reported rates by age (Clode, 2004). The use of alcohol may be declining in medical practitioners, although this is not the case with other substances.

What is most relevant here is the role of actual and impending professional losses in the development of depression and substance use in older and late-career psychiatrists. Being a doctor represents a deeply held and intrinsic sense of identity (Sadavoy, 1994) related to self-worth, contribution to the community, intellectual stimulation and therapeutic and professional relationships, all placed at risk by ageing.

Retirement

Age, finances, health and family/personal reasons were the most frequent criteria for retirement anticipated by working psychiatrists who were significantly more likely to endorse the latter two factors than retired psychiatrists (Draper et al., 1997). Of active RANZCP Fellows aged 55 years and older surveyed by Draper et al. (1997), 60% had commenced retirement planning, mainly by financial planning, reducing working hours or acquiring hobbies and other interests. This is consistent with proactive planning demonstrated by a group of doctors deemed to be ‘successfully ageing’ by their peers (Peisah et al., 2009a).

Of concern, however, is the 17% of psychiatrists who reported no intention to retire, meaning that up to 40% of late-career Australian and New Zealand psychiatrists may make no end-of-career plans (Draper et al., 1997). While the authors did not explore this, possible reasons include a desire to avoid the impending, multiple professional losses noted earlier, a lack of financial independence, a failure to develop non-professional interests, or the need to retain control in determining the time of retirement.

It is possible that the results of this 16-year-old study are now outdated and that subsequent cohorts of psychiatrists have a view of retirement that reflects wider societal changes. In particular, the baby boomer generation, which has come of retirement age in the meantime, may be more resistant to retirement than previous generations.

Possible solutions

While the social, workforce and personal benefits of continuing to work are clear, psychiatrists need to be aware of the potential for age-related changes to cognition, performance, physical and psychological health that affect the capacity to practise safely. A number of preventative measures need to be considered.

Mandatory retirement

In Australia, pilots and federal judges are subject to mandatory retirement; an exemption to the Age Discrimination Act 2004 would need to be sought by the Commonwealth to apply this to medical practitioners. Notwithstanding age-related decline in even the best performers in explicit memory and processing speed, neurocognitive performance in ageing is associated with increased inter-individual heterogeneity (Christensen, 2001; Weintraub et al., 1994). An arbitrary retirement age would then run the risk of losing a cognitively well-performing group of late-career psychiatrists from the workforce.

Mandatory cognitive screening

Such a policy may be targeted at either all psychiatrists at a specific age threshold, or selected ‘high-risk’ groups (Peisah and Wilhelm, 2007). For instance, older practitioners charged with drink-driving, the subject of patient complaints or conduct/performance issues, or those working in isolated solo practice may be considered at increased risk of impairment.

Such a policy would be difficult to implement. Neuropsychological assessment is time-consuming, expensive and requires highly skilled practitioners. There remain uncertainties in the ecological validity of neuropsychological assessment, namely how test results may be used in the evaluation of professional capacity, and whether a practitioner, for instance, in the superior intellectual range premorbidly may retain capacity despite decline into the high average range.

While the use of alternative, more user-friendly screening tools that target cognitive vulnerabilities, such as processing speed and executive function, might be useful, it is unclear how the ‘task-specific’ cognitive requirements of each medical speciality would be considered. Fine motor skills and processing speed may be a priority for procedural specialties, while social cognition, flexibility and analytical diagnostic strategies may be a priority for psychiatrists. This remains speculative; the way in which the skills of a psychiatrist may be cognitively operationalised need to be empirically investigated.

Mandatory performance assessment

The UK General Medical Council has recently introduced a 5-year cycle of revalidation for all medical practitioners, for the purpose of demonstrating that each medical practitioner is ‘up to date and fit to practise’ (GMC, 2011). The framework on which revalidation is based consists of knowledge, skills and performance; safety and quality; communication, partnership and teamwork; and maintaining trust. It remains unclear whether this process will adequately evaluate fitness to practise.

Although no such process has been introduced in Australia or New Zealand, it is important that a novel model of demonstrating continuing competence is adopted to replace the prevailing model of assuming competence until proven otherwise. It is also essential that processes such as revalidation and continuing professional development are modified to fit the specific educational needs of late-career psychiatrists. This may include teaching the capacity to reflect and analyse critically long-held medical knowledge.

Monitoring of personal health

Given the strong association between physical health, mood and substance use disorders, and cognitive impairment, psychiatrists should be encouraged to monitor their physical health, such as consulting an independent GP and engaging in health screening. This should be both non-specific, for instance, the assessment of vascular risk factors, vision and hearing, and targeted at individual problems.

It should be recognised that practitioners may have difficulty accepting professional help for mood and substance use disorders. In many cases, such help is only recommended once a notification has been made to a regulatory authority. There is a need to introduce a well-resourced health programme for medical practitioners, and one that is voluntary, divorced from regulatory authorities, encourages early consultation and ensures long-term follow-up.

Workplace modification

Modification of clinical duties has been identified by other Colleges and should be considered for late-career psychiatrists (Skowronski and Peisah, 2012). Appropriate measures may include reducing total working hours, not practising in more acute clinical areas like psychiatric emergency care and stopping overnight and on-call duties. Practitioners should continue in familiar workplaces and in familiar areas of practice and subspecialisation.

Specific career planning for late-career psychiatrists is required. Administrative work, teaching, supervision and mentoring may be other avenues to pursue professionally with a concomitant reduction in clinical load. Voluntary clinical supervision, in addition to peer review, should also be considered. Just as the whole issue of retirement poses significant workforce issues in the current environment, so will practice adaptation issues, such as exemption from on-call duties. Systemic ratification of such measures is absolutely crucial to avoid resentment by junior colleagues who must understand that ‘all will have their turn’ (Skowronski and Peisah, 2012)

Encore careers

An encore career – one providing personal fulfilment, social impact and continued income (Freedman, 2007) – should be considered by late-career psychiatrists. Such careers may be an extension of interests or capacities built through the practice of medicine and psychiatry, or in an entirely different field, perhaps fulfilling long-held non-professional interests.

Long-term retirement planning

This should commence in a psychiatrist’s sixth decade, focusing on the provision of a plan from a finance industry professional and career counselling from peers.

The way forward

In the absence of dialogue or systemically ratified guidelines regarding an ageing workforce, there is apathy within the profession regarding support for late-career psychiatrists, workplace adaptations, retention strategies or retirement planning. The most catastrophic consequence of this lack of thought – represented by those psychiatrists who are identified as impaired – occurs in patients who may be harmed. It must also be recognised that colleagues and the practitioner’s family will also bear the burden of an impaired practitioner or the sudden retirement of a practitioner.

We urge our speciality to commence the process of discussing the challenges of ageing and retirement. Formal guidelines need to be generated and ratified by the profession as a group, and minimal content should include:

1. acknowledgment that ageing is associated with cognitive, physical and psychological changes, the extent of which is associated with inter-individual variability;

2. workplace adaptations for late-career psychiatrists that are acceptable to the workforce as a whole;

3. an endorsement of the need for long-term, albeit flexible, retirement planning; and

4. an investment in the empirical study of issues relating to cognition, ageing and the practice of psychiatry.