An observational study of emergency department intern activities

Interns in most Australian states must complete an emergency medicine rotation at an accredited emergency department (ED) before they can attain full medical registration.1 In Victoria, the ED rotation is of 8 weeks’ duration. As graduates emerge from the increased medical school places offered by the Australian Government since 2004, intern numbers are expected to rise,2 and there is concern that EDs will be unable to meet future demand for intern placements. Some have proposed substituting emergency medicine rotations in accredited EDs with alternative positions with some emergency exposure (eg, through rural general practices and non-accredited EDs).2

The ED is recognised as a key training ground, providing exposure to clinical skills that are transferable across many medical fields.3,4 If substitution of the ED rotation is to be considered, it would be prudent to analyse the current ED experience of interns.

Emergency experience at undergraduate level is limited, with, for example, 2 weeks allocated in the current 6-year University of Melbourne undergraduate medical course. Limited emergency experience has been linked to reduced clinical competence and procedural confidence among medical students.5 Procedural confidence is positively correlated with exposure to procedures,6,7 whereas formal theoretical teaching that is disconnected from direct clinical cases during internship is perceived to be of limited value.8

Intern procedural training usually follows an apprenticeship model whereby junior doctors’ clinical activities are directly supervised by more senior clinicians.7,8 Time constraints of supervisors have been identified as major barriers to adequate supervision in the ED.4

There is little published information on how interns spend their time during emergency medicine terms in Australia or elsewhere. Further, the contexts in which ED activities are performed (eg, supervised or independent, face-to-face with patients) are yet to be described, but are important for understanding the nature of the ED intern rotation. We aimed to identify the range, frequency, duration and contexts of activities performed by ED interns in order to calculate the expected number of activities an intern is likely to undertake on a standard ED clinical rotation of 8 weeks working 38 hours per week (304 hours).

We undertook a descriptive, observational study between 1 August and 31 December 2006, in the EDs of three public hospitals in Victoria — St Vincent’s Hospital Melbourne (39 000 adult attendances per annum), Royal Melbourne Hospital, also a designated major trauma service (48 000 adult attendances) and Austin Hospital (45 000 adult and paediatric attendances).

All interns rostered to work in one of the three participating EDs were eligible for inclusion in the study. Sampling was based on intern rosters to incorporate every hour of every day of the week once only at each participating ED. This translated to 168 potential hours of intern activity data at each ED (504 hours total). Each data collection period was set at 8 hours, with a set maximum collective observation time of 16 hours per intern. At one ED, interns were rostered on during normal business hours only. Thus, data collection there was restricted to a third of that at other EDs to prevent excessive representation of morning shift hours.

Eligible interns were identified from ED rosters at least 2 weeks before their anticipated participation. An information statement and invitation to participate were emailed to the interns about a week before data collection, and informed consent was sought on the day of data collection.

The observation tool used for identifying tasks performed by ED interns was developed by two medical students and refined using a Delphi panel of nine doctors from one participating ED. Three feedback rounds occurred. One week of training in data collection for J N Z and another researcher, followed by a week of interrater reliability assurance testing took place in August 2006 (Cohen’s κ, 0.86; P < 0.001).

Following training, participants’ activities and respective activity contexts were observed by one researcher (J N Z), who measured the duration of these with a stopwatch. All data were recorded on a predesigned table with variable definitions verified by ED physicians. Where simultaneous activities were undertaken, only the principal activity was recorded. Interns were instructed to continue their duties as usual, and communication between the participant and researcher was minimised. All collected data were de-identified.

Data were analysed with SPSS, version 15.0 (SPSS Inc, Chicago, Ill, USA). Descriptive statistics were calculated, including median (interquartile range), minimum, maximum and overall total percentage of time spent by ED interns on activities and activity contexts. The number of activities performed was totalled, and the mean duration of tasks was calculated (total activity time/activity frequency). The anticipated number of activities performed by interns in a 304-hour 8-week ED rotation was calculated by multiplying each relevant result by the total rostered hours over 8 weeks (304 hours) divided by the total hours observed (390.8 hours).

This study was authorised by the ethics committees of each of the participating institutions.

Fifty-one interns met the inclusion criteria. Of these, 42 were rostered during data collection periods and were subsequently observed. From a possible 504 hours of potential observation hours (all the hours in each day of 1 week at each site), 390.8 hours of data were recorded because of the absence of afternoon and night shifts for interns at one ED, and about an hour of unidentified time loss. The mean continuous observation period was 7.1 hours per intern, with observations ranging from 1.0–8.4 hours per intern (Box 1). Age and sex of the participants were comparable across hospitals (mean age, 26 years; range, 23–32; 20 [46.7%] were male).

Our study shows that most ED intern time (86.6%) is spent on patient-related clinical tasks. This is comparable with an Australia-wide ED-based finding (84.1%),2 but exceeds those of United States-based studies of ED,9 surgical,10 and internal medicine11 intern rotations (58%, 40%, 30%–62%).

While the observation method used may have elicited practice changes secondary to the presence of an observer (the Hawthorne effect), covert observation of interns is unlikely to be ethical or possible. Indeed, the importance of direct observation has been previously emphasised.11

Other potential threats to the study validity were the lack of urban, provincial, or rural EDs, a limited paediatric base and a non-uniform intern roster. Also, in two of the participating EDs, interns were rostered to manage short-stay units adjacent to their departments. The three EDs may be more representative of the intern ED experience than previous single-site studies,8 but extrapolation to all Australian EDs should be made with caution.

Data were collected over 4 months despite possible yearly variation in ED casemix and clinical tasks. Interns were in the 6th to 10th month of their intern year, and hence already had some experience. Samples of interns on their first or last rotations of an intern year may be different.

Our use of a single researcher to sample intern activities without secondary verification may have led to some observer bias, but was unavoidable. Attempts were made to mitigate this through pre-observation training and inter-rater reliability checks.

A study by the Postgraduate Medical Education Council of Queensland (PMCQ) reported that ED interns across four Australian states spent 18.9% of their time taking histories and performing examinations combined,2 compared with 26.8% in our study. Patient history-taking alone consumed more time (17.5%) than that reported in earlier international studies (8%12 and 15.2%9).

Although the ED is perceived as an environment rich in procedural opportunities,6-9,13 procedures took up only 5.6% of intern time, and ranked sixth in terms of activity frequency, similar to other studies.2,9 Senior staff maybe more likely to manage the critically ill and may become distracted from involving interns in advanced procedures. Additionally, competition from other ED staff such as registrars training in emergency medicine may be an additional barrier to intern procedural exposure. Lack of procedural experience has been linked with lower confidence and competence.5,14,15

It is often perceived that intern time is wasted on so-called “mindless” tasks8 that could be more efficiently performed by support staff. However, we found that such tasks, including non-patient-related administrative activities and procedural preparation and clean-up took up less than 3% of intern time. If looking for efficiencies, greater impact maybe obtained by exploring ways of decreasing the time required for obtaining patient histories (17.5%), documentation in clinical notes (11.3%) and the other mainly communication processes required for patient management (32.6% of intern time).

Our study shows that synchronous communication modes (ie, when both individuals participate in the conversation simultaneously) accounted for almost two-thirds of total intern time, with face-to-face communication accounting for most of these. Interruptions created by immediate requests to speak are a characteristic of synchronous communication,16 the consequences of which include forgetfulness and errors.17 The ED may be construed as an event-driven environment with a high demand for immediate communication between both staff and patients. Therefore, training in the safe and efficient use of synchronous communication in the ED context may be useful, such as the recently described Situation-Background-Assessment-Recommendation technique.18

Seeking advice from more senior ED staff (6.7% of intern time) or peers (0.7%) consumed a comparable amount of intern time to that reported in the PMCQ study (9.8%).2 Lack of immediate availability of more senior doctors could result in delays in providing guidance to interns, and hence delays in patient flow. In 8 weeks, each intern would have an average of 683 consultations with more senior ED staff. Each extra intern in an ED could mean a requirement for existing ED staff to service around 20 more consultations per 10-hour shift. An increase in the number of Australian interns may reduce or dilute access to senior medical advice, or decrease the time senior staff have for their other tasks, unless accompanied by a similar increase in the number of senior medical staff. Designated educators are already available for junior US doctors9 and Australian nurses.19 This approach could be adopted for Australian ED junior medical staff.

At less than 2%, structured and incidental educational tasks accounted for substantially less intern time than that indicated by the PCMQ report (9.1%)2 and an Australian study conducted almost two decades ago (12%).8 This may reflect the interns’ working week being substantially shorter than a generation ago, with less time available for educational tasks.8

Personal or miscellaneous tasks (breaks, downtime and “other” activities) accounted for just 8.0% of total intern time, comparable to the PMCQ finding (6.9%)2 but much less than that reported in the US (20%).9 In contrast to Victorian ED interns, those in the US are often on call, and so sleeping was also considered personal time use.9 The small amount of “downtime” when there were no tasks observed (3.7%) indicates that interns are usually performing a work-related task, and that they are therefore not available to see new emergency patients about 96% of the time. Other emergency services (fire, ambulance) aim for a much higher percentage of downtime to ensure their availability for emergencies. Notwithstanding the supervision requirements, this low amount of downtime implies that there could be capacity within EDs for more interns, as it is likely that there would be patients available to be seen; however, further analysis looking at all factors involved would be needed to determine this.

Breaks were only 27 minutes on average per day, including a 17-minute meal break. This is just over half of the meal break entitlement for shifts exceeding five consecutive hours according to Victorian industrial relations law.20 Although EDs had the safest working hours of all hospital departments according to the Australian Medical Association’s Safe Hours Audit in 2006,21 departmental heads may need to consider rostering 30-minute meal breaks to meet industrial requirements.

While not directly recorded in our study, another Australian study has observed that registrars provide the bulk of intern supervision.22 It has been suggested that reliance on “the voluntary contributions of many clinicians who teach” is unsustainable.23 The recognition by a recent Productivity Commission report that quality procedural supervision and education take time and appropriate resources is important.24

The ED offers a unique environment for interns to undertake a comprehensive range of clinical and procedural activities. The expected increase in intern numbers should not precipitate hasty decisions to devise alternatives to the ED rotation without further assessing the impact on EDs of hosting more interns, or the effect on interns of being deprived of their ED experience. Currently, there appears to be no shortage of patients available to be seen in the ED by interns. It is likely that ED attendances will continue to increase, providing more exposure to clinical opportunities for interns and more need for ED manpower. To maintain the positive educational and clinical outcomes for interns in the ED, sufficient preparation and supervision of the increased numbers of interns will be required. An analysis of EDs across Australia in terms of their ability to accept increasing intern rotations may be beneficial.

4 Percentage of total time spent on each activity by emergency department interns