INTRODUCTION
Traumatic injuries place a high toll on the lives and pocketbooks of Canadians. Within British Columbia (BC), one-half million ED visits, 35,000 hospitalizations, and 2000 deaths were attributed to injury in 2010 at an estimated cost of $3.7 billion.Reference Rajabali, Ibrahimova, Barnett and Pike 1
Modern trauma systems comprise several hospitals stratified by resource availability and linked through field triage and inter-facility transfer criteria, all designed to match injured patients to the best facility able to meet their needs. Accreditation Canada is responsible for credentialing trauma centres across the country. Hospitals are assigned a Level I through V based on the resource and subspecialty availability at each site. The provincial adult trauma system of BC is built around ten hospitals: two Level I, three Level II, and five Level III sites. Level I and II sites provide definitive care, and lower level centres initiate injury stabilization prior to transfer. 2 Although the accreditation level of a hospital dictates the scope of treatment available at a particular site, many factors influence the quality of patient care. The “model of care,” the manner by which patients are admitted, overseen, and handed over, is an important determinant of good patient outcomes. A recent review attributed 34% of patient safety–related in-hospital deaths to deficiencies attributable to models of care, including inadequate handover, poor patient observation, and failure to recognize patient deterioration.Reference Donaldson, Panesar and Darzi 3
A variety of models of care for admitted trauma patients have been described in medical literature. At larger centres, a single trauma service would generally oversee all injured patients, and at smaller sites, oversight is provided by a variety of surgical and hospitalist services depending on injury pattern.Reference Simons, Eliopoulos, Laflamme and Brown 4 - Reference Salottolo, Slone and Howell 7 However, although national accreditation guidelines endorse this distribution of care models, the real-world practice remains unknown; available existing literature has examined only individual, almost exclusively Level I trauma hospitals. Given the influence that the model of patient oversight has upon the management of injured patients, we sought to describe the various models of care currently operating within our provincial trauma system. To determine the ideal model of care for each of our variably sized centres, we explored the sustainability of each model and investigated its compatibility with the standards recommended for each trauma accreditation level.
METHODS
Prior to this study, meetings among stakeholders at trauma centres in BC revealed an inadequate understanding of the variety of models of care in existence. A study group was convened, and a questionnaire was developed in consultation with regional and provincial trauma medical directors. The study protocol was approved by the Fraser Health Information Privacy and Security Office. Informed consent was obtained from all study participants. In November 2015, an electronic survey was distributed by email to the medical directors of trauma at ten receiving adult trauma centres in BC. The 39-item questionnaire explored characteristics of models of care for injured patients, including the specialty and composition of the admitting team, service intervals, and additional clinical responsibilities. The survey scrutinized determinants of the existing model and barriers to its overall sustainability, specifically asking if site directors thought their trauma program was sustainable for >5 years in its current form. The model of care for each care was examined alongside the relevant criteria of the 2014 Accreditation Canada Trauma Distinction Centre Standards to establish guideline compliance. 2 Follow-up telephone interviews with respondents were built upon survey results to clarify responses.
RESULTS
All ten sites returned completed surveys and responded to telephone follow-ups. With one exception, the oversight of admitted patients at these centres conformed to one of two models of care that are hereafter referred to as the “admitting trauma service” and “on-call consultant” models. The third novel model was labelled the “short-stay trauma unit” (Table 1).
General specifications of the main models of care in practice in BC trauma centres
ED=emergency department; EP=emergency physician; ICU=intensive care unit.
The “admitting trauma service” was an admitting service led by an emergency physician, a general surgeon, or a trauma surgeon. All these services employed a full-time trauma nurse to assist with the patient care process, and all but one rotated residents or fellows through the service. Except for time spent within a closed intensive care unit, this team remained the admitting and most responsible service from the time of consultation or “activation” (rapid in-person consultation at any time, triggered by anatomic, mechanistic, or physiologic injury criteria) in the emergency department (ED) until the patient had progressed beyond the acutely injured period and was discharged from the hospital or transferred to a hospitalist service. The service interval for the attending trauma physician lasted one-half to 1 week, during which time their clinical responsibilities were restricted to the daily care of admitted patients, re-assessing recently discharged individuals at once-weekly trauma clinics, and assessing new admissions during the daytime and a portion of the evening hours. The attending physician’s extraneous clinical responsibilities (associated with their usual ED or elective surgical practices) were minimized to the greatest extent possible to ensure prompt, attentive trauma care.
The “on-call consultant” model comprised an emergency physician or a general practitioner available to provide supplementary care for trauma patients in the ED. On-call consultants were “called in” at any time and at the discretion of the attending emergency physician. Common scenarios for call in included trauma “activation,” multiple casualties, pre-arrival paramedic notification of major trauma, or the anticipated need for multiple ED procedures. The on-call consultant attended to the injured patients only while they were in the ED until they were transferred to a higher level of care or admitted by another service. These other services could be the intensivist for the severely injured patient, a surgical service overseeing the most significantly injured organ system, or the hospitalist for all non-surgical patients. The on-call consultant service interval spanned 12–24 hours, during which time additional clinical duties were at least tolerated or at the quieter centres even encouraged. Examples of these responsibilities included conducting one’s personal office clinic or serving as an attending physician in a different ED.
The “short-stay trauma unit” model employed emergency physicians or general practitioners who were available to provide supplementary care in the ED in the same fashion as the on-call consultant model. However, these physicians also attended to a short-stay trauma unit, which admitted a cohort of injured patients whose hospital length of stay was estimated to be ≤48 hours. Similar to the on-call consultant model, various other services would admit from the ED those trauma patients with longer anticipated stays, as well as assuming care at the 48-hour mark for those patients whose hospital discharge had been unexpectedly delayed. Trauma physicians at this site remained on-call for 24 hours, during which time they were expected to round on those patients admitted to the short-stay trauma unit, as well as the adjacent observation unit (which housed uncomplicated medical patients requiring an admission of ≤48 hours). Additional non-trauma, non-observation unit clinical responsibilities were discouraged while on-call.
The distribution of the three models by trauma level designation was as follows: all Level I trauma centres utilized the admitting trauma service model (2/2), and all Level II centres employed an on-call consultant (3/3). Level III centres relied upon the admitting trauma service (2/5), on-call consultant (2/5), and short-stay trauma unit models (1/5) (Table 2). If arranged by annual BC Trauma Registry case volume (injured patients aged >15 years who required ≥48 hours of hospitalization or dying as a result of their injuries), we discovered that sites with >1000 cases per year all utilized admitting trauma services (2/2), those with 500–1000 cases per year employed both admitting trauma services (1/4) and on-call consultants (3/4), and those with <500 cases per year operated all three models of care: admitting trauma service (1/4), on-call consultant (2/4), or short-stay trauma unit (1/4) (Table 3).Reference Gezer, Thakore, McMillan and Robinson 8
Distribution of trauma models of care within different levels of trauma centres
Distribution of trauma models of care separated by patient load
Only one-half (5/10) of the medical directors of trauma centres in BC believed that their current model of care was sustainable in its existing form, with significant discordance depending on the type of model currently in situ. All medical directors (4/4) of the admitting trauma services and short-stay trauma unit (1/1) sites believed that these sites were sustainable beyond 5 years. On the contrary, all medical directors of on-call consultant facilities either did not believe their facilities were sustainable (4/5) or “did not know” (1/5). The most commonly reported barriers to program sustainability were inadequate compensation (6/10); limited access to hospital resources such as operating rooms, beds, and consultants (6/10); and difficulty in recruiting new medical staff to their site (2/10). All trauma medical directors of on-call consultant centres desired to transition toward an admitting trauma service but reported difficulties securing physician reimbursement.
DISCUSSION
This study describes the three models by which admitted injured patients are overseen at ten Level I to III adult trauma centres in BC: the admitting trauma service, on-call consultant, or short-stay trauma unit. The distribution of these models was not completely explained by the accreditation level, and at some centres, the model in situ did not satisfy accreditation standards.
Level I and II sites
Both Level I trauma hospitals in the province, each treating >1000 BC Trauma Registry cases annually, utilized an admitting trauma service. Numerous retrospective cohort studies have supported this concept as the ideal model for tertiary centres: up to an 8% mortality reduction, and reductions in-hospital length of stay, per-patient cost, and complications have been demonstrated following the adoption of an admitting trauma service (if it was replacing a model in which trauma patients were admitted by the general surgery service).Reference Simons, Eliopoulos, Laflamme and Brown 4 , Reference Davis, Cabbad and Schuster 6 , Reference Ursic, Curtis, Zou and Black 9 , Reference Abernathy, McGwin, Acker and Rue 10 However, these findings also challenge the reliance upon on-call consultant models of care at all Level II sites used in our trauma system. Two of these sites are the definitive trauma centres within their geographic region and health authority, with the closest neighbouring Level I site being 90–250 kilometres away. The patient volume, geographic isolation, and capability of these centres make them comparable to sites that have demonstrated improved patient outcomes with the implementation of an admitting trauma service.Reference Daniel, West, Craig-Blanco, Myers and Stewart 11 In fact, these on-call consultant sites fail to satisfy current credentialing guidelines, as Accreditation Canada recommends that all Level II centres provide “comprehensive inpatient trauma services . . . [utilizing an] interdisciplinary team . . . [that] manages trauma patients on a dedicated trauma unit or a clustered area,” a definition not consistent with that of the on-call consultant models currently operating within our trauma system. 2
Reassuringly, respondents at all Level II sites reported ongoing efforts to initiate an admitting trauma service model of care and thereby satisfy accreditation standards, as none of them believed their current on-call consultant model was sustainable. However, cited barriers to this transition included insufficient funds for program operation and staff reimbursement, as well as insufficient access to necessary hospital beds. These difficulties are evidenced by one Level II site that continued to employ an on-call consultant model at the time of publication despite initially planning an admitting trauma service start date of January 2016.
Inadequate funding was a commonly cited impediment to establishing an admitting trauma service. The inpatient care component of this model incurs additional physician duty hours and often necessitates the employment of a trauma nurse practitioner or case manager to ensure continuity of care. However, these ancillary positions have been independently shown to improve patient satisfaction and reduce the length of stay by up to 11%.Reference Collins, Miller and Kapu 12 Furthermore, retrospective cost analyses have suggested that overall per-patient hospital spending is up to 38% lower with an admitting trauma service, as compared with other models of care.Reference Davis, Cabbad and Schuster 6 Therefore, the upfront funds required for model transition might be recuperated over time. Nevertheless, if an admitting trauma service remained fiscally impossible, it stands to reason that an on-call consultant or short-stay trauma unit model may still be preferable to historic practices (which involve transferring care directly from the emergency physician to the surgical service overseeing the most severely injured organ system). These two models have the benefit of providing trauma-centred, multi-system oversight at least in the ED or during the first 48-hour portion of the hospital stay.
Level III sites
Within the Level III centres, both the on-call consultant and admitting trauma service models were equally common (2/5), and a unique “short-stay trauma unit” was also in situ (1/5). The ideal model of care for Level III centres remains uncertain, a result of insufficient evidence. Except for the 2010 study by Daniel et al.Reference Daniel, West, Craig-Blanco, Myers and Stewart 11 at the Level III Medical Center Hospital demonstrating a 33% reduction in the mortality of severely injured patients after the implementation of an admitting trauma service, all other studies that demonstrated improved patient outcomes with an admitting trauma service have been conducted in Level I or II facilities. However, a 2010 study by Hameed et al.Reference Hameed, Schuurman and Razek 13 suggested that geography may demand that some Level III centres emulate the care—and perhaps the model of care—of Level I and II hospitals, as 22% of Canadians and 24.3% of British Columbians reside outside the 1-hour road travel catchments of Level I and II centres, making them dependent upon rural and remote centres for initial stabilization. In summary, the limited data suggests that Level III trauma centres, particularly those who are geographically isolated from Level I and II sites, might benefit from an admitting trauma service model of care.
If organized by patient volume, we discovered that the highest-volume Level III centres in BC employed an admitting trauma service, while low-volume centres had an on-call consultant and the median sites utilized a short-stay trauma unit.Reference Jones, Evans and Thakore 14 However, differences in patient volumes between sites are subtle enough to cast doubt on its significance in determining the model of care. Indeed, admissions at the short-stay trauma unit sites outnumbered those at an admitting trauma service site 4 years out of the last 6 years.Reference Gezer, Thakore, McMillan and Robinson 8 , Reference Jones, Evans and Thakore 14 Interestingly, both Level III admitting trauma service sites belonged to a regional health authority employing a Level I admitting trauma service. It is possible that transitioning to an admitting trauma service at an affiliated hospital is more attractive than the de novo creation of such a service within a hospital network because of the following: satisfaction and comfort with a known model, incremental cost savings by adding to an existing model rather than starting a new one, or fewer barriers to implementing a familiar model.
Interestingly, the short-stay trauma unit site was the only centre without an admitting trauma service that indicated this model of care was sustainable for the next 5 years. To our knowledge, this is the first time this model of care has been described for injured patients. The short-stay trauma unit is an attractive and likely more sustainable alternative than the on-call consultant model for small centres for several reasons. First, it allows a greater number of patients such as those who are admitted to the short-stay trauma unit to receive care from physicians devoted to trauma. The size of this subgroup is not inconsequential, as Level III sites care for a relatively less injured cohort that remains following the transfer of more severely injured patients to higher level centres. A study of two rural hospitals in Kentucky noted that the length of stay for trauma patients was on average 2.3 and 3.8 days.Reference Richardson, Cross and Lee 15 Second, the creation of a dedicated trauma unit may facilitate the retention of some injured patients at smaller centres, specifically the 24% of patients transferred to Level I trauma centres who are discharged home within 48 hours without ever requiring surgical intervention.Reference Sorensen, von Recklinghausen, Fulton and Burchard 16 Multiple studies have demonstrated that Level III centres can provide safe, definitive care for this less injured population with no significant difference in outcomes, lower per-patient costs, and improved convenience to patients and their families.Reference Arthur, Newgard and Mullins 17 , Reference Newgard, Staudenmayer and Hsia 18
Sustainability
All medical directors of admitting trauma service and short-stay trauma unit sites believed that their sites were sustainable for >5 years in their current form, but none of the trauma medical directors at the five on-call consultant sites believed that their model was sustainable. Inadequate remuneration, difficulty in recruiting additional on-call physicians, limited access to hospital resources such as beds and consultants, and consulting services are reluctant to admit some non-operative injured patients were all cited as barriers to the sustainability of their current model of care. Respondents from on-call consultant sites unanimously believed that adopting an admitting trauma service would alleviate these difficulties. With respect to physician remuneration, improvements have been noted following the transition to an admitting trauma service, but only if accompanied by strategies to improve diagnostic and fee coding.Reference Davis, Cabbad and Schuster 6 We were unable to find any research examining the effects of an admitting trauma service on physician recruitment or hospital resource access.
CONCLUSIONS
The ten adult trauma centres in BC utilized three different models of overseeing admitted injured patients: the admitting trauma service, the on-call consultant, and the short-stay trauma unit models of care. The first system was employed at both Level I sites and several of their affiliated Level III sites and was universally believed to be sustainable. The on-call consultant model, found at all Level II and several Level III hospitals, was not sustainable at any site, and it further precluded Level II centres from satisfying accreditation guidelines. If financial and resource barriers can be overcome, implementing admitting trauma services at Level I and II sites, along with either admitting trauma services or short-stay trauma units at Level III centres, would improve the sustainability and accreditation compliance of our provincial trauma system.
Acknowledgements
The authors would like to acknowledge Dr. David Evans for inspiring this project and the BC Trauma Registry data analysts, Joanna Szpakowski and Karli Gamble, for their assistance. This research received no specific grant from any funding agency, commercial entity, or not-for-profit sector.
Competing interests: None declared.
