WTA 2014 PLENARY PAPER
Implementation of a surgical intensive care unit service is associated with improved outcomes for trauma patients Amanda L. Klein, MD, Carlos V.R. Brown, MD, Jayson Aydelotte, MD, Sadia Ali, MPH, Adam Clark, and Ben Coopwood, MD, Austin, Texas
Our trauma service recently transitioned from a pulmonary intensive care unit (ICU) service to a surgical ICU (SICU) service. We hypothesized that a newly formed SICU service could provide comparable outcomes to the existing pulmonary ICU service. A specific aim of this study was to compare outcomes of trauma patients admitted to the ICU before and after implementation of a SICU service. METHODS: We performed a retrospective study of trauma patients admitted to the ICU of our urban, American College of SurgeonsY verified, Level 1 trauma center during a 4-year period (2009Y2012). Patients managed by the pulmonary ICU service (2009Y2010) were compared with patients managed by a SICU service (2011Y2012). The primary outcome was mortality, while secondary outcomes included complications (pulmonary, infectious, cardiac, and thromboembolic), hospital and ICU length of stay, ventilator days, and need for reintubation. RESULTS: There were 2,253 trauma patients admitted to the ICU during the study period, 1,124 and 1,129 managed by the pulmonary ICU and SICU services, respectively. When comparing outcomes for SICU and pulmonary ICU patients, there was no difference in mortality (11% vs. 13%, p = 0.41), but patients managed by the SICU service had fewer pulmonary complications (3% vs. 6%, p G 0.001), fewer days on the ventilator (3 vs. 4, p = 0.002), and less often required reintubation after extubation (4% vs. 9%, p G 0.001). CONCLUSION: Transition from a pulmonary ICU service to a SICU service at our institution was associated with no change in mortality but an improvement in pulmonary complications, ventilator days, and reintubation rates. Trauma centers currently staffed with a pulmonary ICU service should feel comfortable converting to SICU service and should expect comparable or improved outcomes for trauma patients admitted to the ICU. (J Trauma Acute Care Surg. 2014;77: 964Y968. Copyright * 2014 by Lippincott Williams & Wilkins) LEVEL OF EVIDENCE: Therapeutic/care management study, level IV. KEY WORDS: ICU; trauma; outcomes; intensivist; intesive care. BACKGROUND:
T
rauma patients admitted to an intensive care unit (ICU) may be managed by a variety of intensivists including anesthesiology, pulmonary, and surgical intensivists. However, according to the American College of Surgeons’ Committee on Trauma (ACS COT), Resources for Optimal Care of the Injured Patient,1 ‘‘[t]he trauma surgeon must remain in charge of the patient in the ICU’’ (criteria deficiency 11Y46) and ‘‘[i]n all trauma centers, the trauma service must retain responsibility for the patient and coordinate therapeutic decisions’’ (criteria deficiency 11Y53). The easiest way for a trauma center to meet these ACS COT requirements is to maintain a surgical ICU (SICU) service staffed by the trauma surgeon’s panel. However, many trauma centers may not have an existing SICU service and may require a transition from a medical or anesthesia ICU service to a SICU service. This transition can be met with resistance and skepticism from nurses, doctors, and administrators who have grown accustomed to a medical or anesthesia critical care service. Submitted: January 14, 2014, Revised: May 7, 2014, Accepted: July 28, 2014. From the University Medical Center Brackenridge, University of Texas SouthwesternY Austin, Austin, Texas. This study was presented at the 44th Annual Meeting of the Western Trauma Association, March 2Y7, 2014, in Steamboat Springs, Colorado. Address for reprints: Carlos V.R. Brown, MD, University of Texas SouthwesternV Austin, University Medical Center at Brackenridge, 601 East 15th St, Austin, Texas; email: [email protected]. DOI: 10.1097/TA.0000000000000460
964
From its inception, our trauma center’s ICU was staffed by a highly skilled group of pulmonary intensivists. However, as we transitioned from an ACS-verified Level 2 trauma to Level 1 trauma center, we received feedback from COT Verification Review Committee that our trauma service needed to be more involved in the intensive care of our severely injured patients. We decided that the simplest way to improve the critical care coverage of our severely injured patients was to implement a surgical critical care service staffed by our trauma surgeons. In 2011, our trauma service made the transition from a pulmonary ICU service to a full-time surgical critical care service. We hypothesized that a newly formed SICU service could provide comparable outcomes to the existing pulmonary ICU service. A specific aim of this study was to compare outcomes of trauma patients admitted to the ICU before and after implementation of a SICU service.
PATIENTS AND METHODS We performed a retrospective study of all trauma patients admitted to the ICU at THE University Medical Center Brackenridge, an urban ACS-verified Level 1 trauma center located in Austin, Texas. Our study spanned 4 years (2009Y2012), including two separate 2-year periods. During the first period (2009Y2010), trauma patients admitted to the ICU were managed by the pulmonary ICU service (pulmonary ICU group), J Trauma Acute Care Surg Volume 77, Number 6
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
J Trauma Acute Care Surg Volume 77, Number 6
Klein et al.
composed of 10 rotating pulmonary intensivists providing coverage of trauma ICU patients 24/7. During the second period (2011Y2012), trauma patients admitted to the ICU were managed by the SICU service (SICU group), composed of four rotating general/trauma surgeons covering trauma ICU patients 24/7. Data were extracted from the trauma registry and electronic medical record. Data collected for both periods include admission demographics, mechanism of injury, emergency department (ED) vital signs, Abbreviated Injury Scale (AIS) score for each body region, and Injury Severity Score (ISS). The primary outcome was mortality, while secondary outcomes included complications (pulmonary, infectious, cardiac, and thromboembolic), ventilator days, reintubation rates, as well as length of stay (LOS) in the hospital and ICU. The SICU and pulmonary ICU groups were compared by univariate analysis for all variables using the unpaired Student’s t test or Mann-Whitney rank-sum test for the continuous variables and Pearson’s r2 with Yates correction for categorical variables. Two subgroups were also analyzed, elderly patients (Q65 years old) and patients requiring at least 1 day of mechanical ventilation. Values are reported as mean (SD) or raw percentages, and statistical significance was set at p G 0.05. The local institutional review board approved this study.
RESULTS There were a total of 2,253 trauma patients admitted to our ICU during the 4-year study period, 1,124 from 2009 to 2010 managed by the pulmonary ICU service and 1,129 from 2011 to 2012 managed by the SICU service. Patients managed by the pulmonary ICU service were younger (44 [20] years vs. 47 [21] years, p = 0.01), but there was no difference in male sex (72% vs. 71%, p = 0.56) or white race (64% vs. 63%, p = 0.85). The pulmonary ICU group less often sustained blunt trauma (87% vs. 91%, p = 0.002), but there was no difference in ISS (21 [12] vs. 21 [12], p = 0.88) or AIS score of 3 or greater for any body region: head (55% vs. 57%, p = 0.32), chest (33% vs. 36%, p = 0.19), abdomen (18% vs. 17%, p = 0.35), and extremities (24% vs. 22%, p = 0.27). In the ED, there was no difference in hypotension (9% vs. 8%, p = 0.24), pulse rate (97 [25] vs. 95 [25], p = 0.28), Glasgow Coma Scale (GCS) of 8 or less (27% vs. 26%, p = 0.59), or the need for prehospital or ED endotracheal intubation (58% vs. 55%, p = 0.18) (Tables 1). There was no difference in mortality, 13% for patients managed by the pulmonary ICU service and 11% for patients managed by the SICU service (p = 0.41). Patients managed by the SICU service had fewer pulmonary complications (3% vs. 6%, p G 0.001), but there was no difference in infectious (20% vs. 20%, p = 0.87), cardiac (7% vs. 7%, p = 0.89), or thromboembolic (2% vs. 3%, p = 0.22) complications. Patients managed by the SICU service had fewer ventilator days (3 [5] days vs. 4 [6] days, p = 0.002), but there was no difference in ICU (5 [6] days vs. 5 [6] days, p = 0.39) or hospital (11 [12] days vs. 12 [12] days, p = 0.87) LOS. Despite spending one less day on the ventilator, the patients managed by the SICU service less often failed extubation and required intubation (4% vs. 9%, p G 0.001). Within the subgroup analyses, there were 443 elderly trauma patients, 202 managed by the pulmonary ICU service and 241 managed by the SICU service. When comparing pulmonary
TABLE 1. Population Characteristics for Pulmonary ICU and SICU Groups
Age, mean (SD), y Male sex Blunt mechanism Admission systolic blood pressure, mean (SD), mm Hg Hypotension (G90 mm Hg) Admission pulse, mean (SD) Admission GCS score, mean (SD) Intubated in ED Head AIS score, mean (SD) Head AIS score Q 3 Face AIS score, mean (SD) Face AIS score Q 3 Chest AIS score, mean (SD) Chest AIS score Q 3 Abdominal AIS score, mean (SD) Abdomen AIS score Q 3 Extremities AIS score, mean (SD) Extremity Q 3 External AIS score, mean (SD) ISS, mean (SD) ISS Q 16
Pulmonary ICU
SICU
n = 1,124
n = 1,129
p
44 (20) 72% 87% 133 (33)
47 (21) 71% 91% 133 (29)
.01 0.56 0.002 0.96
9% 97 (25) 11 (5) 58% 2 (2) 620 (55) 0.46 (0.88) 41 (4) 1 (2) 374 (33) 0.85 (1.4) 204 (18) 1 (1) 267 (24) 0.57 (0.55) 21 (12) 68%
8% 95 (25) 12 (5) 55% 2 (2) 646 (57) 0.43 (0.86) 46 (4) 1 (2) 405 (36) 0.80 (1.4) 188 (17) 1 (1) 246 (22) 0.56 (0.53) 21 (12) 70%
0.24 0.28 0.40 0.18 0.28 0.32 0.42 0.60 0.19 0.19 0.37 0.35 0.29 0.27 0.46 0.88 0.25
ICU patients to SICU patients, there was no difference in mortality (22% vs. 21%, p = 0.87), pulmonary complications (3% vs. 1%, p = 0.31), infectious complications (23% vs. 23%, p = 0.99), cardiac complications (15% vs. 12%, p = 0.38), thromboembolic complications (3% vs. 3%, p = 0.74), ventilator days (3 [6] days vs. 4 [6] days, p = 0.16), or LOS in the ICU (5 [6] days vs. 6 [7] days, p = 0.49) or hospital (9 [8] days vs. 10 [9] days, p = 0.22). However, patients managed by the SICU service less often required reintubation (15% vs. 7%, p = 0.009). In the other subgroup analysis, there were 1, 232 patients who required at least 1 day of mechanical ventilation, 633 were managed by the pulmonary ICU service and 599 managed by the SICU service. While there was no difference in mortality (21% vs. 20%, p = 0.66), patients managed by the SICU service had fewer pulmonary complications (5% vs. 9%, p = 0.002) but no difference in infectious (30% vs. 30%, p = 0.89), cardiac (10% vs. 9%, p = 0.48), or thromboembolic (3% vs. 5%, p = 0.15) complications. The patients managed by the SICU service also had one less day on the ventilator (5 [6] days vs. 6 [6] days, p = 0.02) but no difference in ICU (7 [7] days vs. 7 [7] days, p = 0.35) or hospital (14 [14] days vs. 15 [14] days, p = 0.66) LOS.
DISCUSSION The primary outcome of our study was to look for differences in mortality among trauma patients managed by the pulmonary ICU service versus the SICU service. Despite having an older population with an increased incidence of blunt trauma, there was no difference in mortality between the two services. The patients on the SICU service had a decrease in days
* 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
965
J Trauma Acute Care Surg Volume 77, Number 6
Klein et al.
TABLE 2. Outcomes for Pulmonary ICU and SICU Groups Pulmonary ICU
SICU
n = 1,124
n = 1,129
Mortality Complications Pulmonary Infectious Cardiac Thromboembolic Reintubation Hospital days, mean (SD) ICU days, mean (SD) Ventilator days, mean (SD)
TABLE 4. Population Characteristics for Subgroup of Patients Requiring at Least 1 Day of Mechanical Ventilation
p
13%
11%
0.41
6% 20% 7% 3% 9% 12 (12) 5 (6) 4 (6)
3% 20% 7% 2% 4% 11 (12) 5 (6) 3 (5)
0.0006 0.87 0.89 0.22 G0.0001 0.87 0.39 0.002
on mechanical ventilation, less pulmonary complications, and a decreased rate of failed extubation and reintubation. There were no differences in ICU or hospital LOS, nor were there any differences in infectious, cardiac, or thromboembolic complications. Similar outcomes were seen for both elderly patients and patients requiring at least 24 hours of mechanical ventilation. Compared with previous studies, ours was the first to compare the care of the trauma patient under the management of a surgical intensivist versus a pulmonary intensivist. Previous studies have either compared outcomes of pulmonary intensivists and surgical intensivists in the care of all types of patients (medical and surgical) or examined the care of the trauma patient in different types of ICUs while still having their care guided by a surgical intensivist. TABLE 3. Population Characteristics for Elderly (Q65 Years Old) Subgroup
Age, mean (SD) Male Blunt Admission systolic blood pressure, mean (SD), mm Hg Hypotension (G90 mm Hg) Admission pulse, mean (SD) Admission GCS score, mean (SD) Intubated in ED Head AIS score, mean (SD) Head AIS score Q 3 Face AIS score, mean (SD) Face AIS score Q 3 Chest AIS score, mean (SD) Chest AIS score Q 3 Abdominal AIS score, mean (SD) Abdomen AIS score Q 3 Extremities AIS score, mean (SD) Extremity Q 3 External AIS score, mean (SD) ISS, mean (SD) ISS Q 16
966
Pulmonary ICU
SICU
n = 202
n = 241
p
76 (7) 99 (49) 197 (98) 145 (37)
77 (7) 142 (59) 239 (99) 145 (31)
0.047 0.04 0.25 0.95
15 (7) 89 (21) 13 (4) 104 (51) 3 (2) 117 (58) 0.26 (0.70) 4 (2) 1.1 (1.6) 58 (29) 0.52 (1.1) 17 (8) 1.2 (1.3) 54 (27) 0.51 (0.55) 20 (11) 131 (65)
9 (4) 86 (20) 13 (4) 110 (46) 3 (2) 156 (65) 0.27 (0.73) 7 (3) 1.1 (1.6) 64 (27) 0.44 (0.99) 15 (6) 0.90 (1.2) 46 (19) 0.51 (0.51) 19 (10) 164 (68)
0.09 0.20 0.19 0.22 0.21 0.14 0.96 0.53 0.67 0.61 0.42 0.37 0.03 0.06 0.99 0.43 0.48
Age, mean (SD) Male Blunt Admission systolic blood pressure, mean (SD), mm Hg Hypotension (G90 mm Hg) Admission pulse, mean (SD) Admission GCS score, mean (SD) Intubated in ED Head AIS score, mean (SD) Head AIS score Q 3 Face AIS score, mean (SD) Face AIS score Q 3 Chest AIS score, mean (SD) Chest AIS score Q 3 Abdominal AIS score, mean (SD) Abdomen AIS score Q 3 Extremities AIS score, mean (SD) Extremity Q 3 External AIS score, mean (SD) ISS, mean (SD) ISS Q 16
Pulmonary ICU
SICU
n = 633
n = 599
p
44 (19) 483 (76) 539 (85) 130 (37)
45 (20) 429 (72) 536 (89) 129 (32)
0.57 0.06 0.02 0.60
82 (13) 100 (27) 9 (5) 626 (99) 2.7 (2.1) 377 (60) 0.57 (0.96) 33 (5) 1.6 (1.8) 260 (41) 0.96 (1.5) 125 (20) 1.2 (1.3) 164 (26) 0.58 (0.55) 25 (13) 494 (78)
71 (12) 100 (26) 9 (5) 592 (99) 2.7 (2) 369 (62) 0.54 (0.93) 30 (5) 1.7 (1.8) 255 (43) 0.93 (1.4) 116 (19) 1.1 (1.3) 145 (24) 0.60 (0.53) 25 (13) 458 (77)
0.55 0.93 0.87 0.92 0.68 0.46 0.58 0.87 0.58 0.59 0.71 0.87 0.18 0.49 0.50 0.48 0.54
Across the country, trauma patients are managed in the ICU by a variety of intensivists, in both an open and closed ICU format. There has been a definite trend in recent years toward the use of intensivists trained in critical care for managing patients requiring ICU care and monitoring, both for trauma and nontrauma patients. A large review article examining physician staffing patterns in the ICU found improved outcomes when ICU care is provided by a critical careYcertified physician, including decreased ICU and hospital mortality as well as decreased hospital LOS.2 Another article, however, found that patients managed by critical care physicians had an increased risk of death.3 It is the only article on this topic to describe a negative outcome associated with intensive care physicians. The discrepancy of these results could result from many factors, including available resources and variation in the standard of practice among different hospital systems. Another trend includes the use of ‘‘closed’’ ICUs. There are numerous models that exist across the country. In the closed model, once a patient is transferred to the ICU, all care is provided by a critical care fellowship-trained physician. An open model allows any physician to admit and care for a patient in the ICU. Other varieties exist, including mandatory consultation of a critical care physician when admitting a patient to the ICU.4 For surgical patients, a study found that the closed format resulted in decreased mortality in high-risk surgical patients, even despite statistically significant higher ages and APACHE II scores in the patients admitted to the closed ICU. The closed ICU group also had a higher number of cardiopulmonary complications but had decreased mortality related to them. In this particular study, the total LOS and ICU LOS were both * 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
J Trauma Acute Care Surg Volume 77, Number 6
increased, but this is attributed to the increased severity of these patients.5 Most of these studies have not distinguished between the base training of the intensivists. A 2009 study found that intensivists trained in a 3-year pulmonary and critical care fellowship versus a 1-year critical care fellowship after completion of internal medicine residency or after general surgery, anesthesiology, or emergency medicine had decreased ICU mortality of their patients. The pulmonary-trained physicians performed less invasive procedures, while the internal medicineYcritical care group was more likely to change people to a ‘‘do-not-rescucitate’’ status.6 This article, however, did not comment on the type of patient (surgical/trauma vs. medical), and the surgical intensivist also took care of medical patients. Another study found no significant difference in mortality when comparing outcomes of medicine versus surgical/anesthesiology intensivisists.7 Again, however, the surgical critical care physicians took care of medical patients, and surgical patients received care from medicine critical care physicians; it was simply luck of the draw who was the assigned ICU physician that week. While no studies that examine strictly surgical patients and surgical intensivist exist, this has been studied in neurosurgical ICUs. Mirski et al.8 found decreased mortality, better outcomes, decreased hospital LOS, and decreased additional subspecialty consultation when patients with nontraumatic intracranial hemorrhage were admitted to a neurosurgical ICU staffed by neurosurgeons trained in critical care as well as support staff (nursing, etc.) who were also trained in the care of neurosurgical patients. There was an increase in invasive procedures, such as cerebrospinal fluid draining and emergent craniotomies, which led to improved outcomes. Another study also found decreased mortality for patients with intracranial hemorrhage admitted to the neurosurgical ICU, although they had an increased LOS in the neurosurgical ICU since there were no stepdown options at that institution.9 A 2008 study examined the question, ‘‘Are trauma patients better off in a trauma ICU?’’ The authors suggested that severely injured trauma patients should be treated in a surgical/trauma ICU and have their care guided by surgical critical care physicians.10 However, all trauma patients were ultimately cared for by a surgical intensivist and multidisciplinary team, whether they were admitted to the surgical/trauma ICU, medical ICU, cardiac ICU, or neuroscience ICU. There was no comparison between the type of intensivist caring for the trauma patient. Transition from a pulmonary ICU service to a SICU service at our institution was associated with no change in mortality but an improvement in pulmonary complications, ventilator days, and reintubation rates. Trauma centers currently staffed with a pulmonary or anesthesia ICU service should feel comfortable converting to SICU service and should expect comparable or improved outcomes for trauma patients admitted to the ICU.
AUTHORSHIP C.V.R.B. conceived and designed this study. S.A. and A.C. acquired the data. C.V.R.B., A.L.K., J.A., S.A., and B.C. contributed to the data analysis and interpretation. C.V.R.B., A.L.K., and S.A. drafted the manuscript. C.V.R.B., J.A., and B.C. contributed to the critical revision.
Klein et al.
DISCLOSURE The authors declare no conflicts of interest.
REFERENCES 1. Committee on Trauma, American College of Surgeons. Resources for Optimal Care of the Injured Patient. Chicago, IL: American College of Surgeons; 2006:65Y66. 2. Provonost PJ, Angus DC, Dorman T, Robinson KA, Dremsizov TT, Young TL. Physician staffing patterns and clinical outcomes in critically ill patients: a systematic review. JAMA. 2002;288:2151Y2162. 3. Levy MM, Rapoport J, Lemeshow S, Chalfin DB, Phillips G, Danis M. Association between critical care physician management and patient mortality in the intensive care unit. Ann Intern Med. 2008;148:801Y809. 4. Penkoske PA, Buchman TG. The relationship between the surgeon and the intensivist in the surgical intensive care unit. Surg Clin North Am. 2006: 1351Y1357 5. Van der Sluis F, Slagt C, Liebman B, Beute J, Mulder JW, Engel AF. The impact of open versus closed format ICU admission practices on the outcome of high risk surgical patients: a cohort analysis. BMC Surg. 2011;11:18Y23. 6. Billington EO, Zygun DA, Stelfox HT, Peets AD. Intensivists’ base specialty of training is associated with variations in mortality and practice patterns. Critical Care. 2009;13:R209. 7. Lee J, Iqbal S, Gursahaney A, Nouh T, Khwaja K. Medicine versus surgery/anesthesiology intensivist: a retrospective review and comparison of outcomes in a mixed medical-surgical-trauma ICU. Can J Surg. 2013; 56(4):275Y279. 8. Mirski MA, Chang CW, Cowan R. Impact of a neuroscience intensive care unit on neurosurgical patient outcomes and cost of care. J Neurosurg Anesthesiol. 2001;13(2):83Y92. 9. Diringer MN, Edwards DF. Admission to a neurologic/neurosurgical intensive care unit is associated with reduced mortality rate after intracerebral hemorrhage. Neuro Crit Care. 2001;29(3):635Y640. 10. Aboutanos M, Duane T, Malhotra A, Rao I, Wolfe L. Are trauma patients better off in a trauma ICU? J Emerg Trauma Shock. 2008:1;74Y77.
EDITORIAL CRITIQUE Critical care of trauma patients impacts significantly on outcomes. Nathens et al (Ann Surg 2006;244:545Y54) determined that intensivist-model ICU care was associated with a 36% reduction in hospital mortality following trauma, particularly in elderly patients (68 centers, n=6789). Importantly, this effect was greatest in trauma center ICUs led by surgical intensivists, suggesting the importance of content expertise in care of critically injured patients. There was no benefit (potential harm) in closed ICUs run by non-surgeons or those not certified in critical care. Klein et al. examined trauma patient (n=2253, 4 years) outcomes comparing management by a pulmonary vs. surgical ICU (SICU) service. They used a before/after study design with retrospective analysis and documented no mortality differences and decreased pulmonary complications/ventilator days and reintubation in patients managed by the SICU service. Two subgroup analyses [elderly (age > 65 years) and respiratory failure (requiring 1 day mechanical ventilation)] documented no mortality differences, but decreased reintubation in elderly SICU patients, and decreased pulmonary complications/ ventilator days in respiratory failure SICU patients. Interestingly, they did not perform a separate cohort analysis of the hemorrhagic shock patients, where trauma surgical intensivists may provide improved care compared to pulmonary critical care physicians.
* 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
967
J Trauma Acute Care Surg Volume 77, Number 6
Klein et al.
A number of questions emerge: 1. Were the 4 SICU general/trauma surgeons ABS-certified in surgical critical care? Were all pulmonary physicians ABMS-certified in Critical Care Medicine? 2. Did the SICU attending surgeon have any responsibilities other than the ICU? 3. What was the ICU coverage model for provision of 24/7 care? 4. What specific definitions were used for ‘‘pulmonary, infectious, cardiac or thromboembolic complications’’ as secondary outcome measures? Statistically significant differences in age (younger in pulmonary) in the total cohort and other baseline characteristics
968
(gender, mechanism, injury type) in the subgroup analyses were identified. Since younger age is associated with improved trauma outcomes, logistic regression analysis is required to determine the impact of intensivist model on trauma patient outcomes.The optimal critical care model to improve trauma patient outcomes remains a very important question, particularly since the total number of surgical intensivists in the nation is currently not adequate to staff all trauma ICUs. Lena M. Napolitano, MD Department of Surgery University of Michigan School of Medicine Ann Arbor, Michigan
* 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Implementation of a surgical intensive care unit service is associated with improved outcomes for trauma patients Amanda L. Klein, MD, Carlos V.R. Brown, MD, Jayson Aydelotte, MD, Sadia Ali, MPH, Adam Clark, and Ben Coopwood, MD, Austin, Texas
Our trauma service recently transitioned from a pulmonary intensive care unit (ICU) service to a surgical ICU (SICU) service. We hypothesized that a newly formed SICU service could provide comparable outcomes to the existing pulmonary ICU service. A specific aim of this study was to compare outcomes of trauma patients admitted to the ICU before and after implementation of a SICU service. METHODS: We performed a retrospective study of trauma patients admitted to the ICU of our urban, American College of SurgeonsY verified, Level 1 trauma center during a 4-year period (2009Y2012). Patients managed by the pulmonary ICU service (2009Y2010) were compared with patients managed by a SICU service (2011Y2012). The primary outcome was mortality, while secondary outcomes included complications (pulmonary, infectious, cardiac, and thromboembolic), hospital and ICU length of stay, ventilator days, and need for reintubation. RESULTS: There were 2,253 trauma patients admitted to the ICU during the study period, 1,124 and 1,129 managed by the pulmonary ICU and SICU services, respectively. When comparing outcomes for SICU and pulmonary ICU patients, there was no difference in mortality (11% vs. 13%, p = 0.41), but patients managed by the SICU service had fewer pulmonary complications (3% vs. 6%, p G 0.001), fewer days on the ventilator (3 vs. 4, p = 0.002), and less often required reintubation after extubation (4% vs. 9%, p G 0.001). CONCLUSION: Transition from a pulmonary ICU service to a SICU service at our institution was associated with no change in mortality but an improvement in pulmonary complications, ventilator days, and reintubation rates. Trauma centers currently staffed with a pulmonary ICU service should feel comfortable converting to SICU service and should expect comparable or improved outcomes for trauma patients admitted to the ICU. (J Trauma Acute Care Surg. 2014;77: 964Y968. Copyright * 2014 by Lippincott Williams & Wilkins) LEVEL OF EVIDENCE: Therapeutic/care management study, level IV. KEY WORDS: ICU; trauma; outcomes; intensivist; intesive care. BACKGROUND:
T
rauma patients admitted to an intensive care unit (ICU) may be managed by a variety of intensivists including anesthesiology, pulmonary, and surgical intensivists. However, according to the American College of Surgeons’ Committee on Trauma (ACS COT), Resources for Optimal Care of the Injured Patient,1 ‘‘[t]he trauma surgeon must remain in charge of the patient in the ICU’’ (criteria deficiency 11Y46) and ‘‘[i]n all trauma centers, the trauma service must retain responsibility for the patient and coordinate therapeutic decisions’’ (criteria deficiency 11Y53). The easiest way for a trauma center to meet these ACS COT requirements is to maintain a surgical ICU (SICU) service staffed by the trauma surgeon’s panel. However, many trauma centers may not have an existing SICU service and may require a transition from a medical or anesthesia ICU service to a SICU service. This transition can be met with resistance and skepticism from nurses, doctors, and administrators who have grown accustomed to a medical or anesthesia critical care service. Submitted: January 14, 2014, Revised: May 7, 2014, Accepted: July 28, 2014. From the University Medical Center Brackenridge, University of Texas SouthwesternY Austin, Austin, Texas. This study was presented at the 44th Annual Meeting of the Western Trauma Association, March 2Y7, 2014, in Steamboat Springs, Colorado. Address for reprints: Carlos V.R. Brown, MD, University of Texas SouthwesternV Austin, University Medical Center at Brackenridge, 601 East 15th St, Austin, Texas; email: [email protected]. DOI: 10.1097/TA.0000000000000460
964
From its inception, our trauma center’s ICU was staffed by a highly skilled group of pulmonary intensivists. However, as we transitioned from an ACS-verified Level 2 trauma to Level 1 trauma center, we received feedback from COT Verification Review Committee that our trauma service needed to be more involved in the intensive care of our severely injured patients. We decided that the simplest way to improve the critical care coverage of our severely injured patients was to implement a surgical critical care service staffed by our trauma surgeons. In 2011, our trauma service made the transition from a pulmonary ICU service to a full-time surgical critical care service. We hypothesized that a newly formed SICU service could provide comparable outcomes to the existing pulmonary ICU service. A specific aim of this study was to compare outcomes of trauma patients admitted to the ICU before and after implementation of a SICU service.
PATIENTS AND METHODS We performed a retrospective study of all trauma patients admitted to the ICU at THE University Medical Center Brackenridge, an urban ACS-verified Level 1 trauma center located in Austin, Texas. Our study spanned 4 years (2009Y2012), including two separate 2-year periods. During the first period (2009Y2010), trauma patients admitted to the ICU were managed by the pulmonary ICU service (pulmonary ICU group), J Trauma Acute Care Surg Volume 77, Number 6
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
J Trauma Acute Care Surg Volume 77, Number 6
Klein et al.
composed of 10 rotating pulmonary intensivists providing coverage of trauma ICU patients 24/7. During the second period (2011Y2012), trauma patients admitted to the ICU were managed by the SICU service (SICU group), composed of four rotating general/trauma surgeons covering trauma ICU patients 24/7. Data were extracted from the trauma registry and electronic medical record. Data collected for both periods include admission demographics, mechanism of injury, emergency department (ED) vital signs, Abbreviated Injury Scale (AIS) score for each body region, and Injury Severity Score (ISS). The primary outcome was mortality, while secondary outcomes included complications (pulmonary, infectious, cardiac, and thromboembolic), ventilator days, reintubation rates, as well as length of stay (LOS) in the hospital and ICU. The SICU and pulmonary ICU groups were compared by univariate analysis for all variables using the unpaired Student’s t test or Mann-Whitney rank-sum test for the continuous variables and Pearson’s r2 with Yates correction for categorical variables. Two subgroups were also analyzed, elderly patients (Q65 years old) and patients requiring at least 1 day of mechanical ventilation. Values are reported as mean (SD) or raw percentages, and statistical significance was set at p G 0.05. The local institutional review board approved this study.
RESULTS There were a total of 2,253 trauma patients admitted to our ICU during the 4-year study period, 1,124 from 2009 to 2010 managed by the pulmonary ICU service and 1,129 from 2011 to 2012 managed by the SICU service. Patients managed by the pulmonary ICU service were younger (44 [20] years vs. 47 [21] years, p = 0.01), but there was no difference in male sex (72% vs. 71%, p = 0.56) or white race (64% vs. 63%, p = 0.85). The pulmonary ICU group less often sustained blunt trauma (87% vs. 91%, p = 0.002), but there was no difference in ISS (21 [12] vs. 21 [12], p = 0.88) or AIS score of 3 or greater for any body region: head (55% vs. 57%, p = 0.32), chest (33% vs. 36%, p = 0.19), abdomen (18% vs. 17%, p = 0.35), and extremities (24% vs. 22%, p = 0.27). In the ED, there was no difference in hypotension (9% vs. 8%, p = 0.24), pulse rate (97 [25] vs. 95 [25], p = 0.28), Glasgow Coma Scale (GCS) of 8 or less (27% vs. 26%, p = 0.59), or the need for prehospital or ED endotracheal intubation (58% vs. 55%, p = 0.18) (Tables 1). There was no difference in mortality, 13% for patients managed by the pulmonary ICU service and 11% for patients managed by the SICU service (p = 0.41). Patients managed by the SICU service had fewer pulmonary complications (3% vs. 6%, p G 0.001), but there was no difference in infectious (20% vs. 20%, p = 0.87), cardiac (7% vs. 7%, p = 0.89), or thromboembolic (2% vs. 3%, p = 0.22) complications. Patients managed by the SICU service had fewer ventilator days (3 [5] days vs. 4 [6] days, p = 0.002), but there was no difference in ICU (5 [6] days vs. 5 [6] days, p = 0.39) or hospital (11 [12] days vs. 12 [12] days, p = 0.87) LOS. Despite spending one less day on the ventilator, the patients managed by the SICU service less often failed extubation and required intubation (4% vs. 9%, p G 0.001). Within the subgroup analyses, there were 443 elderly trauma patients, 202 managed by the pulmonary ICU service and 241 managed by the SICU service. When comparing pulmonary
TABLE 1. Population Characteristics for Pulmonary ICU and SICU Groups
Age, mean (SD), y Male sex Blunt mechanism Admission systolic blood pressure, mean (SD), mm Hg Hypotension (G90 mm Hg) Admission pulse, mean (SD) Admission GCS score, mean (SD) Intubated in ED Head AIS score, mean (SD) Head AIS score Q 3 Face AIS score, mean (SD) Face AIS score Q 3 Chest AIS score, mean (SD) Chest AIS score Q 3 Abdominal AIS score, mean (SD) Abdomen AIS score Q 3 Extremities AIS score, mean (SD) Extremity Q 3 External AIS score, mean (SD) ISS, mean (SD) ISS Q 16
Pulmonary ICU
SICU
n = 1,124
n = 1,129
p
44 (20) 72% 87% 133 (33)
47 (21) 71% 91% 133 (29)
.01 0.56 0.002 0.96
9% 97 (25) 11 (5) 58% 2 (2) 620 (55) 0.46 (0.88) 41 (4) 1 (2) 374 (33) 0.85 (1.4) 204 (18) 1 (1) 267 (24) 0.57 (0.55) 21 (12) 68%
8% 95 (25) 12 (5) 55% 2 (2) 646 (57) 0.43 (0.86) 46 (4) 1 (2) 405 (36) 0.80 (1.4) 188 (17) 1 (1) 246 (22) 0.56 (0.53) 21 (12) 70%
0.24 0.28 0.40 0.18 0.28 0.32 0.42 0.60 0.19 0.19 0.37 0.35 0.29 0.27 0.46 0.88 0.25
ICU patients to SICU patients, there was no difference in mortality (22% vs. 21%, p = 0.87), pulmonary complications (3% vs. 1%, p = 0.31), infectious complications (23% vs. 23%, p = 0.99), cardiac complications (15% vs. 12%, p = 0.38), thromboembolic complications (3% vs. 3%, p = 0.74), ventilator days (3 [6] days vs. 4 [6] days, p = 0.16), or LOS in the ICU (5 [6] days vs. 6 [7] days, p = 0.49) or hospital (9 [8] days vs. 10 [9] days, p = 0.22). However, patients managed by the SICU service less often required reintubation (15% vs. 7%, p = 0.009). In the other subgroup analysis, there were 1, 232 patients who required at least 1 day of mechanical ventilation, 633 were managed by the pulmonary ICU service and 599 managed by the SICU service. While there was no difference in mortality (21% vs. 20%, p = 0.66), patients managed by the SICU service had fewer pulmonary complications (5% vs. 9%, p = 0.002) but no difference in infectious (30% vs. 30%, p = 0.89), cardiac (10% vs. 9%, p = 0.48), or thromboembolic (3% vs. 5%, p = 0.15) complications. The patients managed by the SICU service also had one less day on the ventilator (5 [6] days vs. 6 [6] days, p = 0.02) but no difference in ICU (7 [7] days vs. 7 [7] days, p = 0.35) or hospital (14 [14] days vs. 15 [14] days, p = 0.66) LOS.
DISCUSSION The primary outcome of our study was to look for differences in mortality among trauma patients managed by the pulmonary ICU service versus the SICU service. Despite having an older population with an increased incidence of blunt trauma, there was no difference in mortality between the two services. The patients on the SICU service had a decrease in days
* 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
965
J Trauma Acute Care Surg Volume 77, Number 6
Klein et al.
TABLE 2. Outcomes for Pulmonary ICU and SICU Groups Pulmonary ICU
SICU
n = 1,124
n = 1,129
Mortality Complications Pulmonary Infectious Cardiac Thromboembolic Reintubation Hospital days, mean (SD) ICU days, mean (SD) Ventilator days, mean (SD)
TABLE 4. Population Characteristics for Subgroup of Patients Requiring at Least 1 Day of Mechanical Ventilation
p
13%
11%
0.41
6% 20% 7% 3% 9% 12 (12) 5 (6) 4 (6)
3% 20% 7% 2% 4% 11 (12) 5 (6) 3 (5)
0.0006 0.87 0.89 0.22 G0.0001 0.87 0.39 0.002
on mechanical ventilation, less pulmonary complications, and a decreased rate of failed extubation and reintubation. There were no differences in ICU or hospital LOS, nor were there any differences in infectious, cardiac, or thromboembolic complications. Similar outcomes were seen for both elderly patients and patients requiring at least 24 hours of mechanical ventilation. Compared with previous studies, ours was the first to compare the care of the trauma patient under the management of a surgical intensivist versus a pulmonary intensivist. Previous studies have either compared outcomes of pulmonary intensivists and surgical intensivists in the care of all types of patients (medical and surgical) or examined the care of the trauma patient in different types of ICUs while still having their care guided by a surgical intensivist. TABLE 3. Population Characteristics for Elderly (Q65 Years Old) Subgroup
Age, mean (SD) Male Blunt Admission systolic blood pressure, mean (SD), mm Hg Hypotension (G90 mm Hg) Admission pulse, mean (SD) Admission GCS score, mean (SD) Intubated in ED Head AIS score, mean (SD) Head AIS score Q 3 Face AIS score, mean (SD) Face AIS score Q 3 Chest AIS score, mean (SD) Chest AIS score Q 3 Abdominal AIS score, mean (SD) Abdomen AIS score Q 3 Extremities AIS score, mean (SD) Extremity Q 3 External AIS score, mean (SD) ISS, mean (SD) ISS Q 16
966
Pulmonary ICU
SICU
n = 202
n = 241
p
76 (7) 99 (49) 197 (98) 145 (37)
77 (7) 142 (59) 239 (99) 145 (31)
0.047 0.04 0.25 0.95
15 (7) 89 (21) 13 (4) 104 (51) 3 (2) 117 (58) 0.26 (0.70) 4 (2) 1.1 (1.6) 58 (29) 0.52 (1.1) 17 (8) 1.2 (1.3) 54 (27) 0.51 (0.55) 20 (11) 131 (65)
9 (4) 86 (20) 13 (4) 110 (46) 3 (2) 156 (65) 0.27 (0.73) 7 (3) 1.1 (1.6) 64 (27) 0.44 (0.99) 15 (6) 0.90 (1.2) 46 (19) 0.51 (0.51) 19 (10) 164 (68)
0.09 0.20 0.19 0.22 0.21 0.14 0.96 0.53 0.67 0.61 0.42 0.37 0.03 0.06 0.99 0.43 0.48
Age, mean (SD) Male Blunt Admission systolic blood pressure, mean (SD), mm Hg Hypotension (G90 mm Hg) Admission pulse, mean (SD) Admission GCS score, mean (SD) Intubated in ED Head AIS score, mean (SD) Head AIS score Q 3 Face AIS score, mean (SD) Face AIS score Q 3 Chest AIS score, mean (SD) Chest AIS score Q 3 Abdominal AIS score, mean (SD) Abdomen AIS score Q 3 Extremities AIS score, mean (SD) Extremity Q 3 External AIS score, mean (SD) ISS, mean (SD) ISS Q 16
Pulmonary ICU
SICU
n = 633
n = 599
p
44 (19) 483 (76) 539 (85) 130 (37)
45 (20) 429 (72) 536 (89) 129 (32)
0.57 0.06 0.02 0.60
82 (13) 100 (27) 9 (5) 626 (99) 2.7 (2.1) 377 (60) 0.57 (0.96) 33 (5) 1.6 (1.8) 260 (41) 0.96 (1.5) 125 (20) 1.2 (1.3) 164 (26) 0.58 (0.55) 25 (13) 494 (78)
71 (12) 100 (26) 9 (5) 592 (99) 2.7 (2) 369 (62) 0.54 (0.93) 30 (5) 1.7 (1.8) 255 (43) 0.93 (1.4) 116 (19) 1.1 (1.3) 145 (24) 0.60 (0.53) 25 (13) 458 (77)
0.55 0.93 0.87 0.92 0.68 0.46 0.58 0.87 0.58 0.59 0.71 0.87 0.18 0.49 0.50 0.48 0.54
Across the country, trauma patients are managed in the ICU by a variety of intensivists, in both an open and closed ICU format. There has been a definite trend in recent years toward the use of intensivists trained in critical care for managing patients requiring ICU care and monitoring, both for trauma and nontrauma patients. A large review article examining physician staffing patterns in the ICU found improved outcomes when ICU care is provided by a critical careYcertified physician, including decreased ICU and hospital mortality as well as decreased hospital LOS.2 Another article, however, found that patients managed by critical care physicians had an increased risk of death.3 It is the only article on this topic to describe a negative outcome associated with intensive care physicians. The discrepancy of these results could result from many factors, including available resources and variation in the standard of practice among different hospital systems. Another trend includes the use of ‘‘closed’’ ICUs. There are numerous models that exist across the country. In the closed model, once a patient is transferred to the ICU, all care is provided by a critical care fellowship-trained physician. An open model allows any physician to admit and care for a patient in the ICU. Other varieties exist, including mandatory consultation of a critical care physician when admitting a patient to the ICU.4 For surgical patients, a study found that the closed format resulted in decreased mortality in high-risk surgical patients, even despite statistically significant higher ages and APACHE II scores in the patients admitted to the closed ICU. The closed ICU group also had a higher number of cardiopulmonary complications but had decreased mortality related to them. In this particular study, the total LOS and ICU LOS were both * 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
J Trauma Acute Care Surg Volume 77, Number 6
increased, but this is attributed to the increased severity of these patients.5 Most of these studies have not distinguished between the base training of the intensivists. A 2009 study found that intensivists trained in a 3-year pulmonary and critical care fellowship versus a 1-year critical care fellowship after completion of internal medicine residency or after general surgery, anesthesiology, or emergency medicine had decreased ICU mortality of their patients. The pulmonary-trained physicians performed less invasive procedures, while the internal medicineYcritical care group was more likely to change people to a ‘‘do-not-rescucitate’’ status.6 This article, however, did not comment on the type of patient (surgical/trauma vs. medical), and the surgical intensivist also took care of medical patients. Another study found no significant difference in mortality when comparing outcomes of medicine versus surgical/anesthesiology intensivisists.7 Again, however, the surgical critical care physicians took care of medical patients, and surgical patients received care from medicine critical care physicians; it was simply luck of the draw who was the assigned ICU physician that week. While no studies that examine strictly surgical patients and surgical intensivist exist, this has been studied in neurosurgical ICUs. Mirski et al.8 found decreased mortality, better outcomes, decreased hospital LOS, and decreased additional subspecialty consultation when patients with nontraumatic intracranial hemorrhage were admitted to a neurosurgical ICU staffed by neurosurgeons trained in critical care as well as support staff (nursing, etc.) who were also trained in the care of neurosurgical patients. There was an increase in invasive procedures, such as cerebrospinal fluid draining and emergent craniotomies, which led to improved outcomes. Another study also found decreased mortality for patients with intracranial hemorrhage admitted to the neurosurgical ICU, although they had an increased LOS in the neurosurgical ICU since there were no stepdown options at that institution.9 A 2008 study examined the question, ‘‘Are trauma patients better off in a trauma ICU?’’ The authors suggested that severely injured trauma patients should be treated in a surgical/trauma ICU and have their care guided by surgical critical care physicians.10 However, all trauma patients were ultimately cared for by a surgical intensivist and multidisciplinary team, whether they were admitted to the surgical/trauma ICU, medical ICU, cardiac ICU, or neuroscience ICU. There was no comparison between the type of intensivist caring for the trauma patient. Transition from a pulmonary ICU service to a SICU service at our institution was associated with no change in mortality but an improvement in pulmonary complications, ventilator days, and reintubation rates. Trauma centers currently staffed with a pulmonary or anesthesia ICU service should feel comfortable converting to SICU service and should expect comparable or improved outcomes for trauma patients admitted to the ICU.
AUTHORSHIP C.V.R.B. conceived and designed this study. S.A. and A.C. acquired the data. C.V.R.B., A.L.K., J.A., S.A., and B.C. contributed to the data analysis and interpretation. C.V.R.B., A.L.K., and S.A. drafted the manuscript. C.V.R.B., J.A., and B.C. contributed to the critical revision.
Klein et al.
DISCLOSURE The authors declare no conflicts of interest.
REFERENCES 1. Committee on Trauma, American College of Surgeons. Resources for Optimal Care of the Injured Patient. Chicago, IL: American College of Surgeons; 2006:65Y66. 2. Provonost PJ, Angus DC, Dorman T, Robinson KA, Dremsizov TT, Young TL. Physician staffing patterns and clinical outcomes in critically ill patients: a systematic review. JAMA. 2002;288:2151Y2162. 3. Levy MM, Rapoport J, Lemeshow S, Chalfin DB, Phillips G, Danis M. Association between critical care physician management and patient mortality in the intensive care unit. Ann Intern Med. 2008;148:801Y809. 4. Penkoske PA, Buchman TG. The relationship between the surgeon and the intensivist in the surgical intensive care unit. Surg Clin North Am. 2006: 1351Y1357 5. Van der Sluis F, Slagt C, Liebman B, Beute J, Mulder JW, Engel AF. The impact of open versus closed format ICU admission practices on the outcome of high risk surgical patients: a cohort analysis. BMC Surg. 2011;11:18Y23. 6. Billington EO, Zygun DA, Stelfox HT, Peets AD. Intensivists’ base specialty of training is associated with variations in mortality and practice patterns. Critical Care. 2009;13:R209. 7. Lee J, Iqbal S, Gursahaney A, Nouh T, Khwaja K. Medicine versus surgery/anesthesiology intensivist: a retrospective review and comparison of outcomes in a mixed medical-surgical-trauma ICU. Can J Surg. 2013; 56(4):275Y279. 8. Mirski MA, Chang CW, Cowan R. Impact of a neuroscience intensive care unit on neurosurgical patient outcomes and cost of care. J Neurosurg Anesthesiol. 2001;13(2):83Y92. 9. Diringer MN, Edwards DF. Admission to a neurologic/neurosurgical intensive care unit is associated with reduced mortality rate after intracerebral hemorrhage. Neuro Crit Care. 2001;29(3):635Y640. 10. Aboutanos M, Duane T, Malhotra A, Rao I, Wolfe L. Are trauma patients better off in a trauma ICU? J Emerg Trauma Shock. 2008:1;74Y77.
EDITORIAL CRITIQUE Critical care of trauma patients impacts significantly on outcomes. Nathens et al (Ann Surg 2006;244:545Y54) determined that intensivist-model ICU care was associated with a 36% reduction in hospital mortality following trauma, particularly in elderly patients (68 centers, n=6789). Importantly, this effect was greatest in trauma center ICUs led by surgical intensivists, suggesting the importance of content expertise in care of critically injured patients. There was no benefit (potential harm) in closed ICUs run by non-surgeons or those not certified in critical care. Klein et al. examined trauma patient (n=2253, 4 years) outcomes comparing management by a pulmonary vs. surgical ICU (SICU) service. They used a before/after study design with retrospective analysis and documented no mortality differences and decreased pulmonary complications/ventilator days and reintubation in patients managed by the SICU service. Two subgroup analyses [elderly (age > 65 years) and respiratory failure (requiring 1 day mechanical ventilation)] documented no mortality differences, but decreased reintubation in elderly SICU patients, and decreased pulmonary complications/ ventilator days in respiratory failure SICU patients. Interestingly, they did not perform a separate cohort analysis of the hemorrhagic shock patients, where trauma surgical intensivists may provide improved care compared to pulmonary critical care physicians.
* 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
967
J Trauma Acute Care Surg Volume 77, Number 6
Klein et al.
A number of questions emerge: 1. Were the 4 SICU general/trauma surgeons ABS-certified in surgical critical care? Were all pulmonary physicians ABMS-certified in Critical Care Medicine? 2. Did the SICU attending surgeon have any responsibilities other than the ICU? 3. What was the ICU coverage model for provision of 24/7 care? 4. What specific definitions were used for ‘‘pulmonary, infectious, cardiac or thromboembolic complications’’ as secondary outcome measures? Statistically significant differences in age (younger in pulmonary) in the total cohort and other baseline characteristics
968
(gender, mechanism, injury type) in the subgroup analyses were identified. Since younger age is associated with improved trauma outcomes, logistic regression analysis is required to determine the impact of intensivist model on trauma patient outcomes.The optimal critical care model to improve trauma patient outcomes remains a very important question, particularly since the total number of surgical intensivists in the nation is currently not adequate to staff all trauma ICUs. Lena M. Napolitano, MD Department of Surgery University of Michigan School of Medicine Ann Arbor, Michigan
* 2014 Lippincott Williams & Wilkins
Copyright © 2014 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
