REVIEW ARTICLE ANZJSurg.com
Acute surgical unit: the Australasian experience Dean E. Page,* Dilshad Dooreemeah* and Dhan Thiruchelvam*† *Department of Surgery, St Vincent’s Hospital, Melbourne, Victoria, Australia †Department of Surgery, Epworth Hospital, Melbourne, Victoria, Australia
Key words acute surgical unit, ASU, emergency general surgical unit, emergency surgery, general surgery. Correspondence Mr Dhan Thiruchelvam, Department of Surgery, St Vincent’s Hospital, 41 Victoria Pde, Fitzroy, Vic. 3065, Australia. Email: [email protected] Accepted for publication 15 October 2013. doi: 10.1111/ans.12473
Abstract Background: The acute surgical unit (ASU) model of care is a new paradigm shift in the provision of emergency surgery. Clinical and non-clinical outcomes have been described after the introduction of the ASU model in Australia and New Zealand. This paper reviews and analyses the current published literature and methods of implementation of contemporary ASU models. Method: We conducted a comprehensive database search to identify all relevant published papers pertaining to the ASU. Included papers compared ASU models to emergency surgery’s traditional model of care. Relevant clinical and non-clinical end points were extracted for analysis. Results: Seven papers and two abstracts published data assessing clinical and nonclinical end points within the ASU. Four out of six studies reported a reduction in hospital length of stay. Two out of three studies showed reduction in mean time to emergency department review and two out of four studies reported a reduction in time to surgery. Additionally, four out of five studies showed a reduction in after hours operating with an ASU model. Conclusion: Trends in clinical outcomes are seen including reduced length of stay, time to emergency department assessment and surgery, supplemented by non-clinical outcomes including reduced after hours operating and the potential for increased training opportunities. The published data presents certain weaknesses and further information is required to appreciate the applicability of certain aspects of the ASU model to smaller centres.
Introduction Across Australia and New Zealand, patients requiring emergent surgical interventions are significant consumers of health resources. With an ageing Australian population,1 a snowballing pressure on health services creates considerable demand for emergency and elective surgery. This challenge is further compounded by an ageing surgical workforce and reduced on-call commitments.2 There has been a recent resurgence of interest in emergency surgery and the restructuring of surgical services in Australia.3 The acute surgical unit (ASU) model of care is a new paradigm shift in the provision of emergency general surgery and has been in existence in Australia since 2005.4 It is in stark contrast with the current and most common model of care, whereby the emergency general surgery workload is superimposed in an ‘ad hoc’ fashion on the elective list.4 This system also challenges the established notion that emergency surgery should be delayed to after-hours.5 While ASUs vary across different centres, as shown in Table 1, the core elements of an ASU model consist of (i) onsite consultant-led © 2013 Royal Australasian College of Surgeons
service with a defined period of on-call time; (ii) a rigid handover process; and (iii) separation of the emergency and elective surgery lists. Several clinical and non-clinical outcomes have been described after the implementation of an ASU across a number of institutions in Australia and New Zealand. Length of stay (LOS) and after-hours theatre time utilization were identified as the overarching outcomes at most centres. This paper reviews the current published literature on contemporary ASU models to assess their current state and help advise future implementation of such models.
Methods We conducted a comprehensive search to identify all relevant studies addressing ASUs in Australia and New Zealand using the following terms (TI = title word search): S1: TI acute or TI emergency S2: TI surgery or TI surgical or TI surgery S3: TI unit(s) or TI team(s) or TI model(s) or TI service(s) ANZ J Surg 84 (2014) 25–30
No, elective lists 70% filled and rest with ASU cases Shared with renal transplant 1830
0800–1600 24 h 24 h (may be shared between two consultants) 23 Yes 14
9 No 0
Sunshine Hospital, Melbourne†† Royal Adelaide Hospital‡‡
Yes if required 0 Auckland City Hospital, NZ¶
†Dr S. Ahmed, personal communication, 24 April 2013. ‡Dr A. Parasyn, personal communication, 27 May 2013. §Dr K. Epari, personal communication, 23 April 2013. ¶Dr T. Oregan, personal communication, 14 January 2013. ††Dr J. Choi, personal communication, 6 June 2013. ‡‡Dr C. Worthley, personal communication, 16 April 2013.
Yes, most days 0730 0730–1700
24 hrs (rotating from general surgical units) N/A M–W W–F (perm ASU surgeon only) 24 h
1600
Yes 0700–1800 F–M M–F
16 1 perm. surgeon 15–17 3 perm. surgeons 16
0700–1900
No No
1900 End of duty shift 0700 0700–1900 0800–1800 24 h F–M 24 h M–W, W–F 15 9
Nepean Hospital, NSW† Prince of Wales Hospital, NSW‡ Fremantle Hospital, WA§
0 4
Yes Bypass from other hospitals only Yes
Weekend ASU consultant roster No. of dedicated ASU beds Hospital
Table 1 Description of sample ASUs
Provision of trauma services?
Approx. no. of consultant surgeons rotating on unit
Weekday ASU consultant roster
Approx. hours consultant in hospital
Use of dedicated operating theatre?
Thiruchelvam et al.
Timing of consultant handover
26
S4: S1 and S2 and S3 S5: ‘(acute surgical unit)’ or ‘(surgical assessment unit)’ or ‘(acute care surgery)’ or ‘(acute care surgical service)’ or ‘(emergency general surgery)’ S6: S4 or S5 Initial results yielded 2576 papers from Medline, Cinahl, PubMed, Embase, AustHealth and PsychInfo. Our search method also included reviews of related topics so that reference lists could be hand searched to identify studies pertaining to ASUs that may not have been identified in our electronic searches.
Study selection Two researchers independently screened the title and abstracts of all citations identified from the searches and relevant studies were subsequently obtained in full text. Studies were included if they fulfilled all of the following criteria: (i) Australian and New Zealand studies in English; (ii) compared ASU models with traditional models for managing acute general surgical conditions; and (iii) reported objective clinical outcomes. Three researchers including a current Australian consultant general surgeon formed consensus opinion regarding the papers for inclusion in results. The most frequently presented and clinically relevant outcomes were extracted from the papers for qualitative description and analysis.
Results A total of seven published papers and two abstracts were found on the implementation of an ASU in Australia and New Zealand.4–12 All papers were published between 2009 and 2012 (Table 2).
Total hospital LOS Six studies published data on total hospital LOS before and after the implementation of an ASU (Table 3).5,8–12 Four studies showed a reduction in LOS with implementation of an ASU5,8,10,12 with two of these studies statistically significant.8,12
Time to ASU review in the emergency department (ED) Two studies and one abstract presented data on the time to ASU review in ED.7,10,11 Hsee et al.11 showed a reduction in mean time to review from 2.4 to 1.6 h (P < 0.001) while Von Conrady et al.10 showed a reduction in mean time to review from 3.17 to 2.05 h (P not presented). However, Geere et al.7 showed an increase in mean time to review with the ASU from 3.30 to 4.38 h (P < 0.0001).
Time to operation As shown in Table 4, three studies reported a reduction in time for patients to reach theatre5,9,12 with two of these statistically significant.5,12 Conversely, Geere et al.7 reported a statistically significant increase in the time to reach theatre.
Operative timing Five studies published data on in-hours and after-hours operative caseload before and after implementation of an ASU (Table 5).4–6,9,10 © 2013 Royal Australasian College of Surgeons
N/A
45.1 38.9%
N/A
44.4 28.1%
N/A N/A
271 Nepean Hospital, NSW 2012 Pepingco et al.12
48
N/A Auckland City Hospital, NZ 2012 Hsee et al.11
33
3450 Nepean Hospital, NSW 2010 Cox et al.4
6 2010 Von Conrady et al.10
36 2010 Gandy et al.9
34
1875
402
All acute general surgical patients All acute general surgical patients All acute general surgical patients Acute cholecystitis patients
N/A 47 30% 32.8 54.0% N/A N/A 50 38% 33.6 47.2% N/A 569 202 2010 2010 Geere et al. (abstract)7 Lehane et al.8
24 24
2009 Parasyn et al.6
31
Prince of Wales Hospital, NSW Nambour Hospital, Qld Prince of Wales Hospital, NSW Prince of Wales Hospital, NSW Fremantle Hospital, WA
N/A
Acute appendicitis patients
N/A N/A
N/A N/A
Biliary colic/cholecystitis patients Acute general surgical patients Acute appendicitis patients Acute cholecystitis patients 115 John Hunter Hospital, NSW 12 2009
New model Old model
27
Shackleton et al. (abstract)5
Year published Study (n = 9)
Table 2 Published studies and abstracts
Approx. duration of study (months)
Hospital
n
Study group
Demographics age, %male
Acute surgical unit
© 2013 Royal Australasian College of Surgeons
Hsee et al.11 was excluded due to insufficient data presented. Two studies showed a statistically significant reduction in operating between 2400 and 0800.6,9 Cox et al.4 and Von Conrady et al.10 showed a trend towards reduced after-hours operating, while Shackleton et al.5 also showed an increase of 10.5% of in-hours operating.
Complication rate Three studies reported data on complication rates.8,9,12 Lehane et al.8 showed a reduction from 17.2% to 8.7% (P = 0.003) and Gandy et al.9 showed a reduction from 17.0% to 9.3% (P = 0.02). Pepingco et al.12 failed to show a statistically significant difference.
Discussion Key clinical outcomes Patient LOS is a key indicator for success of an ASU model and central to addressing some of the prevailing issues faced by hospitals including increasing emergency surgical demands and increasing hospital costs.5,8–12 Four studies showed a reduction in LOS (absolute reductions ranging from 0.83 to 2 days)5,8,10,12 with two studies showing statistical significance.8,12 These four papers would support a reduced hospital LOS for acute general surgical patients with the implementation of an ASU. This reduction in LOS may reflect the improved patient flow through the surgical pathway in several areas including reduced time to assessment in ED and reduced waiting time for surgery. Two papers demonstrated a reduction in time to ED assessment with the implementation of an ASU10,11 with one of these statistically significant.11 Early assessment to identify surgical candidates, in addition to consultant organization of dedicated ASU surgical lists allows for efficient operative list planning and reduced cancellations and disruptions to elective surgical lists. Additionally, two papers showed a statistically significant reduction in time to surgery with the ASU5,12 with mean and median absolute reductions in time of 19.7 and 84 h, respectively. Reduced waiting time for a definitive procedure shortens overall LOS. Both early ED assessment and quicker time to surgery likely reflect the presence of a consultant led team without the competing interests of elective surgical lists and clinics. The consultant presence at referrals may also allow for more accurate and timely decisions on surgery versus conservative management and can help prevent unnecessary admissions by less experienced registrars.6,9 The efficacy of an ASU model also depends on low complication rates. Two papers showed a statistically significant reduction in postoperative complications with the ASU.8,9 The timing of the complications was absent. Despite small numbers, a reduction in both operative and general complications may reduce the need for ongoing ward treatment and LOS. The reduction in complications may reflect the increased consultant presence in theatre in both studies and the increased use of laparoscopy (shown to reduce wound infections and pain scores13). However, not all data supports a reduced LOS, time to ED assessment and time to theatre with the implementation of an ASU. Gandy et al.9 showed a non-statistically significant increase in median LOS
28
Thiruchelvam et al.
Table 3 Total hospital LOS Study (n = 6)
Year
Shackleton et al. (abstract)5† Lehane et al.8 Gandy et al.9 Von Conrady et al.10† Hsee et al.11†
2009 2010 2010 2010 2012
Pepingco et al.12
2012
Median days
P-value
Old model
ASU
5.07 6 2 4.2 34.58 h
4.24 4 3 3 34.88 h
4.9
N/A P = 0.001 P = 0.92 N/A Reported not significant in text P = 0.002
4.0
†Mean reported.
Table 4 Time to operation Study (n = 4)
Year
Shackleton et al. (abstract)5† Geere et al. (abstract)7‡ Gandy et al.9 Pepingco et al.12
Median hours ED to surgery
2009 2010 2010 2012
P-value
Old model
ASU
54.5 9.95 10.45 134.4 (to definitive procedure)
34.8 13.21 9.6 50.4 (to definitive procedure)
P P P P
< 005 < 0.0001 = 0.29 < 0.001
†Mean time from admission. ‡Mean time from assessment.
Table 5 Operative timing Study (n = 5)
Year
Shackleton et al.5 Parasyn et al.6
2009 2009
Cox et al.4
2010
Gandy et al.9 Von Conrady et al.10
2010 2010
Hours described
% of Operations
In hours 2400–0800 Emergency cases 0800–1800 1800–2400 2400–0800 2400–0800 1801–0659
from 2 to 3 days for appendicectomy and Hsee et al.11 reported a non-significant difference of 0.3 h. The increased LOS in the study by Gandy et al.9 may reflect a delay in surgery for cases presenting afterhours with a reduction in operating between 2400 and 0800, down from 26.1% to 15% (P = 0.006). Geere et al. support this point as a potential reason for the statistically significant increase in time to operation in their study.14
Non-clinical outcomes The ASU model is effective in reducing after hours operating with Gandy et al.9 and Parasyn et al.6 showing a statistically significant reduction in caseload between 2400 and 0800. Additionally, three other papers show a trend in reduced after-hours operating or increased in-hours operating.4,5,10 This data is in line with the Royal Australasian College of Surgeons’ position statement on safe operating hours, on call duties and impact on surgeons’ well-being.15 This reduction in after hours operating is multifactorial. Firstly, with an ageing surgical workforce2 and less acceptance of the demands of acute surgery on lifestyle and family, there may be less tolerance of operating after hours when not for preservation of life or
P-value
Old model
ASU
60 6.2 64.5 26.8 14.7 26.1 28
70.5 4.6 71.7 17.7–20.8 9.3–9.4 15.0 14
N/A P = 0.0074 P < 0.0001 N/A N/A P = 0.006 N/A
limb. Additionally, evidence showing that delaying appropriate cases (e.g. non-perforated appendicitis) to ‘in-hours’ operating does not contribute to morbidity strengthens this mindset.16 Lastly, efficiently planning the dedicated ASU theatre time minimizes the overflow of surgical cases from daytime elective lists to after hours. There are important training opportunities with the implementation of an ASU. General surgical registrars are typically assigned to the unit full time for a 6-month period, the number and seniority of registrars varying across units. Cox et al.4 emphasize the dense surgical exposure with an average of 180 major cases as primary surgeon and a further 123 as assistant during a 6-month ASU registrar rotation. No comparative data is available on emergency operating exposure on traditional, non-ASU units nor the impact the establishment of an ASU has on surgical exposure on other specialty general surgical units. Additionally, the increased consultant presence in theatre (14.5% Gandy et al.9 and 17.6% Lehane et al.8) may provide improved teaching and mentoring opportunities in emergency operating. However, daytime presence of consultants may limit operative exposure for training surgeons when faced with challenging cases. © 2013 Royal Australasian College of Surgeons
Acute surgical unit
The consultant led nature of the ASU model will also directly impact upon the practice of general surgical consultants. The traditional ‘ownership’ of a patient’s care from admission to discharge must be replaced by the ability to handover care once duty on the unit is complete, a concept that may need a period of adjustment. The ASU model also requires consultant surgeons to be dedicated to the unit during rostered hours. The predictable hours can allow planning of extracurricular engagements without disturbance on days off the unit. This promise of reduced after-hours emergent surgical cases and complete handover of patient care may appeal to some consultant demographics. An additional challenge faced by consultants affiliated with an ASU is the potential difficulty dealing with a range of surgical presentations outside their sub-specialization. This increasing trend of sub-specialization across all specialties may deter surgeons from participating in emergency surgery rosters and may place demand on broadly skilled surgeons, creating unsustainable workloads and rosters. Minimal objective data has been published on senior staff satisfaction within an ASU model. We are currently undertaking a survey of Australian consultant general surgeons to appreciate the impact of the ASU on professional practice, personal satisfaction and work– life balance. The results of this survey will be published in a following paper.
General limitations of studies The studies are conducted longitudinally and do not allow for direct comparison of populations. Differences in time to theatre or LOS may reflect changes in mentality towards earlier intervention and quicker discharge rather than direct effects of the new ASU model.8 Enhanced recovery after surgery or ‘fast-track’ surgery pathways have recently been implemented and shown to accelerate recovery and reduce LOS of patients post-operatively.17 Additionally, although early ED review is highlighted as a potential advantage of the ASU model, few studies outline the percentage of consultants present at this review nor indicate if or when consultants first assessed the patient. Only three studies present demographic information;8,9,12 a difference in case age and gender may be confounding variables. Limited information is also available regarding severity of disease and cases managed conservatively. Several studies have small numbers with three studies5,8,12 having fewer than 100 patients in each study arm. This perhaps reflects short study duration (ASU period 12–24 months). Furthermore, many of the studies fail to produce P-values despite showing trends in data, with only three of the six studies presenting data on LOS doing so.8,9,12 The small study groups, short duration of studies and failure to indicate statistical significance make extrapolating the data as proof of success of the ASU model difficult. Furthermore, reducing LOS can only be seen as successful if longer term complication rates and re-presentations (e.g. after conservative management by ASU) are minimized. This data is lacking in the studies presented. Three studies published data on complication rates; however, the population groups are small and despite statistical significance, these studies do not have significant statistical power.8,9,12 © 2013 Royal Australasian College of Surgeons
29
Variations in ASU models and the implementation of an ASU There are notable similarities and differences in the operational configuration of various ASUs described (Table 1). Emergency surgery models that best suit a hospital’s role and its emergency surgery volume need to be implemented to achieve the goals of an ASU. Underpinning the design of an ASU model is the determination of a hospital’s surgical volume, case complexity and urgency.18 This patient load will ultimately be influenced by local referral patterns and geographical catchment areas. These numbers help determine an appropriate model for theatre access and number of dedicated ASU beds required. Theatre access may vary from dedicated full-day ASU operating lists at larger tertiary hospitals (e.g. Fremantle Hospital, Royal Adelaide Hospital) to twice weekly half day lists (e.g. Wagga Wagga Hospital).19 Additionally, it may be viable to have several dedicated ASU beds, numbers varying considerably from centre to centre (Table 1). These beds may allow increased ED throughput as ‘bed block’ problems are reduced and more timely assessment of patients and reduced ward round times. Determining the case mix numbers and requirements for out of hours emergency operating allows for appropriate staff allocation and consultant rostering. Hospitals with predictably larger referral numbers may benefit from 24 h consultant rotations rather than three day or weekly rotations as seen at other hospitals (Table 1). The weekend consultant ASU rosters will also vary. Ultimately the duration and length of time rostered on the unit will reflect the demands of the rotation (e.g. number and complexity of cases) and availability of surgeons. Indeed, the presence of a permanent ASU surgeon allows for fewer rotations on the unit from subspecialist general surgeons. A thorough and standardized handover system at structured times daily is central to the success of an ASU. The consultant to consultant handover system has the benefits of an indirect peer review process and allows complex cases to be managed in a timely fashion with either early referral to a subspecialist unit or ongoing management under the ASU team. Our review shows that a tertiary referral hospital with a large emergency general surgery load can support the establishment of an ASU easily. Hospitals with a lesser load of emergency surgery may incorporate some components of the ASU model without the complete establishment of an ASU depending on number and complexity of cases and resources available including theatre space and number and type of surgeons.
Costs None of the included studies discussed the financial impacts of implementing an ASU model of care. However, initial data from a Department of Health report20 evaluating five consultant-led models of emergency general surgery in Victoria, Australia showed gains of AUD $590 000 per annum across all sites when applying standard costs. Most of the financial benefits resulted from reduced LOS, amounting to AUD $430 000 per annum while minor gains were observed from decreased time spent in ED (AUD $80 000) and fewer after-hours operating cases (AUD $85 000). Ultimately, the report concluded, after taking into consideration the additional costs
30
of implementing new ASU models of care, that the financial outcome was a break-even.20 An additional financial implication may be seen in the study by Pepingco et al.12 who reviewed patients presenting with cholecystitis managed via an ASU. This study showed that more patients admitted to the ASU received surgery on index admission. Although this may reflect a stronger tendency to operate with the ASU, this can prevent recurrent cholecystitis, thus reducing readmission costs and use of hospital resources.
Thiruchelvam et al.
6. 7.
8.
Conclusion Emergency general surgery comprises a significant component of general surgical workload. Australian health care services are being challenged to increase emergency patient throughput whilst maintaining excellent standards of care. The ASU model aims to address these issues. Seven papers and two abstracts have published objective data assessing clinical and non-clinical end points with regards to the ASU.4–12 The results of these studies are mixed and present certain weaknesses. However, certain trends in clinical outcomes can be seen including reduced LOS, reduced time to ED assessment and reduced time to surgery. These objective outcomes supplement nonclinical outcomes including reduced after-hours operating, increased training opportunities and predictable rostered periods. The main goal of this paper is to outline the current state of practice regarding the ASU model in Australia. Future studies are required to appreciate the applicability of certain aspects of the ASU model to smaller centres with lower caseloads (e.g. types of theatre lists, rostering patterns), the cost of implementing such a model, satisfaction of consultants with the model and, importantly, patient outcomes.
References 1. Australian Bureau of Statistics (ABS), Government of Australia. Australian Demographic Statistics, June 2009. [Cited 12 Feb 2013.] Available from URL: http://www.abs.gov.au/ausstats/[email protected]/mf/3101.0 2. Royal Australasian College of Surgeons. The Surgical Workforce. [Cited 12 Feb 2013.] Available from URL: http://www.surgeons.org/ government/workforce-and-activities-reports/#census 3. General Surgeons Australia. 12 Point Plan for Emergency General Surgery. [Cited 12 Feb 2013.] Available from URL: http://www .generalsurgeons.com.au/news/108/12-point-plan 4. Cox MR, Cook L, Dobson J, Lambrakis P, Ganesh S, Cregan P. Acute surgical unit: a new model of care. ANZ J. Surg. 2010; 80: 419–24. 5. Shackleton SD, O’neill CJ, Shah K, Wills VL. The provision of a dedicated acute general surgical unit improves the management of patients with acute gall bladder disease [abstract]. Abstract no GS22. In special issue: abstracts from the Royal Australasian College of Surgeons
9.
10. 11.
12.
13.
14. 15.
16. 17.
18.
19.
20.
Annual Scientific Congress, 6–9 May 2009, Brisbane, Queensland, Australia. ANZ J. Surg. 2009; 79: A29–A. Parasyn AD, Truskett PG, Bennett M et al. Acute-care surgical service: a change in culture. ANZ J. Surg. 2009; 79: 12–8. Geere SL, GD A, Aseervatham RL. The effectiveness of the surgical emergency team at Nambour Hospital in surgical assessment time and time to operation for acute appendicitis [abstract]. Abstract No. GS31P. In: Special Issue. Abstracts of the Royal Australasian College of Surgeons Annual Scientific Congress, 4–7 May 2010, Perth, Western Australia. ANZ J. Surg. 2010; 80: A36–7. Lehane CW, Jootun RN, Bennett M, Wong S, Truskett P. Does an acute care surgical model improve the management and outcome of acute cholecystitis? ANZ J. Surg. 2010; 80: 438–42. Gandy R, Truskett PG, Wong SW, Smith S, Bennett MH, Parasyn AD. Outcomes of appendicectomy in an acute care surgery model. Med. J. Aust. 2010; 193: 281–4. Von Conrady D, Hamza S, Weber D et al. The acute surgical unit: improving emergency care. ANZ J. Surg. 2010; 80: 933–6. Hsee L, Devaud M, Middelberg L, Jones W, Civil I. Acute surgical unit at Auckland City Hospital: a descriptive analysis. ANZ J. Surg. 2012; 82: 588–91. Pepingco L, Eslick GD, Cox MR. The acute surgical unit as a novel model of care for patients presenting with acute cholecystitis. Med. J. Aust. 2012; 196: 509–10. Chung RS, Rowland DY, Li P, Diaz J. A meta-analysis of randomized controlled trials of laparoscopic versus conventional appendectomy. Am. J. Surg. 1999; 177: 250–6. Geere SL, Aseervatham R, Grieve D. Outcomes in appendicectomy in an acute care surgery model. Med. J. Aust. [letter]. 2011; 194: 373–4. Royal Australasian College of Surgeons. Position Statement on Standards for Safe Working Hours and Conditions for Fellows, Surgical Trainees and International Medical Graduates. [Cited 12 Feb 2013.] Available from URL: http://www.surgeons.org/media/312975/position %20paper%20standards%20for%20safe%20working%20hours.pdf Eldar S, Nash E, Sabo E et al. Delay of surgery in acute appendicitis. Am. J. Surg. 1997; 173: 194–8. Varadhan KK, Neal KR, Dejong CHC, Fearon KCH, Ljungqvist O, Lobo DN. The enhanced recovery after surgery (ERAS) pathway for patients undergoing major elective open colorectal surgery: a metaanalysis of randomized controlled trials. Clinical Nutrition. 2010; 29: 434–40. Deane SA, MacLellan DG, Meredith GL, Cregan PC. Making sense of emergency surgery in New South Wales: a position statement. ANZ J. Surg. 2010; 80: 139–44. Meredith G. Establishing an Acute Surgery Unit: Coast and Country. Presentation at Department of Health Emergency Surgery Forum, 22 November 2012, Melbourne, Australia. Department of Health, Government of Victoria. Emergency General Surgery Pilot Project Evaluation. [Cited 14 Nov 2012.] Available from URL: http://docs.health.vic.gov.au/docs/doc/The-Emergency-General -Surgery-Pilot-Project-Evaluation–Executive-Summary
© 2013 Royal Australasian College of Surgeons
Acute surgical unit: the Australasian experience Dean E. Page,* Dilshad Dooreemeah* and Dhan Thiruchelvam*† *Department of Surgery, St Vincent’s Hospital, Melbourne, Victoria, Australia †Department of Surgery, Epworth Hospital, Melbourne, Victoria, Australia
Key words acute surgical unit, ASU, emergency general surgical unit, emergency surgery, general surgery. Correspondence Mr Dhan Thiruchelvam, Department of Surgery, St Vincent’s Hospital, 41 Victoria Pde, Fitzroy, Vic. 3065, Australia. Email: [email protected] Accepted for publication 15 October 2013. doi: 10.1111/ans.12473
Abstract Background: The acute surgical unit (ASU) model of care is a new paradigm shift in the provision of emergency surgery. Clinical and non-clinical outcomes have been described after the introduction of the ASU model in Australia and New Zealand. This paper reviews and analyses the current published literature and methods of implementation of contemporary ASU models. Method: We conducted a comprehensive database search to identify all relevant published papers pertaining to the ASU. Included papers compared ASU models to emergency surgery’s traditional model of care. Relevant clinical and non-clinical end points were extracted for analysis. Results: Seven papers and two abstracts published data assessing clinical and nonclinical end points within the ASU. Four out of six studies reported a reduction in hospital length of stay. Two out of three studies showed reduction in mean time to emergency department review and two out of four studies reported a reduction in time to surgery. Additionally, four out of five studies showed a reduction in after hours operating with an ASU model. Conclusion: Trends in clinical outcomes are seen including reduced length of stay, time to emergency department assessment and surgery, supplemented by non-clinical outcomes including reduced after hours operating and the potential for increased training opportunities. The published data presents certain weaknesses and further information is required to appreciate the applicability of certain aspects of the ASU model to smaller centres.
Introduction Across Australia and New Zealand, patients requiring emergent surgical interventions are significant consumers of health resources. With an ageing Australian population,1 a snowballing pressure on health services creates considerable demand for emergency and elective surgery. This challenge is further compounded by an ageing surgical workforce and reduced on-call commitments.2 There has been a recent resurgence of interest in emergency surgery and the restructuring of surgical services in Australia.3 The acute surgical unit (ASU) model of care is a new paradigm shift in the provision of emergency general surgery and has been in existence in Australia since 2005.4 It is in stark contrast with the current and most common model of care, whereby the emergency general surgery workload is superimposed in an ‘ad hoc’ fashion on the elective list.4 This system also challenges the established notion that emergency surgery should be delayed to after-hours.5 While ASUs vary across different centres, as shown in Table 1, the core elements of an ASU model consist of (i) onsite consultant-led © 2013 Royal Australasian College of Surgeons
service with a defined period of on-call time; (ii) a rigid handover process; and (iii) separation of the emergency and elective surgery lists. Several clinical and non-clinical outcomes have been described after the implementation of an ASU across a number of institutions in Australia and New Zealand. Length of stay (LOS) and after-hours theatre time utilization were identified as the overarching outcomes at most centres. This paper reviews the current published literature on contemporary ASU models to assess their current state and help advise future implementation of such models.
Methods We conducted a comprehensive search to identify all relevant studies addressing ASUs in Australia and New Zealand using the following terms (TI = title word search): S1: TI acute or TI emergency S2: TI surgery or TI surgical or TI surgery S3: TI unit(s) or TI team(s) or TI model(s) or TI service(s) ANZ J Surg 84 (2014) 25–30
No, elective lists 70% filled and rest with ASU cases Shared with renal transplant 1830
0800–1600 24 h 24 h (may be shared between two consultants) 23 Yes 14
9 No 0
Sunshine Hospital, Melbourne†† Royal Adelaide Hospital‡‡
Yes if required 0 Auckland City Hospital, NZ¶
†Dr S. Ahmed, personal communication, 24 April 2013. ‡Dr A. Parasyn, personal communication, 27 May 2013. §Dr K. Epari, personal communication, 23 April 2013. ¶Dr T. Oregan, personal communication, 14 January 2013. ††Dr J. Choi, personal communication, 6 June 2013. ‡‡Dr C. Worthley, personal communication, 16 April 2013.
Yes, most days 0730 0730–1700
24 hrs (rotating from general surgical units) N/A M–W W–F (perm ASU surgeon only) 24 h
1600
Yes 0700–1800 F–M M–F
16 1 perm. surgeon 15–17 3 perm. surgeons 16
0700–1900
No No
1900 End of duty shift 0700 0700–1900 0800–1800 24 h F–M 24 h M–W, W–F 15 9
Nepean Hospital, NSW† Prince of Wales Hospital, NSW‡ Fremantle Hospital, WA§
0 4
Yes Bypass from other hospitals only Yes
Weekend ASU consultant roster No. of dedicated ASU beds Hospital
Table 1 Description of sample ASUs
Provision of trauma services?
Approx. no. of consultant surgeons rotating on unit
Weekday ASU consultant roster
Approx. hours consultant in hospital
Use of dedicated operating theatre?
Thiruchelvam et al.
Timing of consultant handover
26
S4: S1 and S2 and S3 S5: ‘(acute surgical unit)’ or ‘(surgical assessment unit)’ or ‘(acute care surgery)’ or ‘(acute care surgical service)’ or ‘(emergency general surgery)’ S6: S4 or S5 Initial results yielded 2576 papers from Medline, Cinahl, PubMed, Embase, AustHealth and PsychInfo. Our search method also included reviews of related topics so that reference lists could be hand searched to identify studies pertaining to ASUs that may not have been identified in our electronic searches.
Study selection Two researchers independently screened the title and abstracts of all citations identified from the searches and relevant studies were subsequently obtained in full text. Studies were included if they fulfilled all of the following criteria: (i) Australian and New Zealand studies in English; (ii) compared ASU models with traditional models for managing acute general surgical conditions; and (iii) reported objective clinical outcomes. Three researchers including a current Australian consultant general surgeon formed consensus opinion regarding the papers for inclusion in results. The most frequently presented and clinically relevant outcomes were extracted from the papers for qualitative description and analysis.
Results A total of seven published papers and two abstracts were found on the implementation of an ASU in Australia and New Zealand.4–12 All papers were published between 2009 and 2012 (Table 2).
Total hospital LOS Six studies published data on total hospital LOS before and after the implementation of an ASU (Table 3).5,8–12 Four studies showed a reduction in LOS with implementation of an ASU5,8,10,12 with two of these studies statistically significant.8,12
Time to ASU review in the emergency department (ED) Two studies and one abstract presented data on the time to ASU review in ED.7,10,11 Hsee et al.11 showed a reduction in mean time to review from 2.4 to 1.6 h (P < 0.001) while Von Conrady et al.10 showed a reduction in mean time to review from 3.17 to 2.05 h (P not presented). However, Geere et al.7 showed an increase in mean time to review with the ASU from 3.30 to 4.38 h (P < 0.0001).
Time to operation As shown in Table 4, three studies reported a reduction in time for patients to reach theatre5,9,12 with two of these statistically significant.5,12 Conversely, Geere et al.7 reported a statistically significant increase in the time to reach theatre.
Operative timing Five studies published data on in-hours and after-hours operative caseload before and after implementation of an ASU (Table 5).4–6,9,10 © 2013 Royal Australasian College of Surgeons
N/A
45.1 38.9%
N/A
44.4 28.1%
N/A N/A
271 Nepean Hospital, NSW 2012 Pepingco et al.12
48
N/A Auckland City Hospital, NZ 2012 Hsee et al.11
33
3450 Nepean Hospital, NSW 2010 Cox et al.4
6 2010 Von Conrady et al.10
36 2010 Gandy et al.9
34
1875
402
All acute general surgical patients All acute general surgical patients All acute general surgical patients Acute cholecystitis patients
N/A 47 30% 32.8 54.0% N/A N/A 50 38% 33.6 47.2% N/A 569 202 2010 2010 Geere et al. (abstract)7 Lehane et al.8
24 24
2009 Parasyn et al.6
31
Prince of Wales Hospital, NSW Nambour Hospital, Qld Prince of Wales Hospital, NSW Prince of Wales Hospital, NSW Fremantle Hospital, WA
N/A
Acute appendicitis patients
N/A N/A
N/A N/A
Biliary colic/cholecystitis patients Acute general surgical patients Acute appendicitis patients Acute cholecystitis patients 115 John Hunter Hospital, NSW 12 2009
New model Old model
27
Shackleton et al. (abstract)5
Year published Study (n = 9)
Table 2 Published studies and abstracts
Approx. duration of study (months)
Hospital
n
Study group
Demographics age, %male
Acute surgical unit
© 2013 Royal Australasian College of Surgeons
Hsee et al.11 was excluded due to insufficient data presented. Two studies showed a statistically significant reduction in operating between 2400 and 0800.6,9 Cox et al.4 and Von Conrady et al.10 showed a trend towards reduced after-hours operating, while Shackleton et al.5 also showed an increase of 10.5% of in-hours operating.
Complication rate Three studies reported data on complication rates.8,9,12 Lehane et al.8 showed a reduction from 17.2% to 8.7% (P = 0.003) and Gandy et al.9 showed a reduction from 17.0% to 9.3% (P = 0.02). Pepingco et al.12 failed to show a statistically significant difference.
Discussion Key clinical outcomes Patient LOS is a key indicator for success of an ASU model and central to addressing some of the prevailing issues faced by hospitals including increasing emergency surgical demands and increasing hospital costs.5,8–12 Four studies showed a reduction in LOS (absolute reductions ranging from 0.83 to 2 days)5,8,10,12 with two studies showing statistical significance.8,12 These four papers would support a reduced hospital LOS for acute general surgical patients with the implementation of an ASU. This reduction in LOS may reflect the improved patient flow through the surgical pathway in several areas including reduced time to assessment in ED and reduced waiting time for surgery. Two papers demonstrated a reduction in time to ED assessment with the implementation of an ASU10,11 with one of these statistically significant.11 Early assessment to identify surgical candidates, in addition to consultant organization of dedicated ASU surgical lists allows for efficient operative list planning and reduced cancellations and disruptions to elective surgical lists. Additionally, two papers showed a statistically significant reduction in time to surgery with the ASU5,12 with mean and median absolute reductions in time of 19.7 and 84 h, respectively. Reduced waiting time for a definitive procedure shortens overall LOS. Both early ED assessment and quicker time to surgery likely reflect the presence of a consultant led team without the competing interests of elective surgical lists and clinics. The consultant presence at referrals may also allow for more accurate and timely decisions on surgery versus conservative management and can help prevent unnecessary admissions by less experienced registrars.6,9 The efficacy of an ASU model also depends on low complication rates. Two papers showed a statistically significant reduction in postoperative complications with the ASU.8,9 The timing of the complications was absent. Despite small numbers, a reduction in both operative and general complications may reduce the need for ongoing ward treatment and LOS. The reduction in complications may reflect the increased consultant presence in theatre in both studies and the increased use of laparoscopy (shown to reduce wound infections and pain scores13). However, not all data supports a reduced LOS, time to ED assessment and time to theatre with the implementation of an ASU. Gandy et al.9 showed a non-statistically significant increase in median LOS
28
Thiruchelvam et al.
Table 3 Total hospital LOS Study (n = 6)
Year
Shackleton et al. (abstract)5† Lehane et al.8 Gandy et al.9 Von Conrady et al.10† Hsee et al.11†
2009 2010 2010 2010 2012
Pepingco et al.12
2012
Median days
P-value
Old model
ASU
5.07 6 2 4.2 34.58 h
4.24 4 3 3 34.88 h
4.9
N/A P = 0.001 P = 0.92 N/A Reported not significant in text P = 0.002
4.0
†Mean reported.
Table 4 Time to operation Study (n = 4)
Year
Shackleton et al. (abstract)5† Geere et al. (abstract)7‡ Gandy et al.9 Pepingco et al.12
Median hours ED to surgery
2009 2010 2010 2012
P-value
Old model
ASU
54.5 9.95 10.45 134.4 (to definitive procedure)
34.8 13.21 9.6 50.4 (to definitive procedure)
P P P P
< 005 < 0.0001 = 0.29 < 0.001
†Mean time from admission. ‡Mean time from assessment.
Table 5 Operative timing Study (n = 5)
Year
Shackleton et al.5 Parasyn et al.6
2009 2009
Cox et al.4
2010
Gandy et al.9 Von Conrady et al.10
2010 2010
Hours described
% of Operations
In hours 2400–0800 Emergency cases 0800–1800 1800–2400 2400–0800 2400–0800 1801–0659
from 2 to 3 days for appendicectomy and Hsee et al.11 reported a non-significant difference of 0.3 h. The increased LOS in the study by Gandy et al.9 may reflect a delay in surgery for cases presenting afterhours with a reduction in operating between 2400 and 0800, down from 26.1% to 15% (P = 0.006). Geere et al. support this point as a potential reason for the statistically significant increase in time to operation in their study.14
Non-clinical outcomes The ASU model is effective in reducing after hours operating with Gandy et al.9 and Parasyn et al.6 showing a statistically significant reduction in caseload between 2400 and 0800. Additionally, three other papers show a trend in reduced after-hours operating or increased in-hours operating.4,5,10 This data is in line with the Royal Australasian College of Surgeons’ position statement on safe operating hours, on call duties and impact on surgeons’ well-being.15 This reduction in after hours operating is multifactorial. Firstly, with an ageing surgical workforce2 and less acceptance of the demands of acute surgery on lifestyle and family, there may be less tolerance of operating after hours when not for preservation of life or
P-value
Old model
ASU
60 6.2 64.5 26.8 14.7 26.1 28
70.5 4.6 71.7 17.7–20.8 9.3–9.4 15.0 14
N/A P = 0.0074 P < 0.0001 N/A N/A P = 0.006 N/A
limb. Additionally, evidence showing that delaying appropriate cases (e.g. non-perforated appendicitis) to ‘in-hours’ operating does not contribute to morbidity strengthens this mindset.16 Lastly, efficiently planning the dedicated ASU theatre time minimizes the overflow of surgical cases from daytime elective lists to after hours. There are important training opportunities with the implementation of an ASU. General surgical registrars are typically assigned to the unit full time for a 6-month period, the number and seniority of registrars varying across units. Cox et al.4 emphasize the dense surgical exposure with an average of 180 major cases as primary surgeon and a further 123 as assistant during a 6-month ASU registrar rotation. No comparative data is available on emergency operating exposure on traditional, non-ASU units nor the impact the establishment of an ASU has on surgical exposure on other specialty general surgical units. Additionally, the increased consultant presence in theatre (14.5% Gandy et al.9 and 17.6% Lehane et al.8) may provide improved teaching and mentoring opportunities in emergency operating. However, daytime presence of consultants may limit operative exposure for training surgeons when faced with challenging cases. © 2013 Royal Australasian College of Surgeons
Acute surgical unit
The consultant led nature of the ASU model will also directly impact upon the practice of general surgical consultants. The traditional ‘ownership’ of a patient’s care from admission to discharge must be replaced by the ability to handover care once duty on the unit is complete, a concept that may need a period of adjustment. The ASU model also requires consultant surgeons to be dedicated to the unit during rostered hours. The predictable hours can allow planning of extracurricular engagements without disturbance on days off the unit. This promise of reduced after-hours emergent surgical cases and complete handover of patient care may appeal to some consultant demographics. An additional challenge faced by consultants affiliated with an ASU is the potential difficulty dealing with a range of surgical presentations outside their sub-specialization. This increasing trend of sub-specialization across all specialties may deter surgeons from participating in emergency surgery rosters and may place demand on broadly skilled surgeons, creating unsustainable workloads and rosters. Minimal objective data has been published on senior staff satisfaction within an ASU model. We are currently undertaking a survey of Australian consultant general surgeons to appreciate the impact of the ASU on professional practice, personal satisfaction and work– life balance. The results of this survey will be published in a following paper.
General limitations of studies The studies are conducted longitudinally and do not allow for direct comparison of populations. Differences in time to theatre or LOS may reflect changes in mentality towards earlier intervention and quicker discharge rather than direct effects of the new ASU model.8 Enhanced recovery after surgery or ‘fast-track’ surgery pathways have recently been implemented and shown to accelerate recovery and reduce LOS of patients post-operatively.17 Additionally, although early ED review is highlighted as a potential advantage of the ASU model, few studies outline the percentage of consultants present at this review nor indicate if or when consultants first assessed the patient. Only three studies present demographic information;8,9,12 a difference in case age and gender may be confounding variables. Limited information is also available regarding severity of disease and cases managed conservatively. Several studies have small numbers with three studies5,8,12 having fewer than 100 patients in each study arm. This perhaps reflects short study duration (ASU period 12–24 months). Furthermore, many of the studies fail to produce P-values despite showing trends in data, with only three of the six studies presenting data on LOS doing so.8,9,12 The small study groups, short duration of studies and failure to indicate statistical significance make extrapolating the data as proof of success of the ASU model difficult. Furthermore, reducing LOS can only be seen as successful if longer term complication rates and re-presentations (e.g. after conservative management by ASU) are minimized. This data is lacking in the studies presented. Three studies published data on complication rates; however, the population groups are small and despite statistical significance, these studies do not have significant statistical power.8,9,12 © 2013 Royal Australasian College of Surgeons
29
Variations in ASU models and the implementation of an ASU There are notable similarities and differences in the operational configuration of various ASUs described (Table 1). Emergency surgery models that best suit a hospital’s role and its emergency surgery volume need to be implemented to achieve the goals of an ASU. Underpinning the design of an ASU model is the determination of a hospital’s surgical volume, case complexity and urgency.18 This patient load will ultimately be influenced by local referral patterns and geographical catchment areas. These numbers help determine an appropriate model for theatre access and number of dedicated ASU beds required. Theatre access may vary from dedicated full-day ASU operating lists at larger tertiary hospitals (e.g. Fremantle Hospital, Royal Adelaide Hospital) to twice weekly half day lists (e.g. Wagga Wagga Hospital).19 Additionally, it may be viable to have several dedicated ASU beds, numbers varying considerably from centre to centre (Table 1). These beds may allow increased ED throughput as ‘bed block’ problems are reduced and more timely assessment of patients and reduced ward round times. Determining the case mix numbers and requirements for out of hours emergency operating allows for appropriate staff allocation and consultant rostering. Hospitals with predictably larger referral numbers may benefit from 24 h consultant rotations rather than three day or weekly rotations as seen at other hospitals (Table 1). The weekend consultant ASU rosters will also vary. Ultimately the duration and length of time rostered on the unit will reflect the demands of the rotation (e.g. number and complexity of cases) and availability of surgeons. Indeed, the presence of a permanent ASU surgeon allows for fewer rotations on the unit from subspecialist general surgeons. A thorough and standardized handover system at structured times daily is central to the success of an ASU. The consultant to consultant handover system has the benefits of an indirect peer review process and allows complex cases to be managed in a timely fashion with either early referral to a subspecialist unit or ongoing management under the ASU team. Our review shows that a tertiary referral hospital with a large emergency general surgery load can support the establishment of an ASU easily. Hospitals with a lesser load of emergency surgery may incorporate some components of the ASU model without the complete establishment of an ASU depending on number and complexity of cases and resources available including theatre space and number and type of surgeons.
Costs None of the included studies discussed the financial impacts of implementing an ASU model of care. However, initial data from a Department of Health report20 evaluating five consultant-led models of emergency general surgery in Victoria, Australia showed gains of AUD $590 000 per annum across all sites when applying standard costs. Most of the financial benefits resulted from reduced LOS, amounting to AUD $430 000 per annum while minor gains were observed from decreased time spent in ED (AUD $80 000) and fewer after-hours operating cases (AUD $85 000). Ultimately, the report concluded, after taking into consideration the additional costs
30
of implementing new ASU models of care, that the financial outcome was a break-even.20 An additional financial implication may be seen in the study by Pepingco et al.12 who reviewed patients presenting with cholecystitis managed via an ASU. This study showed that more patients admitted to the ASU received surgery on index admission. Although this may reflect a stronger tendency to operate with the ASU, this can prevent recurrent cholecystitis, thus reducing readmission costs and use of hospital resources.
Thiruchelvam et al.
6. 7.
8.
Conclusion Emergency general surgery comprises a significant component of general surgical workload. Australian health care services are being challenged to increase emergency patient throughput whilst maintaining excellent standards of care. The ASU model aims to address these issues. Seven papers and two abstracts have published objective data assessing clinical and non-clinical end points with regards to the ASU.4–12 The results of these studies are mixed and present certain weaknesses. However, certain trends in clinical outcomes can be seen including reduced LOS, reduced time to ED assessment and reduced time to surgery. These objective outcomes supplement nonclinical outcomes including reduced after-hours operating, increased training opportunities and predictable rostered periods. The main goal of this paper is to outline the current state of practice regarding the ASU model in Australia. Future studies are required to appreciate the applicability of certain aspects of the ASU model to smaller centres with lower caseloads (e.g. types of theatre lists, rostering patterns), the cost of implementing such a model, satisfaction of consultants with the model and, importantly, patient outcomes.
References 1. Australian Bureau of Statistics (ABS), Government of Australia. Australian Demographic Statistics, June 2009. [Cited 12 Feb 2013.] Available from URL: http://www.abs.gov.au/ausstats/[email protected]/mf/3101.0 2. Royal Australasian College of Surgeons. The Surgical Workforce. [Cited 12 Feb 2013.] Available from URL: http://www.surgeons.org/ government/workforce-and-activities-reports/#census 3. General Surgeons Australia. 12 Point Plan for Emergency General Surgery. [Cited 12 Feb 2013.] Available from URL: http://www .generalsurgeons.com.au/news/108/12-point-plan 4. Cox MR, Cook L, Dobson J, Lambrakis P, Ganesh S, Cregan P. Acute surgical unit: a new model of care. ANZ J. Surg. 2010; 80: 419–24. 5. Shackleton SD, O’neill CJ, Shah K, Wills VL. The provision of a dedicated acute general surgical unit improves the management of patients with acute gall bladder disease [abstract]. Abstract no GS22. In special issue: abstracts from the Royal Australasian College of Surgeons
9.
10. 11.
12.
13.
14. 15.
16. 17.
18.
19.
20.
Annual Scientific Congress, 6–9 May 2009, Brisbane, Queensland, Australia. ANZ J. Surg. 2009; 79: A29–A. Parasyn AD, Truskett PG, Bennett M et al. Acute-care surgical service: a change in culture. ANZ J. Surg. 2009; 79: 12–8. Geere SL, GD A, Aseervatham RL. The effectiveness of the surgical emergency team at Nambour Hospital in surgical assessment time and time to operation for acute appendicitis [abstract]. Abstract No. GS31P. In: Special Issue. Abstracts of the Royal Australasian College of Surgeons Annual Scientific Congress, 4–7 May 2010, Perth, Western Australia. ANZ J. Surg. 2010; 80: A36–7. Lehane CW, Jootun RN, Bennett M, Wong S, Truskett P. Does an acute care surgical model improve the management and outcome of acute cholecystitis? ANZ J. Surg. 2010; 80: 438–42. Gandy R, Truskett PG, Wong SW, Smith S, Bennett MH, Parasyn AD. Outcomes of appendicectomy in an acute care surgery model. Med. J. Aust. 2010; 193: 281–4. Von Conrady D, Hamza S, Weber D et al. The acute surgical unit: improving emergency care. ANZ J. Surg. 2010; 80: 933–6. Hsee L, Devaud M, Middelberg L, Jones W, Civil I. Acute surgical unit at Auckland City Hospital: a descriptive analysis. ANZ J. Surg. 2012; 82: 588–91. Pepingco L, Eslick GD, Cox MR. The acute surgical unit as a novel model of care for patients presenting with acute cholecystitis. Med. J. Aust. 2012; 196: 509–10. Chung RS, Rowland DY, Li P, Diaz J. A meta-analysis of randomized controlled trials of laparoscopic versus conventional appendectomy. Am. J. Surg. 1999; 177: 250–6. Geere SL, Aseervatham R, Grieve D. Outcomes in appendicectomy in an acute care surgery model. Med. J. Aust. [letter]. 2011; 194: 373–4. Royal Australasian College of Surgeons. Position Statement on Standards for Safe Working Hours and Conditions for Fellows, Surgical Trainees and International Medical Graduates. [Cited 12 Feb 2013.] Available from URL: http://www.surgeons.org/media/312975/position %20paper%20standards%20for%20safe%20working%20hours.pdf Eldar S, Nash E, Sabo E et al. Delay of surgery in acute appendicitis. Am. J. Surg. 1997; 173: 194–8. Varadhan KK, Neal KR, Dejong CHC, Fearon KCH, Ljungqvist O, Lobo DN. The enhanced recovery after surgery (ERAS) pathway for patients undergoing major elective open colorectal surgery: a metaanalysis of randomized controlled trials. Clinical Nutrition. 2010; 29: 434–40. Deane SA, MacLellan DG, Meredith GL, Cregan PC. Making sense of emergency surgery in New South Wales: a position statement. ANZ J. Surg. 2010; 80: 139–44. Meredith G. Establishing an Acute Surgery Unit: Coast and Country. Presentation at Department of Health Emergency Surgery Forum, 22 November 2012, Melbourne, Australia. Department of Health, Government of Victoria. Emergency General Surgery Pilot Project Evaluation. [Cited 14 Nov 2012.] Available from URL: http://docs.health.vic.gov.au/docs/doc/The-Emergency-General -Surgery-Pilot-Project-Evaluation–Executive-Summary
© 2013 Royal Australasian College of Surgeons
