- Email: [email protected]
Early tracheostomy versus late tracheostomy in the surgical intensive care unit
The American Journal of Surgery 189 (2005) 293–296
Scientific paper
Early tracheostomy versus late tracheostomy in the surgical intensive care unit Mecker G. Möller, M.D.a, Jason D. Slaikeu, M.D.a, Pablo Bonelli, M.D., M.B.A.b, Alan T. Davis, Ph.D.a,c, James E. Hoogeboom, D.O.a, Bruce W. Bonnell, M.D.a,* a
Grand Rapids/Michigan State University General Surgery Residency, Grand Rapids Medical Education and Research Center for the Health Professions, 221 Michigan St. N.E., Ste. 200-A, Grand Rapids, MI 49503, USA b Quality Department, Spectrum Health, Grand Rapids, MI, USA c Departments of Surgery, Michigan State University and Spectrum Health, Grand Rapids, MI, USA Manuscript received September 14, 2004; revised manuscript November 19, 2004 Presented at the 47th Annual Meeting of the Midwest Surgical Association, Mackinac Island, Michigan, August 15–18, 2004
Abstract Background: This study’s purpose was to determine if early tracheostomy (ET) of severely injured patients reduces days of ventilatory support, the frequency of ventilator-associated pneumonia (VAP), and surgical intensive care unit (SICU) length of stay (LOS). Methods: This 2-year retrospective review included 185 SICU patients with acute injuries requiring mechanical ventilation and tracheostomy. ET was defined as 7 days or less, and late tracheostomy (LT) as more than 7 days. Results: The incidence of VAP was significantly higher in the LT group, relative to the ET group (42.3% vs. 27.2%, respectively; P ⬍.05). Acute Physiology and Chronic Health Evaluation II scores, hospital and SICU LOS, and the number of ventilator days were significantly higher in the LT group. Conclusions: In patients who required prolonged mechanical ventilation, there was significant decreased incidence of VAP, less ventilator time, and lower ICU LOS when tracheostomy was performed within 7 days after admission to the SICU. © 2005 Excerpta Medica Inc. All rights reserved. Keywords: Early tracheostomy; Surgical ICU; VAP; Timing of tracheostomy
Recent studies have shown that percutaneous tracheostomy is a safe bedside procedure in trained hands [1]. However, consensus for timing of tracheostomy in the critically ill patient has not been reached [2]. Multiple researchers have advocated for the use of percutaneous tracheostomy as an early procedure in the management of patients with severe traumatic brain injury (Glasgow Coma Scale [GCS] ⬍7) [3,4], high spinal cord injuries, and significant maxillofacial trauma [2,3]. Studies have also shown that patients who undergo early tracheostomy (ET; 5–7 days after endotracheal intubation) have shorter lengths of stay in the surgical intensive care unit (SICU) compared to patients who received extubation trials before tracheostomy [5,6]. These patients also had a lower frequency of pneumonias [7]. In * Corresponding author. Tel.: ⫹1-616-391-1405; fax: ⫹1-616-3918613. E-mail address: [email protected]
addition, data also support successful weaning of mechanical ventilation within 48 hours of tracheostomy [5]. The purpose of this study was to determine if ET (defined as tracheostomy within 7 days of admission to the SICU) is associated with diminished incidence of ventilator-associated pneumonia (VAP), reduced days of ventilatory support, and SICU or hospital length of stay (LOS).
Methods This project was designed as a retrospective study from 2000 through 2002 of SICU patients from a tertiary referral hospital, which included a level I trauma center. The study group included 185 patients, ages 16 to 80 years, who were admitted to the SICU by the trauma, general surgery, cardiothoracic, neurology, and neurosurgery services. All of the patients had acute injuries requiring mechanical venti-
0002-9610/05/$ – see front matter © 2005 Excerpta Medica Inc. All rights reserved. doi:10.1016/j.amjsurg.2005.01.002
294
M.G. Möller et al. / The American Journal of Surgery 189 (2005) 293–296
lation and tracheostomy. A percutaneous or open tracheostomy procedure was performed, based on the surgeon’s preferences. For the purposes of this study, ET was defined as 7 days or less, and late tracheostomy (LT) as more than 7 days. Patients who died during the first 48 hours were excluded from the study. Patient data were obtained from a database generated by Project IMPACT, the critical care data system sponsored by the Society of Critical Care Medicine. The data collected included age, gender, Acute Physiology and Chronic Health Evaluation (APACHE) II scores, previous known comorbidities, including chronic obstructive pulmonary disease (COPD), diabetes, chronic renal failure (CRF), and congestive heart failure (CHF), and morbidities, including acute respiratory disease syndromes (ARDS) and acute lung injury (ALI), timing of tracheostomy, termination of ventilator support, VAP ventilator days, and SICU and hospital LOS. Quantitative data are listed as means ⫾ SEM, with differences between groups analyzed with the 2-tailed, unpaired t test. Nominal data were analyzed with the 2 test, except for COPD, which was analyzed with the 2-tailed Fisher exact test. Logistic regression analysis was run with VAP as the dependent variable, and age, gender, ET versus LT, presence of comorbidity, and trauma service admission as the independent variables. For the purpose of the multiple regression analyses, ET versus LT was defined with break points at 7 days, as described above, as well as with break points at 3, 4, 5, and 6 days. Odds ratios generated from multiple regression analysis are reported along with 95% confidence intervals. Significance was assessed at P ⬍0.05. Statistical analyses were run with NCSS 2004 (Number Cruncher Statistical Systems, Kaysville, UT). Table 1 Patient characteristics* Variable
Early tracheostomy
Late tracheostomy
No. of patients Age (y) Male:female (% male) Admitting service Cardiac/thoracic surgery General Surgery Neurology Neurosurgery Trauma Vascular surgery COPD (%) Diabetes (%) CHF (%) CRF (%) APACHE II (%)§
81 49.2 ⫾ 2.2 54:27 (67%)
104 54.6 ⫾ 1.9 61:43 (59%)
4 (5%)† 9 (11%) 5 (6%) 6 (7%) 53 (65%)† 4 (5%) 10 (12%)† 5 (6%) 0 (0%) 4 (5%) 21.7 ⫾ 0.9†
21 (20%)† 16 (15%) 6 (6%) 3 (3%) 51 (49%)† 7 (7%) 4 (4%)† 2 (2%) 4 (4%) 4 (4%) 24.0 ⫾ 0.8†
* Age and APACHE II listed as mean ⫾ SEM. The ET group had a tracheostomy at 7 days or earlier. COPD ⫽ chronic obstructive pulmonary disease; CHF ⫽ congestive heart failure; CRF ⫽ chronic renal failure. † Significant difference between groups, P ⬍ 0.05. § APACHE II values not available for all subjects. Sample size for the ET group ⫽ 72; sample size for the LT group ⫽ 82.
Table 2 Outcome data* Variable
Early tracheostomy
Late tracheostomy
Time to tracheostomy (d) Hospital length of stay (d) ICU LOS (d) Ventilator days VAP ARDS Lung injury
4.3 ⫾ 0.2† 23.8 ⫾ 1.2† 16.7 ⫾ 1.0† 12.2 ⫾ 0.9† 22 (27%)† 2 (3%) 17 (21%)
12.7 ⫾ 0.4† 33.4 ⫾ 1.7† 26.0 ⫾ 1.3† 21.9 ⫾ 1.3† 44 (42%)† 7 (7%) 28 (27%)
* Timed data listed as mean ⫾ SEM. The ET group had a tracheostomy at 7 days or earlier. VAP ⫽ ventilator-associated pneumonia; ARDS ⫽ acute respiratory disease syndrome. † Significant difference between groups, P ⬍ .05.
Results The demographic data for the patients in the 2 groups are shown in Table 1. For the entire data set of 185 patients, the age was 52.2 ⫾ 1.5 years, and 62% of the subjects were males. More than 50% of the admissions were from the trauma service, and 16% of the subjects had at least 1 of the following morbidities: COPD, diabetes, CHF, or CRF. The mean APACHE II score was 22.9 ⫾ 0.6. The APACHE II score was significantly higher in the LT group, relative to patients in the ET group. A significantly higher percentage of patients in the LT group were admitted from the cardio/ thoracic service, whereas a smaller percentage from the same group came from the trauma service, relative to the ET group. In addition, a higher proportion of the ET patients had COPD, relative to the LT group. The outcome data are listed on Table 2. Individuals in the ET group had significantly lower values for VAP, ventilator days, and hospital and ICU LOS, relative to individuals in the LT group. Odds ratios generated from the logistic regression analyses are shown in Table 3. The initial analysis was run using a cutoff of 7 days to distinguish between ET and LT. Additional analyses were run, with the only change being made to the time of the cutoff to distinguish between ET and LT. In all 5 regression analyses, the only significant predictor was time to tracheostomy. The odds ratios ranged from 2.26 using a cutoff of 7 days to an odds ratio of 3.89 using a cutoff of 3 days.
Comments Tracheostomy has an important role in the airway management of ICU patients [8]. Several studies [1,9 –12] have identified the benefits of tracheostomy over endotracheal intubation, such as sparing further injury from translaryngeal intubation, providing a stable airway, facilitating pulmonary toilet, increasing patient comfort and mobility, permitting speech and feedings, and facilitating weaning from the ventilator [5]. Despite several studies [3–7,13–18] advocating ET in the surgical critically ill patient, the timing
M.G. Möller et al. / The American Journal of Surgery 189 (2005) 293–296 Table 3 Odds ratios generated from logistic regression analyses* Cutoff point
Odds ratio (95% CI)
7 6 5 4 3
2.3 (1.2–4.4) 2.7 (1.3–5.5) 3.0 (1.4–6.5) 3.0 (1.3–7.0) 3.9 (1.2–12.2)
days days days days days
* Logistic regression was run with VAP as the dependent variable, and age, gender, admission from the trauma service, comorbidity, and ET vs. LT as the independent variables. Data from Tables 1 and 2 assumed that ET was performed within 7 days of admission to the SICU, while LT was performed later than 7 days. For the purposes of this table, 7 days is the cutoff point. Different cutoff points were used to generate the odds ratios depicted here. 95% CI ⫽ 95% confidence interval.
of tracheostomy continues to be a topic of controversy among surgeons [1,2,19,20]. Historically [2], it has been recommended to wait up to 3 weeks to convert endotracheal intubation to a tracheostomy in order to avoid an unnecessary procedure. However, it has become clear that there are significant risks associated with prolonged endotracheal intubation, such as permanent laryngeal stenosis [21]. ET has been advocated for a number of reasons. There is a reduction in dead space [1], work of breathing, and airway resistance [10]. In addition, ET enables patients to be weaned more rapidly from the ventilator [5,14], leads to decreased tracheobronchial colonization, and hastens recovery from VAP [14]. In those patients with decreased mental status, it provides airway protection [3]. It is also important to take into consideration the economic benefit of ET, which has been shown to decrease the use of resources, without an adverse effect on outcome [3]. In the current study, the ET and LT groups were fairly similar in terms of demographics and pre-admission comorbidities, except for the presence of higher number of patients with COPD in the ET group. This could be explained by the tendency to expect that those patients will do worse on mechanical ventilation, leaning the surgeon towards the decision to perform an ET. Although the difference of COPD was statistically significant, the sample was small and did not correlate with the presence of prolonged mechanical ventilation or development of VAP. In addition to the higher percentage of COPD patients, there was a significantly higher percentage of trauma patients in the early group, while there was a higher percentage of cardiovascular patients in the late group. The disparity between these groups could be a source of bias. However, logistic regression analysis did not determine trauma status to be a significant predictor for incidence of VAP. The presence of preadmission comorbidities and ARDS did not correlate to the development of VAP, and did not predict prolonged mechanical ventilation. There was a significant difference in the timing of tracheostomy between the ET group and the LT group (4.3 ⫾
295
0.2 days vs. 12.7 ⫾ 0.4 days, respectively). Due to the intrinsic limitations of a retrospective study, we were unable to identify those criteria used by the particular surgeon in order to proceed with the clinical decision of performing a tracheostomy. This decision is usually done under the clinical setting of multiple trauma, severe head injury, and the likelihood of significant disability and prolonged recovery from those injuries [2]. However, some authors have identified objective predictors of prolonged mechanical ventilation for more than 14 days, such as an alveolar–arterial oxygen gradient ⱖ175 mmHg (without COPD) and GCS less than 9 at 48 hours of admission as reported by Johnson et al [15], with a positive predictive value of 91% and a negative predictive value of 96%. Gurkin et al [17] identified 2 admitting characteristics in patients with traumatic brain injury who reman intubated at day 7 with a GCS ⱕ 8 and an injury severity score ⱖ 25. Major et al [16] recommended using GCS less than 7 and simplified acute physiologic scores (SAPS) greater than 15, on ICU day 4, in patients with blunt head injury, as their criteria for selecting those patients who should undergo tracheostomy as soon as they can tolerate the procedure. Those with GCS greater than 7 and SAPS less than 13 will likely avoid the procedure. In our study, the APACHE II score was significantly higher in the LT group, relative to patients in the ET group. However, it could be argued that although a statistically significant difference was noted between groups, this does not necessarily represent a clinically relevant difference between the 2 groups. In this study, the incidence of VAP was significantly higher in the LT group (42.3% vs. 27.2% ET group). Our results are similar to those reported by Rodriguez et al [6], who found that ET (⬍7 days) reduced ventilator days, ICU LOS, and hospital LOS. However, they only found a significant difference in VAP for a small group of their patients who had a tracheostomy within the first 48 hours, but not significant differences in the decrease of VAP for those patients who underwent tracheostomy between days 3 and 7. In our study, the odds ratio for developing VAP related to the timing of tracheostomy was significant at all days between 3 and 7. This suggests that ET may provide a benefit in decreasing VAP, even if done earlier than the 7th day, in the patient who is going to require prolonged mechanical ventilation. The most recent prospective randomized study of medically ill patients by Rumbak et al [18], comparing ET within 48 hours versus LT at days 14 to 16, demonstrated that ET has advantages over delayed tracheostomy. Their data showed decreased VAP, time on mechanical ventilation, and ICU LOS in the ET group. Under the limitations of this retrospective review of surgically ill patients, our study adds new data showing the benefit of performing ET in those patients who will need mechanical ventilation or airway protection for more than 1 week. The consideration for earlier tracheostomy should be given as early as day 3, to significantly decrease the risk of
296
M.G. Möller et al. / The American Journal of Surgery 189 (2005) 293–296
VAP. However, more randomized prospective studies are needed to obtain consistent objective data for the early identification of those surgical patients during their ICU course, in order to assist the surgeon with this important decision.
Conclusions Our results reinforce the findings of previous studies showing that ET decreases the incidence of VAP, ventilator time, ICU LOS, and hospital LOS. In addition, our study adds new evidence that earlier tracheostomy may result in greater reduction of VAP when prolonged mechanical ventilation is expected. In those patients who will require mechanical ventilation for more than 1 week, we recommend that tracheostomy be performed between day 3 and 7, in order to obtain greater benefit. More randomized prospective studies are needed to objectively identify those patients. Once the decision has been made to perform a tracheostomy, this procedure should not be delayed due to the hope that an operation could be avoided by waiting.
Acknowledgments The authors wish to express their gratitude for the assistance of Colleen Clonan, R.N. (Database Coordinator, Quality Department, Spectrum Health, Grand Rapids, MI 49503, USA) in the acquisition of the data for the manuscript.
References [1] Vallverdu I, Mancebo J. Approach to patients who fail initial weaning trials. Respir Care Clin North Am 2000;6:365–384. [2] Pryor JP, Reilly PM, Shapiro MB. Surgical airway management in the intensive care unit. Crit Care Clin 2000;16:473– 478. [3] Qureshi AI, Suarez JI, Parekh PD, et al. Prediction and timing of tracheostomy in patients with infratentorial lesions requiring mechanical ventilatory support. Crit Care Med 2000;28:1383–1387.
[4] Lanza DC, Parnes SM, Koltai PJ, et al. Early complications of airway management in head-injured patients. Laryngoscope 1990;100:958 – 961. [5] Koh WY, Lew TW, Chin NM, et al. Tracheostomy in a neurointensive care setting: indications and timing. Anesth Intensive Care 1997;25:365–368. [6] Rodriguez JL, Steinberg SM, Luchetti FA, et al. Early tracheostomy for primary airway management in the surgical critical care setting. Surgery 1990;108:655– 659. [7] Kluger Y, Paul DB, Lucke J, et al. Early tracheostomy in trauma patients. Eur J Emerg Med 1996;3:95–101. [8] Eló G, Pénzes I. Role of percutaneous tracheostomy in intensive care: a review. Orv Hetil 2002;143: 875– 879. [9] Caruso DM, Al-Kasspooles MF, Matthews MR, et al. Rational for ‘early’ percutaneous dilatational tracheostomy in patients with burn injuries. J Burn Care Rehabil 1997;18:424 – 427. [10] Diehl JL, El Atrous S, Touchard D, et al. Changes in the work of breathing induced by tracheotomy in ventilator-dependent patients. Am J Respir Crit Care Med 1999;159:383–388. [11] Wood DE: Tracheostomy. Chest Surg Clin North Am 1996;6:749 – 764. [12] D’Amelio LF, Hammond JS, Spain DA, et al. Tracheostomy and percutaneous endoscopic gastrostomy in the management of the headinjured trauma patient. Am Surgeon 1994;60:180 –185. [13] Armstrong PA, McCarthy MC, Peoples JB. Reduced use of resources by early tracheostomy in ventilator-dependent patients with blunt trauma. Surgery 1998;124:763–767. [14] Teoh WHL, Goh KYC, Chan C. The role of early tracheostomy in critically ill neurosurgical patients. Ann Acad Med Singapore 2001; 30:234 –238. [15] Johnson SB, Kearney PA, Barker DE. Early criteria predictive of prolonged mechanical ventilation. J Trauma 1992;33:95–100. [16] Major KM, Hui T, Wilson MT, et al. Objective indications for early tracheostomy after blunt head trauma. Am J Surg 2003;186:615– 619. [17] Gurkin SA, Parikshak M, Kralovich KA, et al. Indicators for tracheostomy in patients with traumatic brain injury. Am Surg 2002;68: 324 –329. [18] Rumbak MJ, Newton M, Truncale T, et al. A prospective, randomized, study comparing early percutaneous dilational tracheotomy to prolonged translaryngeal intubation (delayed tracheotomy) in critically ill medical patients. Crit Care Med 2004;32:1689 –1694. [19] Gibbons KJ. Tracheostomy: timing is everything. Crit Care Med 2000;28:1663–1664. [20] Heffner JE. Tracheotomy application and timing. Clin Chest Med 2003;24:389 –398. [21] Whited RE. A prospective study of laryngotracheal sequelae in long term intubation. Laryngoscope 1984;94:67–77.
Scientific paper
Early tracheostomy versus late tracheostomy in the surgical intensive care unit Mecker G. Möller, M.D.a, Jason D. Slaikeu, M.D.a, Pablo Bonelli, M.D., M.B.A.b, Alan T. Davis, Ph.D.a,c, James E. Hoogeboom, D.O.a, Bruce W. Bonnell, M.D.a,* a
Grand Rapids/Michigan State University General Surgery Residency, Grand Rapids Medical Education and Research Center for the Health Professions, 221 Michigan St. N.E., Ste. 200-A, Grand Rapids, MI 49503, USA b Quality Department, Spectrum Health, Grand Rapids, MI, USA c Departments of Surgery, Michigan State University and Spectrum Health, Grand Rapids, MI, USA Manuscript received September 14, 2004; revised manuscript November 19, 2004 Presented at the 47th Annual Meeting of the Midwest Surgical Association, Mackinac Island, Michigan, August 15–18, 2004
Abstract Background: This study’s purpose was to determine if early tracheostomy (ET) of severely injured patients reduces days of ventilatory support, the frequency of ventilator-associated pneumonia (VAP), and surgical intensive care unit (SICU) length of stay (LOS). Methods: This 2-year retrospective review included 185 SICU patients with acute injuries requiring mechanical ventilation and tracheostomy. ET was defined as 7 days or less, and late tracheostomy (LT) as more than 7 days. Results: The incidence of VAP was significantly higher in the LT group, relative to the ET group (42.3% vs. 27.2%, respectively; P ⬍.05). Acute Physiology and Chronic Health Evaluation II scores, hospital and SICU LOS, and the number of ventilator days were significantly higher in the LT group. Conclusions: In patients who required prolonged mechanical ventilation, there was significant decreased incidence of VAP, less ventilator time, and lower ICU LOS when tracheostomy was performed within 7 days after admission to the SICU. © 2005 Excerpta Medica Inc. All rights reserved. Keywords: Early tracheostomy; Surgical ICU; VAP; Timing of tracheostomy
Recent studies have shown that percutaneous tracheostomy is a safe bedside procedure in trained hands [1]. However, consensus for timing of tracheostomy in the critically ill patient has not been reached [2]. Multiple researchers have advocated for the use of percutaneous tracheostomy as an early procedure in the management of patients with severe traumatic brain injury (Glasgow Coma Scale [GCS] ⬍7) [3,4], high spinal cord injuries, and significant maxillofacial trauma [2,3]. Studies have also shown that patients who undergo early tracheostomy (ET; 5–7 days after endotracheal intubation) have shorter lengths of stay in the surgical intensive care unit (SICU) compared to patients who received extubation trials before tracheostomy [5,6]. These patients also had a lower frequency of pneumonias [7]. In * Corresponding author. Tel.: ⫹1-616-391-1405; fax: ⫹1-616-3918613. E-mail address: [email protected]
addition, data also support successful weaning of mechanical ventilation within 48 hours of tracheostomy [5]. The purpose of this study was to determine if ET (defined as tracheostomy within 7 days of admission to the SICU) is associated with diminished incidence of ventilator-associated pneumonia (VAP), reduced days of ventilatory support, and SICU or hospital length of stay (LOS).
Methods This project was designed as a retrospective study from 2000 through 2002 of SICU patients from a tertiary referral hospital, which included a level I trauma center. The study group included 185 patients, ages 16 to 80 years, who were admitted to the SICU by the trauma, general surgery, cardiothoracic, neurology, and neurosurgery services. All of the patients had acute injuries requiring mechanical venti-
0002-9610/05/$ – see front matter © 2005 Excerpta Medica Inc. All rights reserved. doi:10.1016/j.amjsurg.2005.01.002
294
M.G. Möller et al. / The American Journal of Surgery 189 (2005) 293–296
lation and tracheostomy. A percutaneous or open tracheostomy procedure was performed, based on the surgeon’s preferences. For the purposes of this study, ET was defined as 7 days or less, and late tracheostomy (LT) as more than 7 days. Patients who died during the first 48 hours were excluded from the study. Patient data were obtained from a database generated by Project IMPACT, the critical care data system sponsored by the Society of Critical Care Medicine. The data collected included age, gender, Acute Physiology and Chronic Health Evaluation (APACHE) II scores, previous known comorbidities, including chronic obstructive pulmonary disease (COPD), diabetes, chronic renal failure (CRF), and congestive heart failure (CHF), and morbidities, including acute respiratory disease syndromes (ARDS) and acute lung injury (ALI), timing of tracheostomy, termination of ventilator support, VAP ventilator days, and SICU and hospital LOS. Quantitative data are listed as means ⫾ SEM, with differences between groups analyzed with the 2-tailed, unpaired t test. Nominal data were analyzed with the 2 test, except for COPD, which was analyzed with the 2-tailed Fisher exact test. Logistic regression analysis was run with VAP as the dependent variable, and age, gender, ET versus LT, presence of comorbidity, and trauma service admission as the independent variables. For the purpose of the multiple regression analyses, ET versus LT was defined with break points at 7 days, as described above, as well as with break points at 3, 4, 5, and 6 days. Odds ratios generated from multiple regression analysis are reported along with 95% confidence intervals. Significance was assessed at P ⬍0.05. Statistical analyses were run with NCSS 2004 (Number Cruncher Statistical Systems, Kaysville, UT). Table 1 Patient characteristics* Variable
Early tracheostomy
Late tracheostomy
No. of patients Age (y) Male:female (% male) Admitting service Cardiac/thoracic surgery General Surgery Neurology Neurosurgery Trauma Vascular surgery COPD (%) Diabetes (%) CHF (%) CRF (%) APACHE II (%)§
81 49.2 ⫾ 2.2 54:27 (67%)
104 54.6 ⫾ 1.9 61:43 (59%)
4 (5%)† 9 (11%) 5 (6%) 6 (7%) 53 (65%)† 4 (5%) 10 (12%)† 5 (6%) 0 (0%) 4 (5%) 21.7 ⫾ 0.9†
21 (20%)† 16 (15%) 6 (6%) 3 (3%) 51 (49%)† 7 (7%) 4 (4%)† 2 (2%) 4 (4%) 4 (4%) 24.0 ⫾ 0.8†
* Age and APACHE II listed as mean ⫾ SEM. The ET group had a tracheostomy at 7 days or earlier. COPD ⫽ chronic obstructive pulmonary disease; CHF ⫽ congestive heart failure; CRF ⫽ chronic renal failure. † Significant difference between groups, P ⬍ 0.05. § APACHE II values not available for all subjects. Sample size for the ET group ⫽ 72; sample size for the LT group ⫽ 82.
Table 2 Outcome data* Variable
Early tracheostomy
Late tracheostomy
Time to tracheostomy (d) Hospital length of stay (d) ICU LOS (d) Ventilator days VAP ARDS Lung injury
4.3 ⫾ 0.2† 23.8 ⫾ 1.2† 16.7 ⫾ 1.0† 12.2 ⫾ 0.9† 22 (27%)† 2 (3%) 17 (21%)
12.7 ⫾ 0.4† 33.4 ⫾ 1.7† 26.0 ⫾ 1.3† 21.9 ⫾ 1.3† 44 (42%)† 7 (7%) 28 (27%)
* Timed data listed as mean ⫾ SEM. The ET group had a tracheostomy at 7 days or earlier. VAP ⫽ ventilator-associated pneumonia; ARDS ⫽ acute respiratory disease syndrome. † Significant difference between groups, P ⬍ .05.
Results The demographic data for the patients in the 2 groups are shown in Table 1. For the entire data set of 185 patients, the age was 52.2 ⫾ 1.5 years, and 62% of the subjects were males. More than 50% of the admissions were from the trauma service, and 16% of the subjects had at least 1 of the following morbidities: COPD, diabetes, CHF, or CRF. The mean APACHE II score was 22.9 ⫾ 0.6. The APACHE II score was significantly higher in the LT group, relative to patients in the ET group. A significantly higher percentage of patients in the LT group were admitted from the cardio/ thoracic service, whereas a smaller percentage from the same group came from the trauma service, relative to the ET group. In addition, a higher proportion of the ET patients had COPD, relative to the LT group. The outcome data are listed on Table 2. Individuals in the ET group had significantly lower values for VAP, ventilator days, and hospital and ICU LOS, relative to individuals in the LT group. Odds ratios generated from the logistic regression analyses are shown in Table 3. The initial analysis was run using a cutoff of 7 days to distinguish between ET and LT. Additional analyses were run, with the only change being made to the time of the cutoff to distinguish between ET and LT. In all 5 regression analyses, the only significant predictor was time to tracheostomy. The odds ratios ranged from 2.26 using a cutoff of 7 days to an odds ratio of 3.89 using a cutoff of 3 days.
Comments Tracheostomy has an important role in the airway management of ICU patients [8]. Several studies [1,9 –12] have identified the benefits of tracheostomy over endotracheal intubation, such as sparing further injury from translaryngeal intubation, providing a stable airway, facilitating pulmonary toilet, increasing patient comfort and mobility, permitting speech and feedings, and facilitating weaning from the ventilator [5]. Despite several studies [3–7,13–18] advocating ET in the surgical critically ill patient, the timing
M.G. Möller et al. / The American Journal of Surgery 189 (2005) 293–296 Table 3 Odds ratios generated from logistic regression analyses* Cutoff point
Odds ratio (95% CI)
7 6 5 4 3
2.3 (1.2–4.4) 2.7 (1.3–5.5) 3.0 (1.4–6.5) 3.0 (1.3–7.0) 3.9 (1.2–12.2)
days days days days days
* Logistic regression was run with VAP as the dependent variable, and age, gender, admission from the trauma service, comorbidity, and ET vs. LT as the independent variables. Data from Tables 1 and 2 assumed that ET was performed within 7 days of admission to the SICU, while LT was performed later than 7 days. For the purposes of this table, 7 days is the cutoff point. Different cutoff points were used to generate the odds ratios depicted here. 95% CI ⫽ 95% confidence interval.
of tracheostomy continues to be a topic of controversy among surgeons [1,2,19,20]. Historically [2], it has been recommended to wait up to 3 weeks to convert endotracheal intubation to a tracheostomy in order to avoid an unnecessary procedure. However, it has become clear that there are significant risks associated with prolonged endotracheal intubation, such as permanent laryngeal stenosis [21]. ET has been advocated for a number of reasons. There is a reduction in dead space [1], work of breathing, and airway resistance [10]. In addition, ET enables patients to be weaned more rapidly from the ventilator [5,14], leads to decreased tracheobronchial colonization, and hastens recovery from VAP [14]. In those patients with decreased mental status, it provides airway protection [3]. It is also important to take into consideration the economic benefit of ET, which has been shown to decrease the use of resources, without an adverse effect on outcome [3]. In the current study, the ET and LT groups were fairly similar in terms of demographics and pre-admission comorbidities, except for the presence of higher number of patients with COPD in the ET group. This could be explained by the tendency to expect that those patients will do worse on mechanical ventilation, leaning the surgeon towards the decision to perform an ET. Although the difference of COPD was statistically significant, the sample was small and did not correlate with the presence of prolonged mechanical ventilation or development of VAP. In addition to the higher percentage of COPD patients, there was a significantly higher percentage of trauma patients in the early group, while there was a higher percentage of cardiovascular patients in the late group. The disparity between these groups could be a source of bias. However, logistic regression analysis did not determine trauma status to be a significant predictor for incidence of VAP. The presence of preadmission comorbidities and ARDS did not correlate to the development of VAP, and did not predict prolonged mechanical ventilation. There was a significant difference in the timing of tracheostomy between the ET group and the LT group (4.3 ⫾
295
0.2 days vs. 12.7 ⫾ 0.4 days, respectively). Due to the intrinsic limitations of a retrospective study, we were unable to identify those criteria used by the particular surgeon in order to proceed with the clinical decision of performing a tracheostomy. This decision is usually done under the clinical setting of multiple trauma, severe head injury, and the likelihood of significant disability and prolonged recovery from those injuries [2]. However, some authors have identified objective predictors of prolonged mechanical ventilation for more than 14 days, such as an alveolar–arterial oxygen gradient ⱖ175 mmHg (without COPD) and GCS less than 9 at 48 hours of admission as reported by Johnson et al [15], with a positive predictive value of 91% and a negative predictive value of 96%. Gurkin et al [17] identified 2 admitting characteristics in patients with traumatic brain injury who reman intubated at day 7 with a GCS ⱕ 8 and an injury severity score ⱖ 25. Major et al [16] recommended using GCS less than 7 and simplified acute physiologic scores (SAPS) greater than 15, on ICU day 4, in patients with blunt head injury, as their criteria for selecting those patients who should undergo tracheostomy as soon as they can tolerate the procedure. Those with GCS greater than 7 and SAPS less than 13 will likely avoid the procedure. In our study, the APACHE II score was significantly higher in the LT group, relative to patients in the ET group. However, it could be argued that although a statistically significant difference was noted between groups, this does not necessarily represent a clinically relevant difference between the 2 groups. In this study, the incidence of VAP was significantly higher in the LT group (42.3% vs. 27.2% ET group). Our results are similar to those reported by Rodriguez et al [6], who found that ET (⬍7 days) reduced ventilator days, ICU LOS, and hospital LOS. However, they only found a significant difference in VAP for a small group of their patients who had a tracheostomy within the first 48 hours, but not significant differences in the decrease of VAP for those patients who underwent tracheostomy between days 3 and 7. In our study, the odds ratio for developing VAP related to the timing of tracheostomy was significant at all days between 3 and 7. This suggests that ET may provide a benefit in decreasing VAP, even if done earlier than the 7th day, in the patient who is going to require prolonged mechanical ventilation. The most recent prospective randomized study of medically ill patients by Rumbak et al [18], comparing ET within 48 hours versus LT at days 14 to 16, demonstrated that ET has advantages over delayed tracheostomy. Their data showed decreased VAP, time on mechanical ventilation, and ICU LOS in the ET group. Under the limitations of this retrospective review of surgically ill patients, our study adds new data showing the benefit of performing ET in those patients who will need mechanical ventilation or airway protection for more than 1 week. The consideration for earlier tracheostomy should be given as early as day 3, to significantly decrease the risk of
296
M.G. Möller et al. / The American Journal of Surgery 189 (2005) 293–296
VAP. However, more randomized prospective studies are needed to obtain consistent objective data for the early identification of those surgical patients during their ICU course, in order to assist the surgeon with this important decision.
Conclusions Our results reinforce the findings of previous studies showing that ET decreases the incidence of VAP, ventilator time, ICU LOS, and hospital LOS. In addition, our study adds new evidence that earlier tracheostomy may result in greater reduction of VAP when prolonged mechanical ventilation is expected. In those patients who will require mechanical ventilation for more than 1 week, we recommend that tracheostomy be performed between day 3 and 7, in order to obtain greater benefit. More randomized prospective studies are needed to objectively identify those patients. Once the decision has been made to perform a tracheostomy, this procedure should not be delayed due to the hope that an operation could be avoided by waiting.
Acknowledgments The authors wish to express their gratitude for the assistance of Colleen Clonan, R.N. (Database Coordinator, Quality Department, Spectrum Health, Grand Rapids, MI 49503, USA) in the acquisition of the data for the manuscript.
References [1] Vallverdu I, Mancebo J. Approach to patients who fail initial weaning trials. Respir Care Clin North Am 2000;6:365–384. [2] Pryor JP, Reilly PM, Shapiro MB. Surgical airway management in the intensive care unit. Crit Care Clin 2000;16:473– 478. [3] Qureshi AI, Suarez JI, Parekh PD, et al. Prediction and timing of tracheostomy in patients with infratentorial lesions requiring mechanical ventilatory support. Crit Care Med 2000;28:1383–1387.
[4] Lanza DC, Parnes SM, Koltai PJ, et al. Early complications of airway management in head-injured patients. Laryngoscope 1990;100:958 – 961. [5] Koh WY, Lew TW, Chin NM, et al. Tracheostomy in a neurointensive care setting: indications and timing. Anesth Intensive Care 1997;25:365–368. [6] Rodriguez JL, Steinberg SM, Luchetti FA, et al. Early tracheostomy for primary airway management in the surgical critical care setting. Surgery 1990;108:655– 659. [7] Kluger Y, Paul DB, Lucke J, et al. Early tracheostomy in trauma patients. Eur J Emerg Med 1996;3:95–101. [8] Eló G, Pénzes I. Role of percutaneous tracheostomy in intensive care: a review. Orv Hetil 2002;143: 875– 879. [9] Caruso DM, Al-Kasspooles MF, Matthews MR, et al. Rational for ‘early’ percutaneous dilatational tracheostomy in patients with burn injuries. J Burn Care Rehabil 1997;18:424 – 427. [10] Diehl JL, El Atrous S, Touchard D, et al. Changes in the work of breathing induced by tracheotomy in ventilator-dependent patients. Am J Respir Crit Care Med 1999;159:383–388. [11] Wood DE: Tracheostomy. Chest Surg Clin North Am 1996;6:749 – 764. [12] D’Amelio LF, Hammond JS, Spain DA, et al. Tracheostomy and percutaneous endoscopic gastrostomy in the management of the headinjured trauma patient. Am Surgeon 1994;60:180 –185. [13] Armstrong PA, McCarthy MC, Peoples JB. Reduced use of resources by early tracheostomy in ventilator-dependent patients with blunt trauma. Surgery 1998;124:763–767. [14] Teoh WHL, Goh KYC, Chan C. The role of early tracheostomy in critically ill neurosurgical patients. Ann Acad Med Singapore 2001; 30:234 –238. [15] Johnson SB, Kearney PA, Barker DE. Early criteria predictive of prolonged mechanical ventilation. J Trauma 1992;33:95–100. [16] Major KM, Hui T, Wilson MT, et al. Objective indications for early tracheostomy after blunt head trauma. Am J Surg 2003;186:615– 619. [17] Gurkin SA, Parikshak M, Kralovich KA, et al. Indicators for tracheostomy in patients with traumatic brain injury. Am Surg 2002;68: 324 –329. [18] Rumbak MJ, Newton M, Truncale T, et al. A prospective, randomized, study comparing early percutaneous dilational tracheotomy to prolonged translaryngeal intubation (delayed tracheotomy) in critically ill medical patients. Crit Care Med 2004;32:1689 –1694. [19] Gibbons KJ. Tracheostomy: timing is everything. Crit Care Med 2000;28:1663–1664. [20] Heffner JE. Tracheotomy application and timing. Clin Chest Med 2003;24:389 –398. [21] Whited RE. A prospective study of laryngotracheal sequelae in long term intubation. Laryngoscope 1984;94:67–77.