Until recently paediatric emergency departments (PEDs) were often staffed primarily by junior medical officers, who are often poorly trained for the task and inadequately supervised. It has been suggested that increasing consultant supervision would result in better outcomes and shorter waiting times might be expected due to the reduced need of junior staff to consult with “outside” specialists. The argument then for increased senior personnel in PEDs relies on the proposition that more consultant staff would lead to better outcomes including decreased admissions, fewer complaints and shorter waiting times. As consultant numbers increased by two and a half times at our institution in a relatively short period of time, we were able to examine this proposition.

METHODS

This retrospective observational study was performed in 2007 and examined the years 1997–2006. Princess Margaret Hospital for Children (PMH) is the only tertiary children’s hospital in Western Australia and serves a total population of approximately 2 million people. The introduction of a computerised management system in 1996 facilitated tracking trends within the department and allowed all attendances from 1997 onwards to be easily assessed.

Data on consultant and other staff numbers and hours of cover were taken from departmental records and rosters. The percentage of children admitted to hospital and waiting times were retrieved from the EDIS computer system. Complaints were taken from records kept by the director of the department. Costs of salaries and admissions were drawn from the hospital finance department.

Outcome measures

Percentages of children admitted to hospital, complaints to the department and waiting times before and after the period during which consultant staff increased were examined. We also assessed the net cost of the consultant increase measured against the savings owing to change in the ED admission rate to the hospital. Due to the nature of the study, it was noted and exempted by the hospital ethics committee.

RESULTS

The number of children presenting to the PMH ED over the study period remained essentially unchanged at around 45 000 per annum. Consultant numbers increased from 2.6 full time equivalent (FTE) staff in 2000 to 6.2 in 2004 (fig 1). Junior medical staff increased from 17 to 23, while the numbers of nursing and clerical staffing were unchanged during this period. The percentage of children admitted from the ED decreased from approximately 26–27% in 1997–2000 to 19% in 2006 (fig 1); the number of children admitted fell by over 2500 per year. Complaints to the department, which had been steady at approximately 33–35 a year between 1997 and 1999, dropped to 20 by 2006 (fig 1). General waiting times, which had been stable at approximately 1 h from 1997 to 2000, dropped to 51 min or by 15% by 2006 (fig 1).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1 Consultant numbers and the emergency department.

We were not able identify other factors or alterations in hospital policies or practices that might have explained the positive changes. The representation rate within 7 days to the department was 7.7% in 1996–2000 and rose modestly to 8.9% for 2004–2006 (fig 1), perhaps in response to the dramatic fall in admission rate.

Net cost

The net cost to the hospital is represented by the increased cost of the consultants compared with the decreased cost of the reduced number of admissions. Costs were indexed to 2005 using the total health price index.1 The number of saved admissions was estimated to be 2570 per annum. Using an average inpatient cost of A$4080 (excluding costs incurred in the ED), the saving in hospital admissions was estimated to be $A10 485 600. The increased cost of consultant cover was based on the increased number of consultants (2.6 to 6.2 FTE staff). As the yearly cost of an additional 3.6 consultants (2005) was $A1 003 490, the net saving to the hospital was over $A9.48 million.

Limitations

The uncontrolled nature of this observational study does not allow us to unequivocally link outcomes with increased consultant numbers. Ideally, a prospective randomised trial would have compared identical departments with and without increased consultant numbers. While a randomised trial is unlikely to occur, a prospective recording of events may have allowed other parameters to be examined, such as change in the number of investigations. As other hospitals are structured differently to PMH, the conclusions of our study may not necessarily apply.

DISCUSSION

This observational study supports the proposition that, in our hospital, increased consultant numbers coincided with improved outcomes and was cost effective. The more than doubling of consultant staff in 4 years at PMH coincided with a decreased percentage of children being admitted to hospital, fewer complaints to the department and shorter waiting times. During 1997–1999, the 2.6 FTE consultant staff provided cover from 9 am to 8:30 pm on weekdays and for 8 h on Sundays; there was no consultant presence on Saturdays. With increased consultant numbers this was extended in stages to cover from 8 am to 9.30 pm on weekdays with 12 h of weekend cover. As well as greater consultant presence “on the floor”, the appointment of more senior staff allowed for the provision of more formal teaching sessions, informal bedside tutorials, the development of departmental guidelines and a more formal ongoing assessment of individual junior staff.

The hospital saved $A10.48 million by reducing admissions by over 2000 patients a year, compared to the years 1997–1999. This compared favourably with the $A1.0 million needed to employ the extra 3.6 FTE consultant staff.

While it would be thought that having more senior staff should improve outcomes, the evidence to support this is hard to find in the literature. Tenner et al2 in a 2-year retrospective study measured length of stay and survival for patients in a teaching hospital paediatric intensive care unit (PICU) where there was a change in after-hours in-house coverage from junior doctors to consultant staff. They controlled for possible confounding effects by measuring the same outcomes in another PICU where patients were managed by the same consultants. No change was found in severity adjusted survival or length of stay in the control PICU, while patients at the study PICU were discharged 21 h sooner and were 2.8 times more likely to survive than under the original arrangement.

Press et al3 examined the effect on risk management of full time consultant physician coverage in a PED. In this retrospective study, data were collected from the PED of an urban teaching hospital. Malpractice cases from 1984 to 1987, when there was part-time attending physician coverage, were compared with those of 1987 through 1990, when full-time attending physician coverage was instituted. There was a 42% decrease in the number of claims filed and a 44% reduction in payments when full time coverage was instituted. While malpractice cases have not been a common feature in Australian PEDs, it is reasonable to assume that the 38% decrease in complaints to our department with increasing consultant presence reflects a similar trend.

In 1996, Sir Charles Gairdner Hospital in Perth, Australia made emergency medicine a key strategic initiative, which resulting in increased senior staffing. Jelinek et al4 and Rogers et al5 in before and after studies, showed improvements in waiting times, time to thrombolysis in acute myocardial infarction, misdiagnosed fracture rate and complaint rate. They also demonstrated significant improvements in teaching and research. They concluded that re-engineering an emergency department (ED), including increasing consultant staff, can be shown to improve the quality of care.

Seniority aside, the presence of any permanent medical staff in the PED leads to efficiencies due to their corporate knowledge of how their particular hospital and consultants work. Difficult long-standing patients are also more likely to be dealt with appropriately and in a more timely fashion if a consultant who has known them over a prolonged period is involved. It is interesting to note that the actual time when any consultant was present in the department did not increase greatly, especially when the department was traditionally most busy, at evenings and weekends. What changed was the ability of junior staff to access a senior staff member during weekdays, as well as increased audit and teaching by senior staff. Previously, when waiting times grew too long there was pressure on sole consultants to see more patients in their own right, thus decreasing their availability to junior staff. With increased senior staff, the division of tasks between consultants leaves the consultant “on the floor” to supervise junior staff exclusively. Other consultant duties, such as teaching, managing the observation ward, taking admitting calls and reviewing investigation results, are managed by a second and sometimes third consultant.

The increase in consultant numbers also allowed time for the development of initiatives such as a protocol for rapid oral rehydration of dehydrated children with gastroenteritis. While previously many of these children were admitted to an inpatient ward, 95% now are treated and discharged home directly from the ED. Also many young children with relatively simple problems, such as fractured limbs or lacerations, who in the past were admitted for a general anaesthetic due to their lack of cooperation and young age, are now managed in the ED. This is possible using ketamine sedation under the care of a consultant trained in advanced airway management.

Mention should be made of the increase in complaints and waiting times in 2002 that appears to argue against the proposition that increased consultant staff results in increased efficiencies. PMH traditionally had not suffered from “access block” whereby significant numbers of children admitted to the hospital were kept for prolonged periods in the ED waiting for inpatient beds. In the winters of 2001 and 2002 this occurred for the first time at PMH with winter weekly average waiting times exceeding 120 min, when traditionally they rarely went above 80 min. This resulted in the yearly average weekly waiting time increasing to 64 min in 2001 and then 71 min in 2002 from a very steady previous 60 min from 1997 to 2000. It should be noted that at that time the consultant staff had only increased from 2.6 to 3.2 FTE staff. The following year the department admitted 957 fewer children, allowing for access block to be less of an issue. This coincided with complaints dropping from 34 in 2001 to 21 for 2002 and the waiting time from 71 to 50 min. The dramatic drop in admissions was thought to be directly due to experienced paediatric consultants being able to safely send home many children whom less experienced junior staff would have admitted. While it is hard to speculate what would have happened if there had been no increase in consultant numbers, we feel that the positive response to access block would not have been so easily achieved with only 2.6 consultant staff within the department.

In summary, the provision of additional consultant medical staff in a paediatric emergency department over a relatively short period coincided with a decrease in the percentage of children admitted, fewer complaints to the department and shorter average waiting times. These improvements were achieved in a cost effective manner.