Participants and questionnaire

Following the methodology of the original study, an anonymous confidential postal questionnaire was sent to all physicians and surgeons carrying a clinical case-load involving regular face-to-face contact with patients and working at St George's Hospital, a teaching hospital in London. Permission was granted by the medical advisory committee of that hospital to carry out a survey of staff. In line with Mayou & Smith's study, paediatricians and pre-registration house officers were excluded. Copies of the questionnaire are available from the authors upon request. The questionnaire was adapted from questions previously used by Mayou & Smith on the comparable population of the Oxford General Hospitals' list of clinical staff, which in turn were adapted from questionnaires used in studies of general practice by Shepherd et al (Reference Shepherd, Cooper and Brown1966) and of referral to psychiatrists by Mezey & Kellett (Reference Mezey and Kellett1971). In contrast to the Oxford hospitals studied by Mayou & Smith, St George's Hospital has received the input of a modestly resourced liaison psychiatry service for over a decade. The questionnaire comprised questions pertaining to hospital doctors' assessment of psychological problems, attitudes to psychosocial care, referral to psychiatrists and treatment of psychological problems. Additional questions included non-identifying background details about respondents. Although the questionnaires were completed anonymously, responses were numbered to permit identification of speciality and grade. A total of 274 doctors were sent the questionnaire.

Statistical analysis

Descriptive statistics were recorded for questions and univariate comparisons were performed using chi-squared or Fisher's exact tests as appropriate. Statistical significance was set at the 5% level.

General attitudes

Hospital doctors' attitudes to psychological factors (Table 1) show heightened awareness of the relevance of these factors compared with 1986. In addition, 95% saw the emotional care of patients as being a key element of their work. In this sample, there was evidence of the enhanced role of nurses in managing social and emotional difficulties of patients.

Table 1 Attitudes to the management of psychological problems

Statement	Percentage in agreement
	2001	1986
Psychological factors can influence the cause and outcome of physical disorders	96	77
Emotional and social aspects of care enhance job interest	58	66
Management of emotional issues is solely a medical responsibility	25	33

In addressing responsibility for management of common problems (Table 2), there appeared to be a greater sense of responsibility for the emotional care of dying patients and for overdoses than was found in Mayou & Smith's sample. More than three-quarters held the view that they had primary responsibility for acute confusional states and the emotional care of dying patients, and over half for overdoses, with less sense of responsibility for depression, alcohol misuse or behavioural disturbance.

Table 2 Responsibility for common types of psychological problems (1986 responses given in parentheses)

Statement	Strongly agree (%)	Agree (%)	Uncertain (%)	Disagree (%)	Strongly disagree (%)
Hospital doctors (other than psychiatrists) have the major responsibility for the management of the following problems:
Depression	10 (5)	21 (21)	10 (20)	49 (41)	9 (3)
Acute confusional state	27 (17)	51 (59)	7 (13)	9 (10)	7 (1)
Overdoses	24 (7)	37 (29)	9 (33)	19 (27)	10 (3)
Chronic drinking problems	6 (5)	24 (20)	18 (33)	36 (38)	16 (5)
Disturbed behaviour	4 (3)	19 (25)	23 (35)	34 (34)	20 (3)
Emotional care of dying patients	15 (10)	59 (23)	14 (26)	8 (37)	3 (4)

Time constraints and assessment

Doctors' time was more constrained than in the original study and this affected the capacity to conduct biopsychosocial assessments, despite awareness of their relevance (Table 3).

Table 3 Time constraints and assessments

Statement	Percentage in agreement
	2001	1986
I would welcome more time to talk to my patients	92	78
It is impractical for hospital doctors to assess and treat emotional problems	52	46
Psychological and social factors should be routinely assessed and recorded for in-patients	78	Not recorded
When psychological factors appear to be an important cause of the presenting problem, I confine myself to physical assessment	16	35
I should concern myself with emotional care of regular attenders with chronic physical illnesses	80	60

Treatment

The majority of respondents felt that hospital doctors should be able to make use of simple psychological methods, with greater use of behavioural therapies. There was greater use of antidepressants (Table 4), and selective serotonin reuptake inhibitors were the most commonly cited drugs of choice, with amitriptyline also used. Two respondents suggested the use of diazepam as an antidepressant. In the 1986 study the most commonly used antidepressants were amitriptyline and mianserin. Similarly, most had treated insomnia, generally favouring short-acting benzodiazepines, as well as sedating tricyclic antidepressants and zopiclone.

Table 4 Attitudes to treatments

Statement	Percentage in agreement
	2001	1986
Hospital doctors should be able to use psychological methods such as:
Listening/reassurance	88	‘Most’
Discussion of anxieties and problems	94	‘Most’
I frequently discuss emotional problems with relatives	73	55
I use cognitive or behavioural methods of treatment	36	<25
Hospital doctors should be able to use psychotropic drugs	78	81
I use antidepressants frequently or occasionally	601	43

Respondents who had treated acute alcohol withdrawal (48%) in the current study favoured the use of benzodiazepines over clomethiazole, and those who had treated anxiety disorders (40%) tended to use benzodiazepines or beta-blockers. Acute confusional states had been managed by 46% of the sample, by ‘treating the underlying cause’ and using traditional neuroleptics such as haloperidol when necessary. Although 21% had experience of treating psychoses with neuroleptics, only three respondents cited experience of atypical antipsychotic drugs.

Attitudes to psychiatry and barriers to referral

There appeared to be a greater desire for and interest in liaison psychiatry (Table 5). Reasons for not referring patients to psychiatric services were similar to those cited in Mayou & Smith's study. The most common reason was the belief that patients dislike referral, followed by fear of stigmatising patients by psychiatric referral. The perceived ineffectiveness of psychiatric interventions was a lesser consideration. As in the earlier study, we received comments requesting greater input from senior psychiatrists. Psychiatrists' insistence on the exclusion of organic causes of disordered behaviour was also a source of adverse comment.

Table 5 Attitudes to psychiatry and barriers to referral

Statement	Percentage in agreement
	2001	1986
I would like more contact with psychiatric services	78	‘Just over half’
I would like to know more about what psychiatry has to offer in the management of medical or surgical patients	73	Not reported
Psychiatrists have little to offer in a general hospital	5	24

Pattern of replies

Differences emerged in responses based on gender. Male doctors were far more likely than female doctors to limit themselves to physical examination, even when psychological factors appeared to be an important cause (χ² ₁=8.56, P=0.003), and to perceive psychiatrists as having little to offer in a general hospital (χ² ₁=8.01, P=0.018). Men were less likely to find that the variety of emotional and social care enhanced their work interest (χ² ₁=3.95, P=0.047). None of these outcomes differed significantly on the basis of seniority or speciality.

In contrast with 1986, there were few significant differences based on speciality (Table 6).

Table 6 Attitudes of physicians and surgeons

Statement	Percentage in agreement
	2001		19861
	Surgeons	Physicians	Surgeons	Physicians
Emotional problems are part of hospital doctor's work	93	97	73	90
Psychological factors are important in the course of physical illness	97	96	67	85
The variety of emotional and social care enhances interest	51	62	60	85
General practitioners are responsible for assessment of emotional problems in new out-patients	31	30	76	35
Management of emotional problems is an important part of my care for chronic out-patients	67	792	47	99
Management of my patients' emotional problems is mainly the responsibility of nursing staff	24	23	49	22

Limitations of the study

This study has four principal limitations. First, it is difficult to extrapolate from the results of a questionnaire to clinical practice. The results of this study may represent the effects of normative social influence and respondents' wishes to make a good impression, rather than true clinical procedure. This form of response bias is technically known as ‘social desirability bias’ and future studies might address it by use of lie scales. Second, the study was conducted at a single centre, and the findings may not apply to other hospitals. Third, because it was not possible to replicate precisely the questionnaire used by Mayou & Smith, responses to questions based on the original questionnaire might have been influenced by factors such as the order of questions. As a result, comparisons between the two studies are qualitative and not quantitative. Fourth, it is regrettable that the study was not repeated in the original hospitals.

Differences between 1986 and 2001

Awareness of the relevance of psychological factors to medical and surgical patients contrasts with the findings of Mayou & Smith. Differences between the two samples could be explained by three elements. First — and most optimistically — they might represent a genuine shift in the culture of the medical profession. Over the intervening 15 years the undergraduate curriculum has moved towards a focus on biopsychosocial constructs of disease and the value of doctor—patient communication. For example, the General Medical Council (1993) has stressed the importance of medical students learning how to carry out a mental state examination. That so many respondents acknowledged the impact of psychopathology on prognosis may exemplify the fruits of these labours in terms of knowledge of psychiatry, although elsewhere there is little evidence that undergraduate education improves attitudes to psychiatry (Reference Calvert, Sharpe and PowerCalvert et al, 1999). In this sample, women were far more likely than men to recognise the relevance of psychological factors, the value of liaison psychiatry input and the contribution of emotional care to their job satisfaction. At present, 58% of applicants and 59% of successful entrants to medical school are women (Reference MooreMoore, 2002). Thus, some differences between 1986 and 2001 may simply represent the increasing proportions of women in the National Health Service (NHS) workforce.

Second, the differences between the two studies might indicate the influence of a liaison psychiatry service on the institution it serves, generating heightened awareness of unmet need among surgeons and physicians (Reference Benjamin, House and JenkinsBenjamin et al, 1994; Reference Storer, Peveler, Feldman and FriedmanStorer, 2000). However, this is an unlikely explanation. In 1986, Oxford was already leading the way in psychological medicine, with the first full-time consultant liaison psychiatrist in Britain, and so this is unlikely to provide a comprehensive explanation of differences.

Third, differences might be an artefact of institutional differences between St George's Hospital in 2001 and Mayou's sample of Oxford hospital doctors in 1986. This seems possible, but not probable, with no reason to expect major cultural differences between the two teaching hospitals.

In conclusion, the most plausible explanation of differences between the two studies lies in genuine changes in hospital doctors' attitudes to and knowledge of psychological problems.

Attitudinal homogeneity

Whereas the 1986 study found widely differing views regarding psychosocial care, our study found greater attitudinal homogeneity. In particular, in the earlier study consultants were less likely than their juniors to see psychiatric referral as serving a useful purpose, and junior doctors were less likely to see emotional problems as part of the hospital doctor's job. Differences between consultants and their juniors did not emerge in the 2001 survey, nor did differences between surgeons and physicians. There appears to be a shrinking minority of hospital doctors who focus on the physical complaint to the exclusion of relevant psychosocial factors.

An enhanced sense of responsibility for the management of overdoses may reflect the continued rise in rates of deliberate self-harm since the 1980s to the point where self-harm (most commonly manifest as overdose) is one of the top five reasons for acute medical admission (NHS Centre for Reviews and Dissemination, 1998). In contrast, the relative neglect of depression is consistent with the time constraints of hospital medicine, given that almost all respondents desired more time to communicate effectively with patients and over half felt unable to address emotional factors under the current limitations of the NHS.

Barriers to psychiatric treatment

The fear of stigmatising patients by providing psychiatric input was the exception to this trend towards better management of patients' psychological needs, and this is consistent with attitudes in the community reported in this journal (Reference ByrneByrne, 2001; Reference CrispCrisp, 2001). Pejorative attitudes to mental disorders among some obstetricians and gynaecologists have previously been reported (Reference MorganMorgan, 1999), and male gender appears to generate pejorative attitudes in this study. The perception of the ineffectiveness of psychiatric interventions and concerns over stigmatisation provide two more barriers to adequate psychiatric care. However, the greatest barrier to treatment seems to be the lack of time to communicate and evaluate the psychological needs of patients.

Psychotherapy and pharmacotherapy

The widespread use of listening, reassuring and discussing anxieties by hospital doctors was heartening, as the claim of a substantial minority to practise ‘behavioural methods’ was unexpected. The study did not address the detail of these interventions, but at the very least this seemed to indicate an awareness of cognitive—behavioural therapy and its efficacy. It also contrasted with the results of the original survey, in which hospital doctors appeared to rely on pharmacological rather than psychological treatments. The move towards the prescription of selective serotonin reuptake inhibitors reflects national prescribing habits, but this was not matched by use of atypical antipsychotic drugs despite their particular value in the medically unwell population.

The national context

Given that the doctors who responded to the questionnaire appeared to appreciate the psychological needs of their patients and yet reported insufficient time to meet those needs, it was unsurprising that the vast majority of respondents desired greater psychiatric input. Well-developed liaison psychiatric services permit health care trusts to achieve essential performance indicators and offer financial savings in excess of the cost of liaison psychiatric services, known as the ‘cost-offset effect’ (Royal College of Physicians & Royal College of Psychiatrists, 1995). Properly resourced liaison psychiatry services are central to the promotion of mental health among medical and surgical patients, and physicians and surgeons appear to acknowledge that this facet of hospital care is a key element of a high-quality health service.

This study implies that hospital doctors have increased their aspirations to provide biopsychosocial care of medical and surgical patients over the past 16 years. Physicians and surgeons have greater awareness of their patients' psychiatric requirements. Most hospital doctors would like more contact with psychiatric services. This provides a powerful argument for the further development of liaison psychiatry services.