Increasing concern regarding medical costs and pay for

Transcription

1 Original Research General Otolaryngology All-Cause Mortality after Tracheostomy at a Tertiary Care Hospital over a 10-Month Period Otolaryngology Head and Neck Surgery 146(6) Ó American Academy of Otolaryngology Head and Neck Surgery Foundation 2012 Reprints and permission: sagepub.com/journalspermissions.nav DOI: / Alexandra E. Kejner, MD 1, Paul F. Castellanos, MD, FCCP 1, Eben L. Rosenthal, MD 1, and Mary T. Hawn, MD 1 No sponsorships or competing interests have been disclosed for this article. Abstract Objective. To evaluate perioperative mortality after tracheostomy in intensive care unit (ICU) patients undergoing routine tracheostomy over a 10-month period. Study Design. Case series with planned data collection. Setting. Tertiary care hospital. Subjects. Mechanically ventilated patients. Methods. Prospective analysis of ICU patients undergoing tracheostomy placement over 10 months was performed. Variables evaluated were demographics, pretracheostomy length of stay, time on ventilator, time to death, preoperative comorbidities, and cause of death. Results. There were 129 consultations resulting in 115 tracheostomies, of which 100 were included for study. The overall 30-day postoperative mortality rate was 25%, including palliative care deaths. Cause of death in all cases was due to a preexisting condition and not from tracheostomy. Patients who died within the 30-day postoperative period were found to have significant differences in age, pretracheostomy length of stay, location of tracheostomy, and preoperative comorbidity scores. No significant difference was found in time on ventilator, sex, or race/ethnicity. Mean time from consultation to tracheostomy was 2.5 days (range, 0-12 days). Conclusion. High rates of mortality after tracheostomy can possibly affect hospital quality ratings for surgical services. There were no deaths directly related to surgery. Despite this, the mortality rate in this population was quite high. This illustrates the significant disease burden in these patients and the need to stratify postoperative mortality as well as to consider comorbidity and age when evaluating patients for tracheostomy. Keywords tracheostomy, postoperative mortality, intensive care unit patients, bedside airway care, percutaneous dilatational tracheostomy, inpatient consult care Received September 6, 2011; revised December 12, 2011; accepted January 10, Increasing concern regarding medical costs and pay for performance has led to the use of mortality rates as a measurable outcome for hospitals providing surgical care. Because of this interest, many surgical subspecialties are evaluating procedures with high mortality rates. 1 At most tertiary care institutions, one of the procedures identified in this category is tracheostomy. About 24% of intensive care unit (ICU) patients undergo tracheostomy during their hospitalization. The most commonly cited indication for tracheostomy is prolonged mechanical ventilation, followed by neurologic devastation. 2 In patients undergoing tracheostomy, 30-day mortality rate has been observed to be as high as 18%. 3 Our institution provides a tracheostomy consult service to the ICUs in the hospital, including the medical intensive care unit, the surgical intensive care unit, the cardiac care unit, the cardiovascular intensive care unit, the neurological intensive care unit, and the heart and lung transplant intensive care unit. These procedures are performed both in the operating room and at the bedside. Because of the high rate of in-hospital posttracheostomy mortality in critically ill patients, we endeavored to investigate the patient characteristics and consultation practices involved in this patient population that may contribute to this phenomenon and to potentially identify scenarios in which tracheostomy should be carefully considered if the operation could be found to contribute to mortality. Methods Approval for this study was obtained from the Institutional Review Board for Human Use through the University of Alabama Birmingham (UAB). From September 2010 to June of 2011, the Otolaryngology Consultation Service at UAB collected data on all new, elective, and nonemergent tracheostomy consults for inclusion in this study. All tracheostomy consults within this time frame were logged with 1 University of Alabama Birmingham, Birmingham, Alabama, USA This article was presented at the 2011 AAO-HNSF Annual Meeting & OTO EXPO; September 11-14, 2011; San Francisco, California. Corresponding Author: Paul F. Castellanos, MD, FCCP, University of Alabama Birmingham, BDB rd Ave S, Birmingham, AL 35294, USA paul.castellanos@ccc.uab.edu

2 Kejner et al 919 the date of consultation and date of tracheostomy. These patients were then followed to the date of discharge, transfer to palliative care, or death. Any emergency tracheostomies were excluded from the study, as were any patients who did not have valid follow-up for at least 30 days. Patient demographics were documented including age, sex, ethnicity, pretracheostomy length of stay, time on ventilator, and Charlson Comorbidity Index (CCI). The CCI was calculated using data collected from the patient s admission and discharge summaries. Delays to tracheostomy either because of patient morbidity or staff availability were documented. For patients who did not survive the postoperative period (defined as within 30 days of tracheostomy), the final cause of death was documented. From these data, the percentage mortality, time to tracheostomy, number of canceled tracheostomies, and extubations prior to tracheostomy were determined. Patients who survived the 30-day postoperative period were then compared with patients who died within 30 days. A t test was performed to evaluate significant differences between these 2 groups for age, comorbidity index, pretracheostomy length of stay, and time on ventilator. A Fisher exact test was used to evaluate differences between the 2 groups with regard to sex, race/ethnicity, and location of tracheostomy (bedside versus operating room). A Kaplan- Meier survival curve was also calculated for 30-day survival. The statistical software system GraphPad was used for statistical analyses. Results A total of 129 consultations for tracheostomy were received over a 10-month period, resulting in 115 tracheostomies (see Figure 1). Reasons for not performing tracheostomy in the remaining 14 patients included extubation or death prior to surgery (n = 6 and 7, respectively) or tracheostomy performance by a different service (n = 1). Of the 115 patients who did undergo tracheostomy, 15 patients were excluded because they did not have 30-day follow-up data available. For the 100 patients included in the study, 63 were men and 37 were women, with a median age of 57 years (range, years). Almost 60% of the patients in the study were white. The mean CCI for the patients in this study was 4.5. From admission to tracheostomy, the mean time was 16.9 days, while the average time on ventilator was 12.9 days. The average time from initial consult to tracheostomy was 2.3 days (range, 0-12 days). Most (74%) tracheostomies were performed within 24 to 48 hours of consultation. Surgery performed after 96 hours was considered delayed. The most common reason for delay of procedure was patient instability (and cancellation by anesthesia), followed by unavailability of attending surgeon to perform a bedside procedure. Four patients whose surgery was delayed ultimately died before tracheostomy could be rescheduled. Within the 30-day postoperative period, 25 (25%) patients died. Mortality associated with withdrawal of care accounted for 11 of these (44% of perioperative deaths). The 25 patients Figure 1. Flow diagram for outcomes following consultation for elective tracheostomy. who died within the 30-day postoperative window died from diseases present at the time of tracheostomy. There were no deaths attributable directly to the tracheostomy such as surgical hemorrhage, wound infection or dehiscence, cannula displacement, mediastinitis, mucus plug, or pulmonary thromboembolism. Forty-eight hour mortality was 2% (1 death on the same day of surgery and 1 death on day 2). Neither died of tracheostomy-related problems, general anesthetic, or the stress of surgery. The patient who died on the same day of surgery was a 34-year-old man with end-stage cystic fibrosis who had been transferred for a lung transplantation evaluation. The patient had active pneumonia but had been medically cleared by the primary team for bedside tracheostomy. However, he rapidly developed refractory septic shock as a result of his pneumonia. The other patient who died within 48 hours was a 57-year-old woman who developed a massive abdominal hemorrhage after cystocele repair requiring transfer to UAB for definitive treatment. She developed multiorgan system failure requiring dialysis and prolonged mechanical ventilation. She was medically cleared for tracheostomy by the ICU team. On postoperative day 2, the patient died and an autopsy was performed. It was found that she had massive, purulent ascites. There was no evidence of mucus plug, tracheal hemorrhage, or pulmonary embolism. Table 1 shows the demographics of the patients who died within the 30-day period and those who survived. A statistically significant difference was noted with regard to age, pretracheostomy length of stay, CCI, and location of tracheostomy. Factors that were not found to be statistically significant were sex, ethnicity, time on ventilator, or time to tracheostomy. Patients who underwent bedside percutaneous dilatational tracheostomy tended to have a slightly higher

3 920 Otolaryngology Head and Neck Surgery 146(6) Figure 2. Kaplan Meier survival plot over 35 days for patients who had undergone tracheostomy. CCI, although this was not found to be significant (P =.29). Figure 2 shows the overall 30-day survival curve for patients undergoing tracheostomy. With the exception of minor peristomal bleeding, which was controlled at the bedside in 2 patients, there were no significant complications directly related to tracheostomy. Sixty-six percent of the patients survived to discharge; however, 24 of these patients (36%) ultimately died. Only 3 of the patients who died after discharge died within the 30-day postoperative period. Discussion In this study, it was found that the 30-day mortality rate of ICU tracheostomies was high at 25%. Compared with data from a study that evaluated survival in a similar data set of tracheostomy patients, the mortality rate between our data set and their in-hospital data (average 42-day stay) shows no significant difference (mortality rate 25% versus 29%, P =.64, Fisher exact test). 4 Given that no death in this study was directly due to tracheostomy, the method for determining the definition of true perioperative mortality comes into question, which has long been an issue of discussion. 5 In addition, these data are salient, as hospitals are being scrutinized based on their postoperative outcomes. 6 The national trend to profile hospitals on surgical quality has increased interest in the use of risk-adjusted outcome measures including the American College of Surgeons National Surgical Quality Improvement Program (ACS- NSQIP). The NSQIP collects data on 136 variables, including preoperative risk factors, intraoperative variables, and 30-day postoperative mortality and morbidity outcomes for patients undergoing major surgery. 7 In addition, risk adjustment may not occur in patients undergoing what is considered a minor surgery. Often not taken into account is the number of procedures performed at a given institution or the scale of illness in a population for a given hospital if it has a large referral base. 8 Multiple studies have shown that risk-adjusted outcomes significantly reduce the overall purported mortality rate, suggesting that postoperative deaths/complications correlate to patient factors. 7,8 Considering that tracheostomy is typically performed in significantly ill patients, a high mortality rate would be expected. It would therefore follow that the higher the disease acuity of a particular medical center, the higher the expected mortality that would coincidentally be attributable to tracheostomy. The CCI has been shown to be a useful tool in predicting mortality in ICU patients and is comparable to the APACHE score, particularly for 30-day mortality. However, it has the potential to exclude significant patient illness factors. 9 Patient characteristics were found to be significantly different between groups that survived the postoperative period and those that did not. Patient age, pretracheostomy length of stay, comorbidity score, and location were significantly different. Age greater than 60 years did not confer any significant difference (P =.25). Length of stay could relate to mortality in multiple ways, including exposure to nosocomial infections, deconditioning leading to poor pulmonary status, and possible pulmonary thromboemboli. Comorbidity scores relate to the overall illness of the patient, which was significantly higher in patients who died more quickly. With regard to the location of tracheostomy, many patients undergoing percutaneous dilatational tracheostomy were often considered poor candidates for transport to the operating room (morbid obesity, hemodynamic instability, tenuous pulmonary status). The CCI was not found to be significantly different between patients undergoing traditional tracheostomy and those undergoing the bedside procedure (P =.6) but could potentially have been underestimated because of the stringent nature of the scoring guidelines. 9 With these data in hand, the consultant might better be able to predict which patients might benefit more from tracheostomy and which patients warrant a family discussion regarding surgical intervention. In addition, a consultation guideline could eventually be formulated for ICU physicians when considering tracheostomy in their mechanically ventilated patients. In our study, a patient who has a CCI of 5.7 or higher or who has been in the hospital for 20 days or more has a significantly increased risk of mortality. This increase in mortality could be secondary to any of these features. A multivariate analysis would be necessary to determine whether these are mutually exclusive factors. As discussed in several articles, tracheostomy in select patients neither reduces nor increases mortality in ICU patients (although in some cases, it may actually increase overall postdischarge mortality). The short-term benefits of decreased trauma to laryngeal structures, decreased need for sedation, and faster wean time must be contrasted with the potential postdischarge mortality. 10 These issues were not directly addressed in this study; however, 36% of the patients who were discharged ultimately died. This study has several limitations. As in most institutions, the referral for tracheostomy is not standardized in our hospital, and we did not collect information on similar patients who did not undergo tracheostomy. Therefore, we

4 Kejner et al 921 Table 1. Effects of Demographic and Clinical Factors on 30-Day Mortality following Tracheostomy Variable Overall N = 100 Survival to 30 Days Death at or before 30 Days P Value Mean age, y Age.60 y Sex (F:M) Female Male Race White African American Other Time from admission to tracheostomy Mean time on ventilator, d Mean CCI Tracheostomy location.002 Operating room Bedside Delayed surgery,.4 d Discharge status Alive 66 a 63 3 b Non palliative care death Palliative care a Twenty-one patients died (nonpalliative) after discharge. b Three patients died (nonpalliative) after discharge. cannot make any conclusions regarding whether tracheostomy is detrimental in some way other than the direct effect of the surgery itself that was not evaluated in this data set. Likewise, we are unable to determine if tracheostomy could be beneficial and cost-saving without further analysis. Our data neither support nor refute the benefits of earlytracheostomyasreportedbyotherssincewedidnot find any statistical impact on perioperative mortality. 11 Future studies should include a control population with additional patient/caregiver outcomes to provide a broader picture of the role of tracheostomy in the treatment of medically ill patients. In any case, the decision to proceed with tracheostomy must be formulated on a case-by-case basis. For some patients, palliative tracheostomy may afford the patient and the patient s family the opportunity to use less sedation and the ability to interact with family members. Increased comfort might also be a benefit from this procedure, although overall patient comfort as a parameter is difficult to quantify. Evidence-based consultation guidelines could help guide the decision to pursue tracheostomy as well as provide caregivers with helpful information for families of critically ill patients. Nevertheless, discussion with the family, the primary team, and the palliative/supportive care team should be pursued to maximize the quality of patient care. Further study is needed to determine whether patients ultimately benefit from tracheostomy both in-hospital and following discharge. Acknowledgments We acknowledge Joshua S. Richman, assistant professor of preventative medicine, University of Alabama Birmingham, Internal Medicine, for statistical analysis; Amalee Smith, PAC, University of Alabama Birmingham, Division of Otolaryngology, for data collection; and Larissa Sweeney, MD, University of Alabama Birmingham, Division of Otolaryngology, for statistical analyses. Author Contributions Alexandra E. Kejner, data collection, interpretation, critical revision, manuscript drafting, final approval of manuscript; Paul F. Castellanos, contributed to, revised, and provided final approval of manuscript; Eben L. Rosenthal, interpretation of data, critical revision, manuscript drafting, final approval of manuscript; Mary T. Hawn, interpretation of data, critical revision, manuscript drafting, final approval of manuscript. Disclosures Competing interests: None. Sponsorships: None. Funding source: None. References 1. Johnson ML, Gordon HS, Petersen NJ, et al. Effect of definition of mortality on hospital profiles. Med Care. 2002;40: Hsu CL, Chen KY, Chang CH, Jerng JS, Yu CJ, Yang PC. Timing of tracheostomy as a determinant of weaning success

5 922 Otolaryngology Head and Neck Surgery 146(6) in critically ill patients: a retrospective study. Crit Care. 2005; 9:R46-R Bacchetta MD, Girardi LN, Southard EJ, et al. Comparison of open versus bedside percutaneous dilatational tracheostomy in the cardiothoracic surgical patient: outcomes and financial analysis. Ann Thorac Surg. 2005;9: Kojicic M, Li G, Ahmed A, et al. Long term survival in patients with tracheostomy and prolonged mechanical ventilation in Olmsted County, Minnesota. Respir Care. 2011;56(11): Mortensen N. Wide variations in surgical mortality. BMJ. 1989;298: Reed K, May R, Taylor H, Brown A. HealthGrades Hospital Quality and Clinical Excellence Study Available at: HospitalQualityReport2011.pdf. Accessed December 9, Khuri SF, Daley J, Henderson W, et al. The Department of Veterans Affairs NSQIP: the first national, validated, outcome-based, risk-adjusted, and peer-controlled program for the measurement and enhancement of surgical care. National VA Surgical Quality Improvement Program. Ann Surg. 1998; 228: Ghaferi AA, Birkmeyer JD, Dimick JB. Variation in hospital mortality associated with inpatient surgery. N Engl J Med. 2009;361: Christensen S, Johansen MB, Christiansen CF, Jensen R, Lemeshow S. Comparison of Charlson Comorbidity Index with SAPS and APACHE scores for prediction of mortality following intensive care. Clin Epidemiol. 2011;3: Clec h C, Alberti C, Vincent F, et al. Tracheostomy does not improve the outcome of patients requiring prolonged mechanical ventilation: a propensity analysis. Crit Care Med. 2007;35: Brook A, Sherman G, Malen J, et al. Early versus late tracheostomy in patients who require prolonged mechanical ventilation. Am J Crit Care. 2000;9: