- R P Caplan
- Accepted 16 September 1994
Objective: To study stress anxiety, and depression in a group of senior health service staff.
Design: Postsal survey.
Subjects: 81 hospital consultants, 322 general practitioners, and 121 senior hospital managers (total 524).
Main outcome measures: Scores on the general health questionnaire and the hospital anxiety and depression scale.
Results: Sixty five (80%) consultants, 257 (80%) general practitioners, and 67 (56%) managers replied. Of all 389 subjects, 183 (47%) scored positively on the general health questionnaire, indicating high levels of stress. From scores on the hospital anxiety and depression scale only 178 (46%) would be regarded as free from anxiety, with 100 (25%) scoring as borderline cases and 111 (29%) likely to be experiencing clinically measurable symptoms. The findigns for depression were also of some concern, especially for general practitioners, with 69 (27%) scoring as borderline or likely to be depressed. General practitioners were more likely to be depressed than managers (69 (27%) v 4 (6%) scored >=8 on hospital anxiety and depression sacle-D; P=0.004) with no sgnificant difference between general practitioners and consultants. General practitioners were significantly more likely to show suicidal thanking than were consultants (36 (14%) v 3 (5%); P=0.04) but not managers (9 (13%)). No significant differnce could be found between the three groups on any other measure.
Conclusions: the levels of stress, anxiety, and depressionin senior doctors and managers in the NHS seem to be high and perhaps higher than expected.
Health service implications
Few studies have examined stress and psychological symptoms in senior hospital doctors
Over half of senior medical staff suffer high levels of stress
Only 46% of those questioned would be regarded as free from anxiety
General practitioners were more likely to be depressed and show suicidal thinking than were managers and consultants
Senior doctors in general suffer more from stress than expected, but this is not restricted to medical staff as managers also suffer. This stress may have far reaching effects on the NHS and patient care.
Many reports have suggested that stress among doctors is high, but careful searches of the literature show that few studies have measured stress with much reliability or validity. Most comment on the postulated causes of stress or on specific subgroups of the medical specialties, and few use standardised rating scales. It is rare to find studies that include a control group. The most frequently quoted British study is confined to junior doctors in their preregistration year.1
It is difficult to find a study that includes consultants across specialties and that provides much information about their current level of stress. A recent study has measured levels of stress among general practitioners before and after the introduction of the new contract and suggests that stress has risen among this group.2 Firth found that nearly a third of medical students and half of junior doctors in their preregistration year were suffering from emotional disturbance.1,3 The BMA has recently published a document on stress within the medical profession,4 but, although comprehensive, it contains no references or information about measured levels of stress or psychological symptoms in senior hospital doctors.
I set out to measure the current levels of stress among three defined groups who by and large lived and worked in similar geographical and demographical circumstances.
Subjects and methods
Three discrete groups of senior NHS staff were included in the study. The first group included all 81 hospital consultants working within North Lincolnshire Health Authority in all medical and surgical specialties. The second group included all 322 general practitioners registered with Lincolnshire Family Health Services Association. A third group included all 121 members of a group representing the senior and middle managers of North Lincolnshire Health Authority. This group, although not a control, was included to balance the “medical” variable. The managers' group included the district general manager, his chief officers, and all the key managers in the district.
All those taking part were asked to complete the general health questionnaire5 and the hospital anxiety and depression scale.6 The general health questionnaire is a widely accepted and reliable method of establishing minor psychiatric disorders among general populations. It has also been validated as a suitable. measure of mental health in occupational studies.7 I used version 28 in this study with a cut off point of 5/6.8 Five of the seven items in the severe depression subscale of the questionnaire measure aspects of suicidal thinking. Respondents who scored on two or more of these five items were regarded as having notable suicidal thoughts. The hospital anxiety and depression scale is a well validated measure of the symptoms of anxiety and depression and produces two subscales: A for anxiety and D for depression. A score of between 8 and 10 on either of the subscales indicates borderline presence of anxiety or depression, with a score of over 10 indicating that these conditions are likely to be present.
Eleven consultants and 24 managers completed their rating scales while at, but before the presentation of, two meetings on stress in the workplace run by me, but most questionnaires were circulated by post. All the questionnaires for general practitioners were circulated by post.
Table I shows the results. I found no significant difference between general practitioners and consultants, general practitioners and managers, or consultants and managers on the hospital anxiety and depression scale- A or the general health questionnaire.
There were highly significant differences between general practitioners and managers on the hospital anxiety and depression scale-D (one way ANOVA, df=1, F=8.5, P=0.004) with no significant difference between general practitioners and consultants. Senior medical staff therefore seem considerably more likely to be depressed as a group than do senior managers.
General practitioners were significantly more like to show suicidal thinking than consultants (36 (14%) 3 (5%); X2=4.3, df=1; P=0.04, but not manage (9 (13%)). There were no other significant difference between groups.
Correlations were measured with Spearman's relation coefficients. Table II shows the results, indicate that the total score on the general questionnaire correlates positively with scores on hospital anxiety and depression scale- A and D suicidal thinking. Scores on the hospital anxiety on the hospital anxiety and depression scale-D suicidal thinking, with scores on the hospital and depression scale-D correlating only with thinking. This is much as one would expect and that there is a strong internal consistency within results. There was no correlation between type and any of the measures.
Overall, the main and disturbing findings are nearly half of the consultants, general practition and managers scored positively on the general questionnaire (version 28), whereas in the population only 26.8% would be expected to positively on the general health questionnaire 30) in the East Midlands.9 A recent major survey British population found that the percentage positively on version 30 for those in the profession and managerial group was 27% for men and 28% women.10 This version, which although not comparable with version 28, shares many and a similar modal cut off score for identi “caseness.” Other studies, including those on hospital doctors by Firth,1,3 used version 12. compared the three versions in the same common population and found high correlations between with version 28 proving to be the most reliable three, showing the highest sensitivity and especially with a cut off point of 5/6.11 I used 28 as it allows for more detailed investigation depression and in particular suicidal thinking. comparisons can probably be drawn between by using either version.
There are no normal population values hospital anxiety and depression scale-A or the anxiety and depression scale-D, but there are accepted cut off values which have been valid several studies.12 With these cut off values only senior doctors and managers would be regarded from anxiety, with 25% scoring as borderline case 29% likely to be experiencing clinically symptoms. The findings for depression are some concern, especially for general practi with 27% scoring as borderline or definitely like depressed.
The high response rate for senior doctors (80 consultants and general practitioners) is good postal survey. The response rate for manager reasonable at 56%. There is a current perception that the medical profession is suffering from increased stress, but most of the attention has been focused on junior doctors and their excessive hours of work. This study shows that senior doctors also suffer from considerable amounts of stress and perhaps more than expected. The inclusion of senior managers in the study, however, has shown that there seems to be an equivalent amount of stress in that group, and this should widen the debate. Stress is not the sole property of the medical profession.
I thank Sheila Teasdale of Lincolnshire Medical Audit Advisory Group for her help in the preparation of this study and Professor Mike Pringle of the department of general practice at Nottingham University for his help with the statistics.