Senior house officers' work related stressors, psychological distress, and confidence in performing clinical tasks in accident and emergency: a questionnaire studyBMJ 1997; 314 doi: https://doi.org/10.1136/bmj.314.7082.713 (Published 08 March 1997) Cite this as: BMJ 1997;314:713
- Susan Williams, research fellowa,
- Jeremy Dale, senior lecturer in primary carea,
- Edward Glucksman, consultantb,
- Amanda Wellesley, senior registrarb
- a Department of general practice and primary care, King's college school of medicine and dentistry, London SE5 9PJ
- b Department of accident and emergency medicine, King's college hospital, London SE5 9RS
- correspondence to: Susan Williams
- Accepted 29 November 1996
Objective: To investigate the relation between accident and emergency senior house officers' psychological distress and confidence in performing clinical tasks and to describe work related stressors.
Design: Questionnaire survey with data collected at four points during senior house officers' six month attachment to accident and emergency departments.
Subjects: 171 newly appointed accident and emergency senior house officers from 27 hospitals in the South Thames region.
Main outcome measures: Psychological distress measured with a 25 item questionnaire; confidence in performing a range of 35 clinical and practical activities (visual analogue scales); reported consultation stress factors, other work related stressors, and personal stressors.
Results: Overall confidence scores in carrying out a range of clinical and practical activities increased significantly between the end of the first and the end of the fourth month (Z=-6.05, P<0.001). Senior house officers with higher psychological distress scores at the end of their first and fourth month had significantly lower confidence scores (Z=-3.20, P<0.001; Z=-1.90, P<0.05). Senior house officers with lower increases in confidence between the first and fourth month had significantly higher distress than those with greater increases (Z=-2.62, P<0.001). Factors identified as causing stress during consultations included difficulties with communication, certain clinical presentations, and department organisational factors (particularly the intensity of workload).
Conclusions: Psychological distress is linked to confidence in senior house officers. This supports the need to monitor and build confidence in senior house officers and to address work related stressors. Additional communication skills training needs to be considered.
The importance of dealing with stress in junior doctors has been emphasised since excessive levels of stress may lead to dissatisfaction, lower morale, and poorer work performance
Senior house officers' psychological distress levels is inversely related to their confidence in carrying out a range of clinical tasks in accident and emergency
Intensity of workload, communication difficulties, and feeling concerned about incorrect or missed diagnoses were common stress factors
Senior house officers may benefit from training in communication skills
There remains a need to reduce high work intensity and provide greater support
Stress affecting junior doctors is receiving increasing attention.1 2 3 4 5 6 7 8 It has been reported that excessive levels of stress may lead to dissatisfaction, lower morale, and poorer work performance.5 Although it has been stated that there is a need to reduce excessive levels, little work has attempted to investigate relations between junior doctors' psychological distress levels and their clinical practice.
We looked at training needs of senior house officers in accident and emergency departments. Our initial aim was to investigate the relation between psychological distress and confidence in relation to a range of clinical and practical activities. Confidence was considered to be an important variable since it has been shown to be a subjective marker of competence and to be one component of the ability to function competently.9 We also set out to report on areas where senior house officers have least confidence and to identify the occupational factors that cause greatest stress.
All 28 accident and emergency departments in the South Thames region were invited to participate in the study. All newly appointed senior house officers employed between August 1994 and February 1995 were sent questionnaires to be completed on their first day and at the end of the first, fourth, and final months of their attachments. Instructions for all four questionnaires stated that all responses would remain totally confidential.
Confidence–The questionnaires sent at the end of the first and fourth months asked subjects to rate their level of confidence in performing a range of 35 clinical and practical activities on 100 mm visual analogue scales (0=not at all confident, 100=very confident). The items were selected to represent a cross section of the range of specialties, types of patient presentations, and technical procedures senior house officers may deal with in accident and emergency.
Psychological distress–All four questionnaire booklets included the psychological distress questionnaire, which was based on existing scales, excluding items that could be confounded by the nature of working as a senior house officer in the accident and emergency setting. The measure consists of 25 items (14 negatively worded and 11 positively worded): 21 items were modified from the mental health inventory10 and four from the 28 item general health questionnaire.11 Scoring is on a six point scale. For analysis scores were reversed for negatively worded items so that for all ratings 1=lowest distress and 6=highest distress. The overall mean score for all completed items was calculated for each subject.
Work related stressors–Questionnaires sent at the end of the first and fourth months asked subjects “at the end of each of your next three shifts” to describe the case presentations that had caused most difficulty. For each, they were asked to indicate whether difficulties were due to “stress and/or worry caused by the consultation” and if so, to describe what they found stressful or worrying. At the end of the fourth month subjects were also asked to report the main factors that had caused them to feel stressed at work over the past month.
Personal stressors–In the final questionnaire subjects were asked to list all major stressful events that had occurred over the past six months.
The first battery of questionnaires was administered to a separate sample of eight accident and emergency senior house officers in July 1994 to test for comprehensibility.
Statistical analysis was carried out with the SPSS-X statistical package.12 Confidence item scores at the end of months 1 and 4 were summed and the mean overall confidence score calculated. Wilcoxon matched pairs tests were used to investigate changes in confidence scores and psychological distress over time; Mann-Whitney U tests were used to analyse the relation between confidence and distress levels. The work related factors that caused subjects to feel stressed were analysed thematically and quantitatively according to the number of times issues were reported.13 This procedure was repeated by an independent researcher to ensure reliability.
The sample included 171 senior house officers from 27 sites; the mean age was 26.5 (range 23.7 to 35.7) years. The response rates to the questionnaires were 82% (140 respondents) at the start, 77% (132) at the end of the first month, 64% (110) at the end of the fourth month, and 67% (115) at the end of the sixth month.
Confidence over time
Figure 1) shows the distribution of overall confidence scores at the end of the first and fourth months for the 105 subjects completing both questionnaires. The median overall confidence score was 54.8 (range 37.8 to 78.5) at the end of month 1 and 59.7 (range 40.3 to 91.4) at the end of month 4. There was a significant increase in overall confidence (Z=-6.05, P<0.001); the median change was 138.5 (range -507 to 1042). Men (n=65) had significantly greater overall confidence than women (n=47) at the end of the first month (median score 57.3 v 53.7; Z=-2.32, P<0.05), but there was no significant sex difference at the end of the fourth month.
Tables 1 and 2 show the change in confidence scores between the end of the first and fourth month. Table 1) presents scores for the 14 items with median scores at or below 50 at the end of month 1, and table 2) shows the 21 items with median scores above 50 at the end of month 1.
Figure 2) shows the distribution of scores on the psychological distress questionnaire at the start of the study, the end of the first month, and the end of the fourth month for the 97 subjects completing all three questionnaires. Distress levels did not differ significantly between these time periods. At the end of the first month, mean stress score had decreased 0.05 (SD 0.53; range -2.28 to 1.72) points; at the end of the fourth month the mean decrease was 0.05 (0.59; -1.96 to 2.12) in comparison with the start. Between the first and fourth months, mean distress score increased 0.04 (0.50; -1.84 to 1.72). The maximum change in distress scores for individual subjects was 2.28; standard deviations were between 0.5 and 0.59), indicating that for most subjects changes in distress scores were less than 1.0 during the study period.
A total of 116 (76%) subjects reported experiencing no stressful personal life events during the first four months of the study. The remaining subjects experienced up to four stressful events such as death of a friend or relative, illness of a relative, divorce, relationship difficulties, personal injury, illness, taking exams, financial difficulties, and moving house.
Distress and confidence
When the sample was split at the median for overall confidence, subjects with lower confidence at the end of the first and fourth months showed significantly higher distress scores than those with higher confidence levels (Z=-3.20 (P<0.001) v Z=-1.90 (P<0.05; fig 3). When this analysis was repeated for subjects reporting no stressful life events the same results were found (Z=-3.19 (P<0.001) v Z=-1.90 (P<0.05)).
Subjects with change scores below the median had significantly higher psychological distress scores at the end of the fourth month than those with scores above the median (Z=-2.62 (P<0.01); fig 4). The same results were found for the group reporting no stressful personal life events (Z=-2.18, P<0.05).
Work related stressors
Of a possible total of 726, 586 (81%) case descriptions of consultations causing difficulty were completed. The factors reported as contributing to the difficulties encountered related to three broad categories: the type of problem presented by the patient, communication difficulties, and department organisational issues (box).
Box 1—Factors causing senior house officers stress during 586 “difficult” consultations
Type of clinical presentation (91;16%)
Dead or dying patients (10)
Psychiatric patients (10)
Road traffic accident or acute multiple trauma (7)
Patient in pain (5)
Trivial condition (5)
Cardiac problems or arrest (4)
Head injuries (4)
Eye injuries (2)
Drug addicts or overdose cases (2)
Patient alcoholic (2)
Bowel obstruction (1)
Cervical spine injuries (1)
Problems entirely social (1)
Unable to/whether to admit patient (12)
Lack of confidence in own skills (11)
No pathology detected (7)
Problems assessing patient (4)
Concern about risk of contracting disease (2)
Embarrassment on examining patient (1)
Lack of experience (1)
Concern about missed or incorrect diagnosis (87)
Lack of experience (7)
Poor response to treatment (2)
Communication with patient or relatives (132;23%)
Dealing with demanding, manipulative, violent, or aggressive patients (50)
Dealing with family or friends; breaking bad news (34)
Poor historian or unwilling to provide history (15)
Uncooperative patient (15)
Patient disagreed with SHO's opinion (11)
Racism among patients (1)
Concern about medicolegal issues (15; 3%)
Medicolegal worries or threat of being sued (5)
Patient complained (5)
Restraining orders (4)
Legal implications of withholding information from police (1)
SHO's own feelings of stress or distress
Own anger or frustration (10)
Feeling of lack of authority or control (3)
Seeing aborted fetus (1)
Seeing decapitated body (1)
Unable to stay calm (1)
Made to feel guilty by patient (1)
Feeling sad or upset (1)
Discharge or referral (74;13%)
Discharging patient; concern about future problems (31)
Problems with referral/when to refer (27)
No GP or problem with GP (4)
Medical/nursing team (20;4%)
Problems with other teams; colleagues reluctant to see or admit patients (6)
Arguing with colleagues or nursing staff (4)
Had to disturb a colleague (2)
Lack of guidance from specialist teams (2)
Had to ask colleague for help (1)
Failure to do something urgently (1)
Racism among staff (1)
Departmental issues (33;6%)
Unable to treat patient; treatment not available on site (10)
Night consultations or disturbed sleep (7)
Fears for safety; no security in department (6)
Busy department (5)
Bed shortage (3)
Problems with defibrillating equipment (1)
Waiting for results (1)
At the end of their fourth month subjects reported the main factors that caused them to feel stressed while working during the preceding four weeks (box). The most frequently reported stressors were related to department organisational factors, in particular the intensity of workload.
Box 2—Main factors causing stress to 100 senior house officers during their fourth month working in accident and emergency
Patient factors (24%)
Violent, abusive, or demanding patients (16)
Trivial conditions (5)
Patients or relatives disagreeing with diagnosis or treatment plans (1)
Repeat attenders (1)
Breaking bad news (1)
Clinical presentation (17%)
Treating paediatric patients (5)
Drug or alcohol dependent patients (2)
Making mistakes (2)
Elderly patients (2)
Seriously ill patients (1)
Cervical spine injuries (1)
Multiple trauma patients (1)
Deliberate self harm (1)
Primary care patients unhappy with GP (1)
Personal concerns (4%)
Personal problems (2)
Low morale (1)
Lack of a challenge (1)
Intensity of workload (33)
Alone working nights; poor support from seniors (10)
Fatigue; lack of sleep (8)
Working alone; lack of support (6)
Lack of time off (6)
Unsociable rota (6)
Working for examinations (2)
Working nights after annual leave (1)
Feeling isolated (1)
Staff factors (19%)
Difficulty referring to specialties (8)
Tensions between staff (4)
Lazy colleagues (2)
Lack of supervision or teaching (1)
Violent nurses (1)
Having to refer to busy colleagues (1)
Difficulty in getting teams to review (1)
No paediatric cover (1)
Bed shortages (6)
Uncertain when to refer or admit (1)
Senior house officers with higher distress scores had significantly lower confidence in carrying out a range of clinical and practical tasks. Those with higher distress scores at the end of the fourth month also had significantly lower increases in confidence between the end of the first month and the end of the fourth month.
Despite the significant increase in senior house officers' confidence for carrying out a range of activities, confidence remained low in several areas. In particular, these related to certain technical skills (intubation, plastering, inserting a central venous pressure line) and managing patients with psychosocial problems (consulting with confused or abusive patients, dealing with patients' complex social problems, psychiatric problems). Confidence was higher for managing patients with trauma or acute medical or surgical needs, reflecting the emphasis in undergraduate training and preregistration house jobs.
A wide range of consultation problems, departmental organisational factors, and personal concerns were reported as stressors. Communication difficulties (dealing with demanding, manipulative, violent, or aggressive patients) were frequently cited. This suggests that training in communication skills may be beneficial. In a recent study of burnout and psychiatric disorder among hospital consultants Ramirez and colleagues found that clinicians who felt insufficiently trained in communication skills had significantly higher levels of distress.14 This finding may also apply in the early stages of medical careers. Training sessions in communication skills could provide the opportunity for case review.
In terms of factors causing most stress overall, the intensity of workload, coping with diagnostic uncertainty, working alone, working unsocial hours, and experiencing fatigue emerged as key stressors. This is consistent with other studies of junior hospital doctors,1 15 general practitioners,16 and cancer clinicians.17
The psychological distress questionnaire was developed from the mental health inventory10 and the general health questionnaire11 because existing scales included items that could be confounded by the nature of working as a senior house officer in accident and emergency. The aim of our study was not to describe psychiatric morbidity but rather to investigate how far psychological distress was associated with confidence levels. The distress questionnaire and confidence rating scale have not been tested for reliability and validity, and there was no control for the potential problem of reactivity in terms of biased response sets. Nevertheless, the distress and confidence scores show that all subjects' scores do not change in the same direction over time. The small number of senior house officers working at each site prevented investigating the extent to which confidence scores, distress scores, and stressors related to departmental features.
In senior house officers in accident and emergency departments, there seems to be an inverse relation between psychological distress and confidence in performing tasks. Future studies are required to establish if there is a causal link, the direction of causality, and to investigate the relation between senior house officers' confidence and their clinical competence. Our study indicates some work related stressors that need to be considered. Although training in communication skills may be beneficial,18 the problems of high work intensity for senior house officers and inadequate senior cover during nights need to be addressed. As stated recently, emergency medical care should not be provided by doctors who are chronically short of sleep or who have not had adequate rest periods during shifts.19
The assessment of confidence, distress levels, and work related stressors would enable training and support to be tailored to individual needs. With the increasing numbers of middle grade and multiconsultant departments, new support initiatives for senior house officers should be considered. Future work is required to establish whether these findings also apply to junior medical staff working in other specialties.
We thank all the accident and emergency medical, clerical, and secretarial staff who helped us during the study; Dr Derek Cooper for his statistical advice; Marilyn Peters, Floss Chittenden, Emma Smith, Kate Cowley, and Patsy Dixon for transcribing and coding the data; and Alex Dionysiou for his graphic work.
Funding: South Thames Primary Care Development Fund.
Conflict of interest: None.