Workplace factors leading to planned reduction of clinical work among emergency physicians

Introduction

The ACEM was formed in 1983 with the first fellowship examinations being held in 1986. Since that time, the number of emergency physicians (EP) with ACEM fellowship has increased steadily and, in April 2002, the number had reached 510. This is consistent with the worldwide trend of increasing numbers of doctors entering training programs^1,2 despite the fact that emergency medicine is generally perceived as a stressful specialty.^3–7

During training, ACEM registrars generally have a more senior physician to call upon for advice and to whom some problems or complaints can be deflected. Upon achieving fellowship, however, responsibility can increase considerably beyond the immediate patient workload. This added responsibility, often coupled with relatively little support, is thought to contribute to increasing levels of perceived stress, in addition to that associated with shift work, night duty and weekend work.^8,9 Consequent upon these and other workplace difficulties, concern exists that some EPs are reducing, or planning to reduce, the number of hours spent in clinical work, opting instead for non-clinical work, part-time work or other totally unrelated work.^7,10

Reports published overseas, mostly in the UK and US, have examined the sources of EP stress and the rates of, and reasons for, EP attrition.^9,11–16 In the UK, these have included lack of funding and beds, understaffing, on-call requirements, conflicts between clinical and administrative responsibilities and the effect of work on family and personal life.^11,16 Alternatively, in the US, violence in the ED,¹² shift work,¹³ night duty, threat of malpractice and responsibility for dead and dying patients are commonly reported.^9,14 Reports of attrition rates overseas range from less than 1% per year¹⁴ to 10.4% over 3.5 years.¹⁵

To date, there has been little work in Australasia relating to EP stress and attrition. This study aimed to investigate the past, present and projected work practices of EP in Australasia. It also aimed to determine the most stressful aspects of clinical work, the impact of clinical work on a range of lifestyle issues, and the reasons why EP consider reducing their clinical load. These findings will indicate the important areas that may need to be targeted in order to improve clinical working conditions and EP lifestyle. Also, the results will relate directly to EP workforce planning, an area of importance to ACEM as the roles and expectations of EP in Australasia continue to evolve.

Methods

The study was a voluntary, anonymous, cross-sectional, mail survey of ACEM Fellows that comprised two separate components. The first, relating to the impact of workplace factors on clinical work, is reported here. The second component, relating to emergency physician psychological health, is reported elsewhere¹⁷ and will not be discussed below. The study was reviewed by the ACEM Scientific Committee and Executive, and authorized by the Swinburne University Ethics Committee.

The survey, including questionnaire dispatch and receipt, was administered through the ACEM head office during April and May, 2002. All ACEM Fellows were mailed a covering letter, the questionnaire, a plain envelope and a coded reply paid envelope. Completed questionnaires were sealed inside the plain and reply paid envelopes for return mailing to the ACEM. The plain envelopes were then forwarded, unopened, to the investigators for analysis. The coded reply paid envelope allowed the identification of respondents. Those who failed to respond were mailed a second questionnaire. This arrangement helped to ensure confidentiality of responses.

The respondents were required to report the number of hours they spend per week in a wide range of professional disciplines, as well as 5 years previously and 5 and 10 years hence. Likert scales were used to determine the impact of a variety of workplace factors on a decision to decrease clinical work. The scales ranged from 1 to 10 (1 = minimal impact, 10 = maximal impact). The list of workplace factors used was comprised of ED stressors previously identified by Whitley et al.¹⁸ Respondents were also required to report, descriptively, those aspects of their work that they find most stressful. Other Likert scales were employed to determine how work impacted on various aspects of life (physical and emotional health, family and social life and quality of work). The questionnaire was trialed for readability and revised prior to the study.

We attempted to survey all ACEM Fellows. This was logistically feasible and was thought to provide more representative results than population sampling. Hence, no sample size calculation was performed. Most results are presented descriptively. Wilcoxon signed ranks test was used to examine changes in physician work practices over time. In these analyses, all past and projected working hours were compared to present working hours. As numerous comparisons were made, the level of statistical significance was set at 0.01. All data analysis was performed using SPSS for Windows (version 11.5).

Results

Of the 510 Fellows surveyed, 323 returned completed, usable questionnaires (response rate 63.5%). The demographics of these respondents are described in Table 1. The majority were male, relatively young and married. Half of all respondents were ‘junior’ Fellows (i.e. ≤ 5 years since obtaining fellowship), an indication of the considerable recent growth in Fellow numbers. The majority of respondents were employed as staff specialists in public, city EDs. The large majority provided on-call cover at night with almost one half having on-call responsibilities more than five times per month. Few respondents reported working night shift. For the demographic variables examined, the respondents were similar to the total Fellow group at March 2003 (76.5%vs 78.9% males, 40.3 vs 41.3 years mean age).

The past, present and projected mean number of hours worked in various disciplines are described in Table 2. In this analysis, respondents were separated into ‘junior’ (≤ 5 years of fellowship) and ‘senior’ (more than 5 years of fellowship) EP groups. This separation was undertaken in order to minimize measurement bias that otherwise would have impacted upon the results. Clearly, many junior respondents would have been in training 5 years prior to the survey, with work and study practices that would have skewed the results. The point of separation provided almost equal group sizes and was consistent with the time periods of the information sought.

The results demonstrate that both groups have significantly reduced their clinical loads in the 5 years prior to the survey. Furthermore, both groups plan to significantly decrease these loads in the future. Notably, there is consistency between the two groups with a phase shift of 5 years between them. For example, the junior EP present, 5 and 10 year projected clinical loads are very similar to the senior EP past, present and 5 year projected loads, respectively. Importantly, the senior EP 10 year projected load (approximately 18 h) equates approximately to what some departments would consider as half a full clinical load.

For both groups, the past and projected total number of hours worked follow a similar pattern to that of clinical hours worked. The junior EP group report significant decreases in total hours with each successive time period. However, while the total hours of the senior EP has not decreased significantly over the past 5 years, this group is planning to decrease their total hours significantly in the future. Notably, the senior EP still aim to work a full (40 h) week 10 years from now.

Some important patterns are apparent among the other total work components. Both groups report significant increases in administrative responsibility over the last 5 years although these are expected to change little in the future. Other changes within the junior EP group, including significant increases in committee work and teaching and a significant decrease in personal education, are consistent with their recent completion of training. Although both groups expect significant increases in their research activity in the future, the absolute increase in the number of research hours per week (less than two) is small. Finally, the commitment to prehospital and retrieval work has, and is expected, to remain constant.

The reasons for the respondents (both groups combined) decreasing or planning to decrease their clinical workload are described in Table 3. It is clear that many factors impact considerably upon this decision especially excess total workload, the effect on family life, night shift, burnout and stress. Importantly, the threat of litigation, violence, lack of support, role ambiguity and co-worker conflict impact to lesser degrees.

Table 4 describes the perceived impact of work on various aspects of the respondents’ lives. While this question related to all work undertaken, Table 2 indicates clearly that clinical work is the largest component of total work and is therefore likely to be the major contributor to any perceived effect of work on the respondents’ lives. The results indicate that family life is impacted most by work. This is consistent with family life effects featuring highly among the respondents’ reasons for reducing clinical workload.

Table 5 describes the most stressful aspects of work as perceived by the respondents. Importantly, access block and dealing with management and other departments were reported by approximately one quarter of respondents. Fewer respondents reported insufficient staffing, workload, supervision of junior staff, clerical and administrative demands. Notably, only six respondents reported fear of litigation. This table is a compilation of descriptive responses that have been categorized into relatively narrow stressor groups in order to best describe the respondents’ perceptions. Accordingly, many stressor groups are not mutually exclusive. For example, access block clearly impacts upon overcrowding which, in turn and in combination with many other stressors, impacts upon workload.

Discussion

It is clear that EPs are planning to reduce their clinical workload over the next 5 and 10 years. This is consistent with the findings of Hall and Wakeman who followed the proportion of time spent in clinical work by EP graduates in the US. They reported a decrease from 86% in the first year after graduation to 60% by the fifteenth year.⁷ There is a clear difference, however, between planning a change in work practice and actually effecting such a change. This study has revealed a highly significant decrease in the clinical workload of the senior EP over the last 5 years. However, it is not possible to know if these decreases will be sustained in the future. Although decreases are clearly planned, it is conceivable that financial commitments and/or staffing difficulties may preclude continuing decreases for some. Presently, the EP workforce in Australasia is in flux with increasing numbers of trainees and EP positions, and an expanding role for EPs generally. The effect of future stabilization of this workforce on clinical responsibilities and workload remains to be seen. If, however, the planned decreases are realized, the workforce provided by today's EPs will diminish in size. This has implications for EP training, as new graduates will be required to accommodate both the reduced clinical workload of senior EPs and the likely continued expansion of alternative clinical opportunities.

The past and planned decreases in total workload are not as large as those for clinical workload. This indicates that other responsibilities are, at least partially, replacing clinical workload. The most notable change has been an increase in administrative responsibilities for both EP groups over the last 5 years. This is also consistent with Hall and Wakeman's findings. They reported an increase from 5% to 25% of time spent in administration over a 15-year period.⁷ In our study, committee work and research have also increased in both groups, albeit to lesser extents. As expected, the greatest changes have occurred within the junior physician group as many of them progressed from trainee to EP status. The planned changes in other responsibilities are small.

Many factors appear to impact upon the decision to decrease clinical workload, especially excess workload, effects on family life and emotional health, shift work and work stress. These important factors are cited commonly in similar studies from the UK and US.^1,9,11–16 Notably, the threats of litigation and physical violence received relatively low ratings in our study, in contrast to the findings of some US studies.^12,14

The finding that family life and emotional health were both rated highly as reasons for reducing clinical workload indicates the perceived importance of these factors among the respondents. This could be considered as a healthy perception. This finding is consistent with the perception that work impacts most upon these two factors.

Importantly, reasons for decreasing clinical workload do not appear to derive from the nature of the work itself. This study provides little evidence that the doctor/patient interaction, with all its attendant responsibilities, is a source of significant concern for the respondents. Indeed, factors most directly related to this interaction (litigation and violence threat, concerns regarding patient satisfaction, decision making and clinical expectations) do not feature prominently in either the reasons for decreasing clinical work or the perceived stressful aspects of work.

Alternatively, it is clear that most dissatisfaction derives from ‘systematic’ workplace difficulties. Overall, two themes are apparent among the results — excessive workload (access block, workload and overcrowding) and lack of resources (insufficient staffing, supervision of junior staff/adequate numbers of senior staff, clerical/administrative demands and lack of resources). Clearly, these two factors are interlinked. This finding may be of little surprise to most EPs. However, its importance lies in the fact that these factors, unlike many of those associated with the doctor/patient interaction, are clearly amenable to change and could be effectively minimized, given the will and political incentive.

It is noteworthy that important variables in this study are perceived workload and stress. It could be argued that doctors are now better trained, have better facilities and more back up then ever before. Despite this, the EPs still perceive heavy workloads and considerable stress. It is possible that the expectations of EPs and patients today are higher than the system can deliver. Accordingly, trainees may need to be instructed in what are considered to be reasonable expectations as well as how to manage patient expectations.

Presently, there are ample opportunities available to ACEM Fellows. Undoubtedly, clinical work is the primary role of an EP and it is recommended that substantial effort be directed to addressing those factors identified in this study as contributing most to dissatisfaction with clinical work. In the meanwhile, it cannot be concluded that the apparent reduction in clinical work with time is necessarily undesirable. Indeed, this may represent a healthy progression towards a balance of clinical and non-clinical work practice that provides for a sustainable and fulfilling practice overall.

The findings of this study have implications for the ACEM. First, the planned decrease in clinical workload is likely to impact upon the EP workforce and should be considered in determinations of workforce strategies in the future. Second, the ACEM needs to maintain its profile in political negotiations aimed at improving the ED workplace. Third, as EDs are likely to remain as relatively difficult environments in which to practice, consideration should be given to the development of training modules that specifically address coping skills for adverse workplace conditions.

This study has several notable limitations. The response rate, although reasonable for this type of study, may have been associated with selection bias. It is encouraging that the age and gender mix of the respondents was very similar to that of all ACEM Fellows. However, there may have been differences in work practices, plans and work perceptions between respondents and non-respondents. It is possible that some respondents falsified their responses, thereby introducing prevarication bias. However, data cleaning and review techniques revealed no obvious inconsistencies likely to have affected the study results. The Likert scales employed are relatively crude instruments for measurement of complex associations between the variables examined in this study. However, more sophisticated instruments would have increased the time required for questionnaire completion and impacted adversely upon response rate.

Conclusions

This study demonstrates that EPs have significantly decreased their clinical workload over the last 5 years and plan to continue this trend in the future. Involvement in other professional activities, especially administration, committees and research, has increased over the last 5 years. The main reasons for reducing clinical workload and the reported stressful aspects of work were numerous. However, most relate to excessive workload and lack of resources, interlinked factors that are amenable to change if appropriately addressed. The findings of this study have implications for workforce planning, high level political negotiation and lobbying, and physician training.