Emergency Medicine

Trauma team leader and early mortality: An interrupted time series analysis

a b s t r a c t

Background: The trauma team leader (TTL) is a “model” of a specifically dedicated team leader in the emergency department (ED), but its benefits are uncertain. The primary objective was to assess the impact of the TTL on 72- hour mortality. Secondary objectives included 24-hour mortality and admission delays from the ED.

Methods: Major trauma admissions (Injury Severity Score (ISS)>=12) in 3 Canadian Level-1 trauma centres were included from 2003 to 2017. The TTL program was implemented in centre 1 in 2005. An interrupted time series (ITS) analysis was performed. Analyses account for the change in patient case-mix (age, sex, and ISS). The two other centres were used as control in sensitivity analyses

Results: Among 20,193 recorded trauma admissions, 71.7% (n=14,479) were males. The mean age was 53.5 +-

22.0 years. The median [IQR] ISS was 22 [16-26]. TTL implementation was not associated with a change in the quarterly trends of 72-hour or 24-hour mortality: adjusted estimates with 95% CI were 0.32 [-0.22;0.86] and

-0.07 [-0.56;0.41] percentage-point change. Similar results were found for the proportions of patients admit- ted within 8 hours of ED arrival (0.36 [-1.47;2.18]). Sensitivity analyses using the two other centres as controls yielded similar results.

Conclusion: TTL implementation was not associated with changes in mortality or admission delays from the ED. Future studies should assess the potential impact of TTL programs on other patient-centred outcomes using different quality of care indicators.

(C) 2022

  1. Introduction

Trauma is often reported as the leading cause of death in patients under 40 years old [1,2]. Every day, 10,000 Canadians are injured and re- quire acute care in the Emergency Department (ED) [3]. Several compli- cations may arise from a trauma, which have been associated with altered long-term functional capacity and lesser quality of life [4]. Specific quality improvement (QI) strategies may help prevent or

* Corresponding author at: CHU de Quebec – Universite Laval, Hopital de l’Enfant-Jesus, 1401, 18e rue, H-608, QC G1J 1Z4, Canada.

E-mail address: [email protected] (M. Emond).

avoid some of these complications [5]. Different approaches have been tested to improve efficiency and quality of care in critically ill ED pa- tients. An example of a currently implemented solution is the creation of dedicated care teams such as a “sepsis team”, a “stroke team”, or a “trauma team” [6,7]. More specifically, the trauma team includes a ded- icated physician who acts as a designated “leader” (Trauma Team Leader (TTL)) during the acute phase of the patient’s trauma care. The presence of a trauma team led by an experienced physician could poten- tially facilitate and improve resuscitation, diagnosis and treatment for trauma patients [8]. During the Initial resuscitation phase of a trauma patient, a designated TTL can also alleviate the workload of the on- duty emergency physician, who can then be available to provide


0735-6757/(C) 2022

medical care to other patients in the ED. A TTL could also improve the medical team’s adherence to advanced trauma life support (ATLS) pro- tocols. [9]. Even though trauma teams and TTLs are widely implemented across Canada, their actual benefits have yet to be assessed. For exam- ple, their impact on time to incision for emergent surgery and ED length of stay is unclear [10-12], despite some studies suggesting that additional human resources could reduce the total hospital LOS [13]. This is of concern because prolonged ED LOS has been identified as an independent risk factor of pneumonia among intubated blunt trauma patients [14], and is associated with increased in-hospital mortality [15]. The literature on trauma teams and TTLs is heterogeneous; while a pre-post implementation study found a statistically significant decrease in intensive care unit admission rates, ED LOS and mortality [16], another Canadian study showed that an on-call TTL did not improve survival rates or ED LOS compared to on-site medical team [10]. Hence, the impacts of the TTL are still unclear. The primary objective of this study is to evaluate the impact of a dedicated TTL on 72-h patient mortality. Secondary objectives are to assess its impact on 24-h mortal-

ity and on time to admission from the ED.

  1. Methods
    1. Study design and setting

We conducted a retrospective multicentre cohort study using Que- bec’s provincial Trauma Registry (RTQ) data. This registry is managed by the Quebec Ministry of Health and Social Services (Ministere de la Sante et des Services Sociaux). It contains information on all trauma pa- tients (International Classification of Diseases codes 800-959) treated at any of the 59 designated trauma centers in the province (population of 8.5 million in a geographic area of approximately 1.7 million km2 [17]). Trauma level designations are based on the American College of Surgeons’ criteria and are periodically revised following on-site visits [18,19].

From 2003 to 2010, the RTQ inclusion criteria were: death following injury, intensive care unit admission, hospital stay >=3 days or transfer from another hospital. From April 2010 onwards, each patient admitted to the hospital after a visit to an ED following a trauma is included in the registry. The results are reported as per the STROBE guidelines.

This project was approved by the [BLINDED] Research ethics board and the Commission d’Acces a l’information; therefore, patient consent was not required. Data were all anonymous before analysis. Results are reported as per the STROBE guidelines [20].

    1. Study population

wrist or ankle, paralysis following the trauma, severe Burn injury (>20% body’s surface area) and at the discretion of the ED physician even if none of the above criteria are present. When a “trauma code” is activated, the on-call TTL is expected to arrive at the ED within 20 min, either before the patient has arrived (if activation is done from the information received by the paramedics) or after the initial assess- ment made by the EP. The TTL coordinates the resuscitation manage- ment, ensures adherence to guidelines, and leads the trauma team.

The pre-implementation phase was from April 2003 to March 2005, and the post-implementation phase was from April 2005 to December 2017.

    1. Outcome measures

The primary outcome was 72-h in-hospital mortality.

Secondary outcomes were: 24-h in-hospital mortality and the cumulative proportion of hospital admissions from the ED within 8 h of arrival.

    1. Statistical analyses

We estimated the effect of the TTL using an interrupted time series (ITS) methodology. We used the prognostic score to obtain standard- ized quarter proportions of mortality to account for potential changes in patient case mix (age, sex, Injury severity score (ISS), pre-hospital systolic blood pressure (SBP), Glasgow coma scale (GCS), number of comorbidities, injury mechanism, type of injury, and transfer-in from another acute care hospital) [23,24].

Time to hospital admission from the ED was calculated using inverse probability weighting to account for the change in patient case mix. We computed the difference in standardized cumulative incidence of admission 8 h after arrival [25,26]. Given the low incidence of death within 8 h of ED arrival (<1.5%), deceased patients were censored. An increase in the cumulative incidence of hospital admissions is consid- ered a good outcome.

Considering that previous work has shown that accreditation seemed beneficial for centres experiencing a decrease in performance preceding accreditation [27], we also adjusted for accreditation (which was a co-intervention occurring in the first quarter of 2012 in all centres).

Segmented regressions were used to estimate the short (change in outcome level) and long-term impact (change in outcome trends) of TTL introduction. They consist of linear regressions with autocorrelated

errors to account for the serial correlation: Yt = ?0 + ?1time + ?2TTL+

?3PostTTL + ?4Accreditation + ?5PostAccreditation + P3 ?q* (I[Quartert =


major trauma patients (defined as an Injury Severity Score (ISS)

>=12) aged >=16 years were included in our analyses if they were admit- ted to one of the three adult Level-I trauma centres of the province of Quebec between 2003/04 and 2017/12.

We used the ISS > 12 threshold based on the literature review ex- ploring the effects of different AIS revisions (1998, 2008 and 2015) on clinical outcome measures. An ISS15 >= 12 performs similarly to a thresh- old ISS98 >= 16 for in-hospital mortality and ICU admission. Since mortal- ity is the primary outcome, an ISS >12 threshold may be adopted to identify major trauma patients [21,22].

Centres 1 and 2 are urban hospitals covering a 4.1 million population and are in the same geographic area of approximately 365 km2. Centre 3 covers a population of almost 2 million in an urban geographic area of approximately 485 km2.

The TTL program was implemented in centre 1 in April 2005 with a 24-h daily emergency physician, trauma surgeon or anaesthesiologist TTL coverage. Trauma team and TTL activation criteria are as follows: advanced airway management needed, systolic blood pressure < 90 mmHg at ED arrival, packed red cell transfusion needed, penetrating trauma to the head, neck, torso or abdomen, amputation above the

q]) + Ru , with Rt= ?1Rt-1 + ?pRtp + ?t. Yt is the quarterly proportion of the outcome at the time unit t, ?0 is the baseline level of the outcome, time is coded 0 to 58, and its coefficient ?1is the baseline trend. Dummy variables TTl and Accreditation indicate whether each

time point occurred before or after the corresponding interventions (0 for all-time prior and 1 for all time after). The coefficients ?2 and ?4 are the change in the level of Yt associated respectively with TTL intro- duction and accreditation. Post_TTL and Post_Accreditation represent the number of time units since the corresponding intervention (0 for all time until the intervention; 1,2,3…for subsequent time points), and their coefficients ?3 and ?5 the change in the trend of the studied outcome. Rt is the autoregressive component, comprised of the autoregressive parameter ?p for lag p, and the random error ?t. Seasonality was investigated and modelled by incorporating dummy variables for each quarter ?q and autocorrelated error terms at a given seasonal lag [28].

Sensitivity analyses: to assess our results’ robustness to possible co- interventions that may have coincided with the TTL program introduc- tion, controlled interrupted time series (CITS) was applied when the pre-accreditation outcome trends for the treated and the non-treated

Time series analysis“>centers were parallel [29]. We also used the two other centres as “pla- cebo”, fitting the ITS model described above even if these centres only introduced a TTL after 2017. The rationale is to identify potential shock as we should not expect changes (particularly in level) of the TTL implementation in centres 2 and 3.

Multiple imputation with chained equations was used to impute missing data on the GCS score (35%), number of comorbidities (3%) and SBP (3%) [27,28,30]. Rubin’s rules were used to combine esti- mates across 35 imputed datasets and to obtain 95% confidence in- tervals [31]. More details on the model specification of the prognostic score, inverse probability weighting, and the CITS are available in Appendix 1.

  1. Results
    1. Patient characteristics

A total of 20,193 patients were included in our analyses, 71.7% of whom were male (n = 14,479). The mean age was 53.5 +- 22.0 years. The most common injury mechanism was fall (46.9%, n = 9479). A total of 11,778 (58.3%) patients were admitted after an inter-hospital transfer, and 10,493 (52.0%) patients were admitted to the ICU. The

median [IQR] ISS was 22 [16-26]. injury patterns and patient character- istics are described in Table 1. In centre 1, a total of 726 patients were included before the TTL implementation, and 6724 patients were in- cluded in the post-TTL implementation period; among those, 4728 (70.3%) were managed by a TTL. The distribution and patient character- istics of centre 1 and the two control centres used in the sensitivity analyses are presented in Table 1.

Figures 1 and 2 display the crude and adjusted quarterly 72-h and 24-h mortality proportion for the adult level 1 trauma centre 1. No decrease in the quarterly mortality trends through time following the TTL program implementation was observed: adjusted estimates with 95% CI were 0.32 [-0.22;0.86] and -0.07 [-0.56, 0.41]

percentage-point increase in centre 1 for 72-h and 24-h mortality, respectively.

The crude and adjusted proportion of patients admitted within 8 h of ED arrival are shown in Fig. 3. No decrease in the quarterly trends through time was observed: adjusted estimates with 95% CI were 0.36 [-1.47;2.18] percentage-point decrease. Level and change trend analyses are shown in Table 2.

Table 1

Sociodemographic and clinical characteristics.

Centre 1

n = 7450 (36.9) n (%)

Centre 2

n = 6268 (31) n (%)

Centre 3

n = 6475 (32.1) n (%)


n = 20,193 n (%)

Year of inclusion 2003

268 (3.6)

255 (4.1)

251 (3.9)

774 (3.7)


364 (4.9)

372 (5.9)

369 (5.7)

1105 (5.5)

2005, TTL implementation

442 (5.9)

409 (6.5)

433 (6.7)

1284 (6.4)


429 (5.8)

418 (6.7)

424 (6.5)

1271 (6.3)


464 (6.2)

370 (5.9)

398 (6.14)

1232 (6.1)


504 (6.8)

433 (6.9)

402 (6.2)

1339 (6.6)


544 (7.3)

443 (7.0)

426 (6.6)

1413 (7.0)


609 (8.2)

374 (6.0)

467 (7.2)

1450 (7.2)


591 (7.9)

436 (6.9)

460 (7.1)

1487 (7.4)


563 (7.6)

425 (6.8)

449 (6.9)

1437 (7.1)


489 (6.6)

404 (6.4)

503 (7.8)

1396 (6.9)


535 (7.2)

409 (6.5)

445 (6.9)

1389 (6.9)


540 (7.2)

492 (7.82)

464 (7.2)

1496 (7.4)


553 (7.4)

461 (7.4)

471 (7.3)

1485 (7.3)


553 (7.4)

567 (9.0)

513 (7.9)

1633 (8.1)

Age, mean +- SD

54.1 +- 22.1

52.8 +- 22.2

53.4 +- 21.7

53.5 +- 22.0

Sex, male

5278 (70.8)

4551 (72.6)

4650 (71.8)

14,479 (71.7)

Injury mechanism Motor vehicle collisions

2393 (32.1)

2722 (43.4)

2813 (43.4)

7928 (39.3)


3722 (50)

2774 (44.3)

2983 (46.1)

9479 (46.9)


467 (6.3)

205 (3.3)

104 (1.6)

776 (3.8)


868 (11.6)

567 (9.0)

575 (8.9)

2010 (10.0)

Injury pattern

Traumatic brain injuries

4591 (61.6)

3241 (51.7)

3839 (59.3)

11,671 (57.8)

spinal cord injuries

238 (3.2)

998 (15.9)

643 (9.9)

1879 (9.3)

Multiple trauma

827 (11.1)

618 (9.8)

558 (8.6)

2003 (9.9)

Thoraco-abdominal injuries

1154 (15.5)

790 (12.6)

761 (11.7)

2705 (13.4)

orthopaedic injuries

604 (8.1)

590 (9.4)

613 (9.5)

1807 (8.9)

Other injuries

36 (0.5)

31 (0.5)

61 (0.9)

128 (0.6)

Transfer from another hospital

3797 (51)

3828 (61.1)

4153 (64.1)

11,778 (58.3)

Intubated before admission

1602 (21.5)

963 (15.3)

656 (10.1)

3221 (15.9)

GCS, n = 12,996 13-15

911 (12.2)

608 (9.7)

539 (8.3)

2058 (10.3)


472 (6.3)

301 (4.8)

293 (4.5)

1066 (5.3)


3936 (52.8)

3008 (48)

2928 (45.2)

9872 (48.9)

SBP, mean +- SD

136.6 (29.3)

130.6 (28.8)

132.8 (27.2)

133.5 (28.6)

ISS, median [IQR]

22 [17-27]

20 [16-26]

22 [16-26]

22 [16-26]

TTL activation

4728 (63.5)



4728 (23.3)

ICU admission

4188 (56.2)

3341 (53.3)

2964 (45.8)

10,493 (52)

ED: Emergency department; GCS: Glasgow coma Scale; ICU: Intensive care unit; ISS: Injury severity score; SBP: Systolic blood pressure; TTL: Trauma team leader; SD: Standard Deviation; IQR: Interquartile range.

Image of Fig. 1

Fig. 1. Quarterly proportions of 72 h in-hospital mortality in centre 1 – TTL exposure. The time axis shows the year and the quarters.

    1. Controlled interrupted time series analysis

Our CITS model was performed as sensitivity analysis to assess the potential unmeasured confounding at the Trauma system level. To do so, pre-TTL period characteristics of centres 2 and 3 were compared to those of centre 1 to identify the best control centre and evaluate the TTL implementation’s impact on the different outcomes. None of the centres could be used as a single control site for each outcome. Centre 3 was the most suitable control to assess the effect of the TTL on the

72-h mortality outcome, while centre 2 was used as a control for the 24-h mortality outcome (Figs. 4 and 5). None of the two-control centres could be used to assess the effect of the TTL on the proportion of patients admitted within 8 h of ED arrival because of divergent mean levels and trends in the pre-TTL implementation period (See Fig. 6).

As shown in Figs. 4 and 5, we found no difference in quarterly trends mortality among the patients exposed to the TTL and those who were not (at 72 h: adjusted mortality: -0.21 [-0.64;0.22] percentage-point decrease; and at 24 h: 0.04 [-0.40;0.49] percentage-point increase).

Image of Fig. 2

Fig. 2. Quarterly proportions of 24 h in-hospital mortality in centre 1. The time axis shows the year and the quarters.

    1. Placebo test

No decrease in the quarterly trends of the 72-h (Appendix 2) or 24-h mortality (Appendix 3) through time was observed for centre 2 (-0.48 [-1.31;0.36] and -0.08 [-0.74;0.57], respectively) nor for centre 3

(0.35 [-0.40;1.10] and 0.15 [-0.41;0.72] respectively). In centre 2, we observed a greater proportion of patients admitted within 8 h of ED arrival (2.14 [0.25;4.03], Appendix 4). This was not observed in cen- tre 3 (0.59 [-1.57;2.75], Appendix 3).

  1. Discussion

Our results suggest that implementing a TTL did not impact 72-h and 24-h inpatient mortality and did not reduce patient ED LOS. To our knowledge, our study is the first to report the lack of impact of the TTL on patient mortality. In line with recent Lavigueur et al., our study also failed to show improvement in ED LOS by adding an “extra” resource [16]. Conversely, the same authors found a statistically significant de- crease in mortality (1.25% absolute reduction, 16% relative reduction)

Image of Fig. 3

Fig. 3. Quarterly proportions of patients discharged within 8 h of ED arrival in all three level-I trauma centres. The time axis shows the year and the quarters.

and ICU admissions (4.46% absolute reduction, 14% relative reduction) despite no change in the overall LOS after the implementation of the TTL program [16]. Lavigueur et al. conducted a pre-post study; this might explain the differences between our results and theirs and limit the interpretation of their findings. Indeed, it is difficult to distinguish whether these findings are related to the TTL implementation or a sec- ular improvement in overall trauma care. The use of controlled interrupted time series analysis in our study reduces the impact of sec- ular improvement in trauma care while providing a better estimate of

TTL’s short-term impact on mortality and LOS. The CITS is considered one of the most robust quasi-experimental designs and is the method of choice when a randomized-controlled study is difficult or unethical to realize [32].

The use of mortality as a performance indicator in traumatology seems adequate. However, there are varioUS time points to assess mor- tality in the literature; this includes using outcomes such as “in-hospital mortality” without specifying the delay. Other authors have also used three and Six-month mortality [13,16,33]. This contributes to the debate

Table 2

Level and trend change of the interrupted time series analyses.

Level change Trend change Interrupted time series 72 h mortality

Centre 1 0.19 [-2.79;3.18] 0.32 [-0.22;0.86]

Centre 2 (placebo) -2.16 [-6.81;2.50] -0.48 [-1.31;0.36]

Centre 3 (placebo) -1.35 [-5.49;2.79] 0.35 [-0.40;1.10]

Interrupted time series 24 h mortality

Centre 1 -0.41 [-3.10;2.28] -0.07 [-0.56;0.41]

Centre 2 (placebo) 0.40 [-3.25;4.05] -0.08 [-0.74;0.57]

Centre 3 (placebo) 0.21 [-2.92;3.34] 0.15 [-0.41;0.72]

Controlled interrupted time series 72 h mortality

Centre 1 vs Centre 3 0.40 [-1.02;1.82] -0.21 [-0.64;0.22]

Controlled interrupted time series 24 h mortality

Centre 1 vs Centre 2 -0.54 [-4.53;3.46] 0.04 [-0.40;0.49]

on the pertinence of performance indicators in acute traumatology [33]. Because trauma patients’ mortality could have been influenced by other in-hospital variables that were not collected in the RTQ, such as inten- sive care and/or surgical management, but also withdrawal of life- support therapies, we determined that the impact of the TTL should be evaluated at an early stage (72 and 24 h after patient arrival). But no consensus has ever been published regarding the most accurate time point. Assessing the impact of TTLs on quality of care, healthcare provider satisfaction, delay before CT scan, ICU LOS and morbidity could also be of interest. The benefits of the TTL should also be assessed in specific groups of patients (i.e., patients needing airway manage- ment, multi-trauma patients, etc.). Because the on-duty emergency physician did not assume the role, the TTL implementation may also im- pact the morbidity and mortality of the other patients being treated in the ED.

In our study, there are several possible explanations for the lack of

impact of the TTL in our findings. One of them would be that only a pro- portion of admitted patients have triggered a TTL activation. Hence, if the impact of TTL on study outcomes is small, it might have been diluted as we included all trauma admissions with an ISS > 12 in our study.

Given that we have a short pre-period (24 months versus 11 years for the post-period), we might have some issues accounting for underlying secular trends. In addition, we only have a pool of two potential control centres, which limits our ability to find a suitable control (which satis- fied the required assumptions) for each outcome. It could also be de- bated whether the benefit of the TTL could have been underestimated because of the heterogenous background specialty (emergency physi- cian and surgeons) of the TTL physicians. Cummings et al. suggested in a retrospective cohort study that surgeons, on-call emergency physi- cians (EPs), or on-shift EPs can act as the TTL without negatively impacting patient survival or ED LOS [34]. A meta-analysis found that there was no difference in survival (OR 0.82, 95% CI [0.61-1.10], p = 0.19) and LOS when surgeon or non-surgeon TTLs led the trauma team [35]. Another hypothesis is that the Experience level among TTL was not evaluated and could vary across sites as reported elsewhere [36]. The absence of ED LOS improvement by adding an “extra” resource may be attributed to the increase in ED crowding, which is unrelated to the TTL itself.

Other QI interventions have been studied in traumatology. Dodek

et al. reported that after Trauma Team Activation (TTA), the median elapsed time from initial nursing assessment in the ED to arrival in the operating room for blunt trauma patients could be decreased. Still, there were no significant differences in the crude or adjusted mortality [11]. Other studies showed that there were no apparent differences in mortality for patients exposed to a TTA and non-TTA either with severe presentation (ISS > 25 group) or less severe presentation (ISS 16-25) [37]. In addition, patients in the TTA group had shorter delays between their ED departure and operation room arrival. Our study did not ex- plore this because data regarding the time to the operation room was not collected. When it comes to QI interventions, evaluating and sus- taining a Quality improvement project in the ED can be quite challeng- ing [37]. However, our results showed an increase in the number of TTL activations over time, weighing in favour of the sustainability of the TTL model in trauma centers.

One limitation of this study is its retrospective design. Nevertheless, to our knowledge, this is the first study performing sensitivity analyses (controlled ITS design, placebo test) to strengthen results. Some biases

Fig. 4. Controlled Interrupted time series analysis – proportion of 72 h in-hospital mortality.

Image of Fig. 5

Fig. 5. Controlled Interrupted time series analysis – proportion of 24 h in-hospital mortality. The time axis shows the year and the quarters.

may remain. We cannot exclude the presence of a co-intervention in the Quebec trauma system around the TTL implementation. Results from placebo test in centres 2 and 3 suggest a slight decrease thought impre- cision of the 72-h mortality, and an increase in the trends of admission from ED, specifically for centre 2. It is also of note that ITS are generally underpowered to detect small-size effect.

Furthermore, measurement errors on patients’ covariates and out- comes cannot be ruled out even though clinical data were extracted from the RTQ, which has strict inclusion criteria, standardized data col- lection procedures by trained medical archivists, and rigorous data qual- ity control. These potential measurement errors may have biased our results in either direction. Some factors, such as prehospital time before

Fig. 6. Controlled Interrupted time series analysis – proportion of patients discharged within 8 h of ED arrival.

ED admission, were not recorded and could have affected our results. We cannot exclude the possibility that the activation guidelines were not rigorously applied. Finally, not all admitted patients in our study during the post-TTL period have required a TTL activation. We attempted to perform a propensity score matched study (matching ad- mission for whom a TTL was activated to a similar admission in the pre- TTL period) to estimate the impact of TTL implementation on admis- sions. However, achieving a balance of measured covariates between treated and controls for admissions was difficult.

In the present study, TTL implementation was not associated with

reduced mortality or discharge delays from the ED. These may not be the appropriate indicators of the benefits of the TTL. Future studies are needed to assess the potential impact of TTL programs on other patient-centred outcomes using different quality of care indicators.


JT&AB received a scholarship from the Fondation du CHU de Quebec (no grant number). JT received a Fellowship award from the Societe francaise de medecine d’urgence. This project was funded by the Fonds de recherche du Quebec – Sante (#33239). These funding agen- cies were in no way involved in this study’s design, data collection and analyses.

Declaration of Competing Interest



We would like to thank the Emergency physicians of the Hopital de l’Enfant-Jesus (CHU de Quebec-Universite Laval) for their support.

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi. org/10.1016/j.ajem.2022.09.048.


  1. Choi J, Carlos G, Nassar AK, Knowlton LM, Spain DA. The impact of trauma systems on patient outcomes. Curr Probl Surg. 2021;58(1):100849. https://doi.org/10.1016/ j.cpsurg.2020.100849.
  2. Centers for Disease Control and Prevention. Injury Prevention & Control – Fatal In- jury and Violence Data. https://www.cdc.gov/injury/wisqars/fatal.html; 2021. [accessed 2021-05-19].
  3. Parachute. Parachute, The cost of injury in Canada. www.parachutecanada.org/ downloads/research/Cost_of_Injury-2015.pdf; 2015.
  4. Holbrook TL, Hoyt DB, Anderson JP. The impact of major in-hospital complications on functional outcome and quality of life after trauma. J Trauma Acute Care Surg. 2001;50(1):91-5. https://journals.lww.com/jtrauma/Fulltext/2001/01000/The_ Impact_of_Major_In_Hospital_Complications_on.16.aspx.
  5. Reicks P, Thorson M, Irwin E, Byrnes MC. Reducing complications in trauma patients: use of a standardized quality improvement approach. J Trauma Nurs. 2010;17(4): 185-90. https://doi.org/10.1097/JTN.0b013e3181ff247e.
  6. Cho HJ, Lee KY, Nam HS, Kim YD, Song TJ, Jung YH, et al. Process improvement to en- hance existing stroke team activity toward more timely Thrombolytic treatment. J Clin Neurol. 2014;10(4):328-33. https://doi.org/10.3988/jcn.2014.10.4.328.
  7. Verhoeff K, Saybel R, Fawcett V, Tsang B, Mathura P, Widder S. A quality- improvement approach to effective trauma team activation. Can J Surg. 2019;62 (5):305-14. https://doi.org/10.1503/cjs.000218.
  8. Georgiou A, Lockey DJ. The performance and assessment of hospital trauma teams. Scand J Trauma Resusc Emerg Med. 2010;18:66. https://doi.org/10.1186/1757- 7241-18-66.
  9. Tsang B, McKee J, Engels PT, Paton-Gay D, Widder SL. Compliance to advanced trauma life support protocols in adult trauma patients in the acute setting. World J Emerg Surg. 2013;8(1):39. https://doi.org/10.1186/1749-7922-8-39.
  10. Cummings GE, Mayes DC. A comparative study of designated trauma team leaders on trauma patient survival and emergency department length-of-stay. CJEM. 2007;9(2):105-10. https://doi.org/10.1017/s1481803500014871.
  11. Dodek P, Herrick R, Phang PT. Initial Management of Trauma by a trauma team: ef- fect on timeliness of Care in a Teaching Hospital. Am J Med Qual. 2000;15(1):3-8. https://doi.org/10.1177/106286060001500102.
  12. Khetarpal S, Steinbrunn BS, McGonigal MD, Stafford R, Ney AL, Kalb DC, et al. Trauma faculty and trauma team activation: impact on trauma system function and patient outcome. J Trauma Acute Care Surg. 1999;47(3):576-81. https://journals.lww.com/ jtrauma/Fulltext/1999/09000/Trauma_Faculty_and_Trauma_Team_Activation Impact.28.aspx.
  13. Mains C, Scarborough K, Bar-Or R, Hawkes A, Huber J, Bourg P, et al. Staff commit- ment to trauma care improves mortality and length of stay at a level I trauma center. J Trauma Acute Care Surg. 2009;66(5):1315-20. https://doi.org/10.1097/TA. 0b013e31819d96d8.
  14. Carr BG, Kaye AJ, Wiebe DJ, Gracias VH, Schwab CW, Reilly PM. Emergency depart- ment length of stay: a major risk factor for pneumonia in intubated blunt trauma pa- tients. J Trauma. 2007;63(1):9-12. https://doi.org/10.1097/TA.0b013e31805d8f6b.
  15. Mowery NT, Dougherty SD, Hildreth AN, JHt Holmes, Chang MC, Martin RS, et al. Emergency department length of stay is an independent predictor of hospital mor- tality in Trauma activation patients. J Trauma. 2011;70(6):1317-25. https://doi.org/ 10.1097/TA.0b013e3182175199.
  16. Lavigueur O, Nemeth J, Razek T, Maghraby N. The effect of a multidisciplinary trauma team leader paradigm at a Tertiary trauma center: 10-year experience. Emerg Med Int. 2020;2020:8412179. https://doi.org/10.1155/2020/8412179.
  17. Institut national d’excellence en sante et services sociaux. Continuum de services en traumatologie. https://www.inesss.qc.ca/thematiques/sante/traumatologie/ continuum-de-services-en-traumatologie-cst.html; 2015.
  18. American College of Surgeons – Committee on Trauma. Resources for Optimal care of the injured patient. https://www.facs.org/-/media/files/quality-programs/trauma/ vrc-resources/resources-for-optimal-care.ashx; 2014. [2021].
  19. Bracewell N, Winchester DE. Accreditation in health care: does it make any differ- ence to patient outcomes? BMJ Qual Saf. 2021. https://doi.org/10.1136/bmjqs- 2020-012533.
  20. von Elm E, Altman DG, Egger M, Pocock SJ, Gotzsche PC, Vandenbroucke JP. The strengthening the reporting of observational studies in epidemiology (STROBE) statement: guidelines for reporting observational studies. Lancet. 2007;370(9596): 1453-7. https://doi.org/10.1016/s0140-6736(07)61602-x.
  21. Van Ditshuizen JC, Sewalt CA, Palmer CS, Van Lieshout EMM, Verhofstad MHJ, Den Hartog D. The definition of major trauma using different revisions of the Abbreviated Injury Scale. Scand J Trauma Resusc Emerg Med. 2021;29(1):71. https://doi.org/10. 1186/s13049-021-00873-7.
  22. Palmer CS, Gabbe BJ, Cameron PA. Defining major trauma using the 2008 abbrevi- ated injury scale. Injury. 2016;47(1):109-15. https://doi.org/10.1016/j.injury.2015. 07.003.
  23. Richardson DB, Keil AP, Cole SR, Kinlaw AC. Assessing exposure-response trends using the disease risk score. Epidemiology. 2020;31(2). https://doi.org/10.1097/ EDE.0000000000001145. e15-e6.
  24. Richardson DB, Keil AP, Kinlaw AC, Cole SR. Marginal structural models for risk or prevalence ratios for a point exposure using a disease risk score. Am J Epidemiol. 2019;188(5):960-6. https://doi.org/10.1093/aje/kwz025.
  25. Batomen B, Moore L, Strumpf E, Nandi A. Addressing competing risks when assessing the impact of health services interventions on hospital length of stay. Ep- idemiology. 2021;32(2):230-8. https://doi.org/10.1097/ede.0000000000001307.
  26. Naimi AI, Moodie EE, Auger N, Kaufman JS. Constructing inverse probability weights for continuous exposures: a comparison of methods. Epidemiology. 2014;25(2): 292-9. https://doi.org/10.1097/ede.0000000000000053.
  27. Batomen B, Moore L, Strumpf E, Champion H, Nandi A. Impact of trauma Centre ac- creditation on mortality and complications in a Canadian trauma system: an interrupted time series analysis. BMJ Qual Saf. 2021;30(11):853-66. https://doi. org/10.1136/bmjqs-2020-011271.
  28. Donnelly N. The use of interrupted time series analysis to evaluate the impact of pharmaceutical benefits scheme policies on drug utilisation in Australia. [Postgrad- uate research thesis for the degree of Doctor of Philosophy]. University of New South Wales; 2005.
  29. Lopez Bernal J, Cummins S, Gasparrini A. The use of controls in interrupted time se- ries studies of public health interventions. Int J Epidemiol. 2018;47(6):2082-93. https://doi.org/10.1093/ije/dyy135.
  30. Van der Kruijk M. Multiple imputation with chained equations and survival out- comes; 2015.
  31. Azur MJ, Stuart EA, Frangakis C, Leaf PJ. Multiple imputation by chained equations: what is it and how does it work? Int J Methods Psychiatr Res. 2011;20(1):40-9. https://doi.org/10.1002/mpr.329.
  32. Kontopantelis E, Doran T, Springate DA, Buchan I, Reeves D. Regression based quasi- experimental approach when randomisation is not an option: interrupted time se- ries analysis. Bmj. 2015;350:h2750. https://doi.org/10.1136/bmj.h2750.
  33. Moore L, Turgeon AF, Emond M, Le Sage N, Lavoie A. Definition of mortality for trauma center performance evaluation: a comparative study. Crit Care Med. 2011; 39(10):2246-52. https://doi.org/10.1097/CCM.0b013e3182227a59.
  34. Cummings GE, Mayes DC. A comparative study of designated trauma team leaders on trauma patient survival and emergency department length-of-stay. Can J Emerg Med. 2007;9(2):105-10. https://doi.org/10.1017/S1481803500014871.
  35. Hajibandeh S, Hajibandeh S. Who should lead a trauma team: surgeon or non- surgeon? A systematic review and meta-analysis. J Inj Violence Res. 2017;9(2): 107-16. https://doi.org/10.5249/jivr.v9i2.874.
  36. Ringen AH, Hjortdahl M, Wisborg T. Norwegian trauma team leaders-training and experience: a national point prevalence study. Scand J Trauma Resusc Emerg Med. 2011;19:54. https://doi.org/10.1186/1757-7241-19-54.
  37. Ham C, Kipping R, McLeod H. Redesigning work processes in health care: lessons from the National Health Service. Milbank Q. 2003;81(3):415-39. https://doi.org/ 10.1111/1468-0009.t01-3-00062.

Leave a Reply

Your email address will not be published. Required fields are marked *