Stress, burnout and doctors' attitudes to work are determined by personality and learning style: A twelve year longitudinal study of UK medical graduates

Sir William Osler (1849–1919), one of the most distinguished physicians of the nineteenth and early twentieth century, recognised that only some doctors are happy in their professional lives:

"To each one of you the practice of medicine will be very much as you make it – to one a worry, a care, a perpetual annoyance; to another, a daily joy and a life of as much happiness and usefulness as can well fall to the lot of man."[1]

The modern medical workplace is a complex environment, and doctors respond differently to it, some finding it stimulating and exciting, whereas others become stressed and burned out from the heavy workload. The medical workplace also provides an environment where new skills are continually being learned, both as a result of medical knowledge evolving and because a doctor's work changes, in part due to career development and progression through different jobs.

In an important study, Delva et al [2] used earlier research [3, 4] to develop two separate instruments for studying how doctors work, the Approach to Work Questionnaire (AWQ) and the Workplace Climate Questionnaire (WCQ). In Canadian physicians [2, 5] the AWQ showed three separate factors, which were called Surface-Rational, Surface-Disorganised, and Deep (see table 1). These approaches related to different methods and motivations for continuing medical education. Those with a deep approach preferred independent and problem-based learning and motivation was internal. Surface-rational and surface-disorganised approaches were primarily driven by external motivation, with the preferred mode of continuing education learning being independent for the surface-rational, and in consultations for the surface-disorganised.

The WCQ showed three dimensions, called Choice-Independence, Supportive-Receptive, and Workload (see table 1), which correlated with the AWQ. Doctors reporting Choice-Independence and Supportive-Receptive work environments had a Deeper approach, whereas those describing an environment dominated by Workload tended to be more Surface-Disorganised.

Some doctors are unhappy with their work, which can manifest as stress (usually assessed by the General Health Questionnaire) or burnout, which has three separate components of emotional exhaustion, depersonalisation and reduced personal accomplishment (see table 2). Greater stress and burnout in doctors are related to the personality trait of neuroticism or 'negative affectivity' [6].

The AWQ and WCQ provide a snapshot of a doctor's learning environment and approach to work at one particular time, as also do measures of stress and burnout. A key question, as Deary et al recognised [6] when considering stress, is the extent to which different approaches to work and the climate of the workplace are consequences of the workplace or of the doctor. At first sight it might seem that the workplace itself has to be the primary force driving both workplace learning and workplace climate. However, it is also possible that approaches to learning and work mainly depend upon pre-existing differences among doctors, differences that may already have manifested earlier in the doctors' careers. The AWQ bears a strong formal similarity to the surface, deep and strategic study habits and learning styles identified by the Study Process Questionnaire (SPQ), which assesses the motivations and approaches used by students in higher education (see table 3). The similarity is not accidental since the AWQ was developed by adapting items from Entwistle and Ramsden's Approaches to Study Inventory [7], which has a similar factor structure to that of the Study Process Questionnaire [8]. It is therefore expected that there may be significant continuities across approaches to study and approaches to work.

In this paper we describe a large cohort of UK doctors, typically aged 29 or 30 at the time of the study, who have been qualified for five or six years, who are practising as SHOs or SpRs in hospital or are in general practice, and who previously had been studied when aged 17 or 18 at application to medical school in autumn 1990 [9], in their final year at medical school [10] and as PRHOs [11]. The main interest here will be in the extent to which a doctor's present approaches to work and their workplace climate, as well as their stress and burnout, relate to earlier measures of study habits and personality at application to medical school and subsequently.

The tracing exercise looked for 2,912 individuals thought to have completed basic medical sciences and entered a clinical course. Eighty-nine had never been on the UK Medical Register, and either had failed finals, had never registered, or had emigrated. Of 2,823 individuals who were traced, 2,754 doctors were on the 2002 Register, 7 returned to the Register during 2002, and 64 were on an earlier Register. Of 2,823 questionnaires sent, 176 were returned by the Post Office as undeliverable, 10 doctors were travelling and hence uncontactable, and2 had died. Of the remaining 2,635 doctors, 1,668 returned questionnaires, giving a response rate of 63.3%. There was no evidence of response bias (see Supplementary Information see Additional file: 1).

Many doctors at the age of 30 are unhappy in their jobs, and a fifth of our sample reached the conventional GHQ criterion of psychiatric 'caseness'. In contrast, many doctors reported high levels of personal accomplishment, choice and independence in their work environment, satisfaction with medicine as a career, and intellectual and emotional satisfaction from their work. That is not new; Sir William Osler in 1905 contrasted doctors "whose stability of character and devotion to duty make one proud of our profession" with those who find it difficult to keep "the flame alive, smothered as it is apt to be by the dust and ashes of the daily routine" [1].

In 2001, Richard Smith asked "Why are doctors so unhappy?" and concluded that "The most obvious cause of doctors' unhappiness is that they feel overworked and undersupported" [23]. Certainly many doctors in our study report a high workload and a work climate that is neither supportive nor receptive, and those doctors also report more stress, burnout and dissatisfaction with medicine as a career. It is tempting therefore to conclude, as did an article in a special edition of BMJ Careers devoted to "Doctors' Wellbeing", that excessive workload and absence of support are directly caused by poor working conditions: "the way in which the NHS is run generates stress for members of the workforce every day" [24]. However, such an interpretation is not straightforward in general [25]. It is particularly difficult for the doctors in our study because the study is longitudinal, and workload and lack of support correlate with stress and burnout reported five or six years earlier, when the doctors were PRHOs and carrying out entirely different jobs. High perceived workload and poor support are therefore determined as much by doctors themselves as by specific working conditions. That view was expressed in another article in the special edition of BMJ Careers: "A critical element contributing to the stress that many conscientious doctors experience is internal..." [26]. A similar conclusion was reached in a previous study of ours when these doctors were PRHOs, and multi-level modelling showed that stress is not a characteristic of jobs but of doctors, different doctors working in the same job being no more similar in their stress and burnout than different doctors in different jobs [11].

If differences in reported workload are partly explained by differences among doctors, what in turn explains those differences? Doctors reporting a high workload also have what Delva et al [2] describe as a surface-disorganised approach to work, which in turn is correlated with being a surface learner at application to medical school, a dozen years previously. Surface-disorganised doctors are also high on the personality trait of neuroticism and low on the trait of conscientiousness; and again those correlations are with measures taken six years earlier when the doctors were PRHOs. Doctors reporting a work climate low in support were lower on the personality scale of agreeableness in the measures collected when they were PRHOs.

Some doctors may be stressed and burned out, but what predicts those others who are happy in their work? Doctors reporting high satisfaction with medicine as a career have a deep approach to work, and that approach is more common in those who also had a deep learning style when they applied to medical school. Satisfaction with medicine also relates directly to the personality traits of greater extraversion and lower neuroticism, and the deep approach to work correlates with greater extraversion and more openness to experience. Doctors who describe their colleagues as receptive and supportive score more highly on the personality trait of agreeableness; and as in many other correlations reported here, that correlation is stable across time – those who are more agreeable at the age of 24 have a more receptive and supportive work environment when aged 30.

An overview of our findings is that approaches to work are predicted by earlier measures of study habits and learning styles, whereas perceived work climate, and its pathologies such as stress and burnout, are predicted mainly by personality. Although unfortunately our study did not measure personality during selection, the high stability of the Big Five measures across the life-span [27‐29] (and across our two measures six years apart), as well as their heritable component [30], means that we have little doubt that personality at selection would also have been predictive, particularly given that a similar pattern of correlations was found in a different cohort of doctors in mid-career [15]. Other studies on very different groups of students have also found, like us, that both strategic and deep learning correlate with conscientiousness, and that deep learning also correlates with extraversion and openness to experience [22, 31]. Our study has, for various reasons, not looked at academic performance in relation to study habits, learning styles and personality, although previous work of ours has found clear correlations between learning styles and examination performance [32]. In contrast we have not found any correlation of undergraduate or postgraduate academic achievement with personality [15], and although some studies have found correlations of conscientiousness with academic achievement [33], this does seem to vary according to the learning context [34, 35]. Although we will be looking at this question again in more detail in a further analysis, it does seem probable that personality mostly has an indirect effect upon academic achievement via approaches to learning [31, 36].

If, as William Wordsworth said, "the child is father to the man", then the seeds of subsequent job satisfaction and dissatisfaction in doctors may be visible in the personality, motivations and learning styles of medical school applicants. This argument may provide some justification for using such measures in selection, particularly given the general association of job performance and satisfaction with personality [37] and motivation [38], and learning styles with personality [22] .

However, just as genes are not destiny, so neither personality nor learning style is destiny. Nurture interacts with nature [39], the environment building upon the genes, and the genes using what is provided by the environment; the poetic complement to William Wordsworth is therefore Alexander Pope, who said, "This education forms the common mind: Just as the twig is bent, the tree's inclined." Extreme introverts can, with sufficient insight, preparation and appropriate training become effective public speakers, less conscientious individuals can learn to be more organised and efficient, and those who are more neurotic can transcend their anxieties (and indeed neuroticism may be beneficial if sublimated into a professional concern for detail in critical situations, rather than merely being undifferentiated personal anxiety). .

Formal education, particularly effective formal education [40], can also alter study habits and learning styles, which are less fixed and 'trait-like' than personality measures [17]. Intercalated degrees increase deep and strategic learning and decrease surface learning at medical school [41], making it likely that they also encourage surface-rational and deep approaches to work. Deep and strategic learning also relate to the clinical experience gained by medical students [32], making it possible that greater patient involvement during undergraduate clinical training, rather than mere reliance on textbook learning to pass exams, a characteristic of surface learners, will also reduce surface-disorganised approaches to work.

Longitudinal data suggest that personality and learning style are not merely correlates of approaches to work, workplace climate, stress, burnout and satisfaction with a medical career, but are causes, events later in time being predicted by events earlier in time [35]. Doctors with greater stress and emotional exhaustion, who were less satisfied with medicine as a career, had higher neuroticism scores and were more likely to be surface-disorganised. Lower conscientiousness on the personality measure also predicted greater stress. Extraverts reported more personal accomplishment and were more satisfied with medicine. The personality measure of agreeableness predicted a more supportive-receptive work environment.

These results imply that differences in approach to work and workplace climate in our study result from differences among doctors themselves, as much as they do from differences in working conditions.