The Laryngoscope C 2013 The American Laryngological, V
Rhinological and Otological Society, Inc.
Tracheostomy as Treatment for Adult Obstructive Sleep Apnea: A Systematic Review and Meta-Analysis Macario Camacho, MD; Victor Certal, MD; Scott E. Brietzke, MD, MPH; Jon-Erik C. Holty, MD, MS; Christian Guilleminault, MD; Robson Capasso, MD Objectives/Hypothesis: To systematically review outcomes for polysomnography, sleepiness, and mortality in patients who undergo tracheostomy for the treatment of adult obstructive sleep apnea (OSA). Data Sources: MEDLINE, Scopus, and the Cochrane Library were searched from inception to March 2013, followed by extensive hand searching for the identification of relevant English language studies that met predefined criteria. Review Methods: Adult studies of tracheostomies or tracheotomies as treatment for OSA with outcomes for apnea index (AI), apnea–hypopnea index (AHI), oxygen desaturation index (ODI), effect on daytime sleepiness or mortality were identified, abstracted and pooled (as appropriate). Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines were followed. Results: The systematic search identified 18 relevant studies that were primarily case series, using retrospective review. Posttracheostomy AI improved from 73.0 6 27.1 to 0.2 6 1.2/h and an AI mean difference of 283.47 (95% CI, 2106.07 to 260.87; P < 0.0001). Mean AHI decreased from 92.0 6 34.8 to 17.3 6 20.5/h and an AHI mean difference 279.82 (95% CI, 263.74 to 295.90; P < 0.0001) compared with preoperative status was observed. Postsurgically, there was a development of central apneas; however, the central AI demonstrated near normalization to a mean of 2.1 6 3.5/h after 14 weeks. ODI decreased from 78.2 6 25.8/h to 20.8 6 25.5/h. Four studies demonstrated a statistically significant improvement in subjective sleepiness posttracheostomy. Postoperative statistically significant reductions in overall and in cardiovascular mortality compared with untreated historical cohorts was reported. Conclusion: Tracheostomies significantly decrease apnea index, oxygen desaturation index, sleepiness, and mortality in OSA subjects. Key Words: Tracheostomy, obstructive sleep apnea. Level of Evidence: NA Laryngoscope, 124:803–811, 2014
INTRODUCTION In 1965, Valero et al. described the resolution of hypersomnia by way of tracheostomy in a thin patient with significant traumatic micrognathia who was observed to have apneas during a hospital admission.1 Burwell et al. described a patient with Pickwickian syn-
From the Department of Otorhinolaryngology (V.C.), Hospital Sao Sebastiao, Sta Maria da Feira; CINTESIS–Center for Research in Health Technologies and Information Systems, University of Porto Porto, Portugal; and Department of Otolaryngology–Head and Neck Surgery (S.E.B.), Walter Reed National Military Medical Center, Bethesda, Maryland; and VA Palo Health Care System, Medical Service (J-E.C.H.), Pulmonary, Critical Care and Sleep Medicine Section, Palo Alto; and Sleep Medicine Division (C.G., M.C.), Stanford University Hospital and Clinics Redwood City; and the Robson Capasso, MD; and Department of Otolaryngology– Head and Neck Surgery, Sleep Surgery Division (R.C.), Stanford University Medical Center, Stanford, California, U.S.A. Editor’s Note: This Manuscript was accepted for publication September 10, 2013. Send correspondence to Macario Camacho, MD, Sleep Medicine Division, 2nd floor, Stanford Hospital and Clinics, 450 Broadway St., Redwood City, CA 94063. E-mail: [email protected]
Laryngoscope 124: March 2013
drome, which included morbid obesity (body mass index [BMI] of 43.76 kg/m2), hypersomnia, cyanosis, periodic respiration, and heart failure.2 In 1969, Kuhlo et al. were the first to describe treating an obstructive sleep apnea (OSA) patient (called Pickwickian syndrome at the time) successfully with a tracheostomy.3 Lugaresi demonstrated increased pulmonary and systemic arterial pressures during apneic events.4 Later, Lugaresi et al. and Coccagna et al. described treating patients with hypersomnia and periodic breathing with tracheostomies who subsequently demonstrated complete resolution of hypersomnia, normalization of systemic and pulmonary pressures, and normalization of blood gas values.4,5 Other studies have also demonstrated that tracheostomies successfully decrease systemic hypertension.6–10 In addition, there is a case study that demonstrated a significant decrease in insulin requirement 9 days after a tracheostomy was performed.10 Tilkian et al. and other studies have demonstrated that tracheostomies improved cardiac arrhythmia control and in some cases even allowed for reversal.8,11,12 Since these early studies, there have been several other case reports and case series that have Certal et al.: Tracheostomy for Obstructive Sleep Apnea
demonstrated the effectiveness of tracheostomies as a treatment for OSA.6,7,9,11,13–21 Between the late 1960s to the early 1980s, tracheostomies were the primary surgical modality utilized to treat OSA subjects failing previous medical management.3,6,8,19,20 In the 1980s, continuous positive airway pressure (CPAP) was introduced by Sullivan et al. and additional surgical modalities were developed such as uvulopalatopharyngoplasty, hypopharyngeal surgeries, and maxillomandibular advancement.22–29 Studies have reported that these newer surgical modalities (other than maxillomandibular advancement) are less effective in treating morbidly obese patients (i.e., BMI >40 kg/m2).30,31 Obesity and sedentary lifestyles are becoming a national and worldwide pandemic.32,33 Given the increase in obesity, traditional surgical modalities are less effective as surgical treatment options.30 Tracheostomies are used to treat OSA in obese or nonobese patients who have failed medical management or who are not candidates for other surgical techniques.26 Tracheostomy is perceived as a highly effective surgical option in treating OSA and associated daytime sleepiness.6–9,14,16,18–21,34 Thus, we performed a systematic review and meta-analysis of tracheostomy studies to critically evaluate the effectiveness in the treatment for OSA.
MATERIALS AND METHODS Search Strategy and Selection Criteria A MEDLINE, Scopus, and Cochrane search was performed (from inception to October 2012 with an update March, 2013) with the words “tracheotomy” or “tracheostomy and sleep,” along with the following search keywords and MeSH terms: “obstructive sleep apnea,” “Pickwickian,” “obese,” “obesity,” “apnea index,” “apnea hypopnea index,” “oxygen desaturation index,” “mortality,” “vascular,” “sleepiness,” “hypersomnolence,” “hypersomnia,” and “Epworth Sleepiness Scale” (ESS). References of all relevant articles, review articles, and relevant nonelectronic literature were hand searched to identify additional relevant studies. The authors of studies not reporting sufficient data were contacted to request additional information.
Study Selection Authors included adult studies of tracheostomies or tracheotomies for the treatment of OSA, with the outcomes for apnea index (AI), apnea–hypopnea index (AHI), oxygen desaturation index (ODI), mortality benefit, or effect on daytime sleepiness. Studies included are those written in English that provided quantitative outcomes pre- and posttracheostomy. All studies that did not include the selected polysomnogram (PSG) data were excluded.
Data Abstraction and Study Quality Assessment Two reviewers (M.C. and V.C.) independently performed a literature search, screened titles and abstracts, and subsequently extracted data from the included studies. The basic data included the following: year of publication, author(s), study site, length of follow-up, sample size, patient age, BMI, AI, AHI, ODI, sleepiness outcomes, and mortality benefit. Case series studies may be affected by various types of biases related to selection, detection, performance, attrition, reporting, and publication. For these reasons, we employed a quality appraisal tool from National Institute for Health and Clinical Excellence (NICE) to assess methodological quality of
Laryngoscope 124: March 2013
included case series studies. NICE includes eight items per case series study in its quality assessment.
Statistical Analysis Means and 95% confidence intervals (CI) for change in posttracheotomy outcomes compared with preoperative status were calculated from mean differences (after–before) and standard deviations (SD). A P value < 0.05 was considered statistically significant. When pooling study level data, we excluded studies with less than two subjects in the calculations or with incomplete or missing pre- or posttracheotomy data. If heterogeneity existed, data was analyzed using a random effects model. In the absence of heterogeneity, a fixed effects model was used. We graphically inspected Forest plots, and calculated the Q statistic (P 0.01) and I2 statistic to assess heterogeneity. A sensitivity analysis was performed by excluding each of the studies from the overall analysis one at a time to measure whether the relevant estimates were strongly influenced by any single study. The data processing and statistical analyses were performed using the Cochrane Collaboration’s Review Manager software version 4.2, and the Statistical Package for Social Sciences software (SPSS version 20.0).
RESULTS We identified 493 titles of potentially relevant articles from our computerized and manual search strategies (Fig. 1). We excluded 448 articles based on our preliminary title and abstract screening. The remaining 45 articles were retrieved for full-text evaluation, and an additional 29 articles were excluded for the following reasons: 13 failed to provide complete pre- or postoperative PSG data, five evaluated other surgical modalities (i.e., not tracheotomy), and 11 were not sleep apnearelated studies. Two studies identified from our hand search were added to the review. Overall, 18 studies were included describing 286 patients with a mean age of 49.4 6 10.9 years and a mean BMI of 35.5 6 8.5 kg/m2.
Methodological Quality of Included Studies Most included studies were retrospective case series and four were case reports (Table I). There was no randomized trial identified in the literature. The level of evidence consisted of level 5 (case studies) and level 4 (case series). Most included studies satisfied only three of the eight items on the NICE quality-assessment tool for case series studies. The primary methodological limitation of the included studies was related to the definition of inclusion and exclusion criteria, lack of prospective design, and stratification of outcomes. No study had a multicenter design and none explicitly stated that subjects were recruited consecutively. Methodological quality appraisal was not performed for the case report studies.
Polysomnography Results–Apnea Index, Apnea– Hypopnea Index, and Oxygen Desaturation Index Obstructive apnea index. Twelve studies (120 patients) demonstrated posttracheostomy AI improvement from 73.0 6 27.1 to 0.2 6 1.2/h (P