Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2020 Sep 1.
Published in final edited form as: J Craniomaxillofac Surg. 2018 May 24;46(8):1303–1312. doi: 10.1016/j.jcms.2018.05.017

Complications associated with surgical treatment of sleep-disordered breathing among hospitalized U.S. adults

Hind A Beydoun 1,*, May A Beydoun 1, Hong Cheng 1, Anjum Khan 1, Shaker M Eid 1, Carolina Alvarez-Garriga 1, Colin Anderson-Smits 1, Alan B Zonderman 1, Danica Marinac-Dabic 1
PMCID: PMC7462435  NIHMSID: NIHMS1620776  PMID: 29803366

Abstract

The purpose of this cross-sectional study is to examine the relationship between surgical treatments for sleep-disordered breathing (SDB) and composite measure of surgical complications in a nationally representative sample of hospital discharges among U.S. adults. We performed secondary analyses of 33,679 hospital discharges from the 2002–2012 Nationwide Inpatient Sample that corresponded to U.S. adults (≥18 years) who were free of head-and-neck neoplasms, were diagnosed with SDB and had undergone at least one of seven procedures. Multivariate logistic regression models were constructed to estimate adjusted odds ratios (aOR) and 95% confidence intervals (CI), controlling for age, sex, race/ethnicity, obstructive sleep apnea (OSA) and obesity diagnoses. Positive associations were found between composite measure of surgical complications and specific procedures: palatal procedure (aOR = 12.69, 95% CI: 11.91,13.53), nasal surgery (aOR = 6.47, 95% CI: 5.99,6.99), transoral robotic assist (aOR = 5.06, 95% CI: 4.34–5.88), tongue base/hypopharynx (aOR = 4.24, 95% CI: 3.88,4.62), maxillomandibular advancement (MMA) (aOR = 3.24, 95% CI: 2.74,3.84), supraglottoplasty (aOR = 2.75, 95% CI: 1.81,4.19). By contrast, a negative association was found between composite measures of surgical complications and tracheostomy (aOR = 0.033, 95% CI: 0.031,0.035). In conclusion, most procedures for SDB, except tracheostomy, were positively associated with complications, whereby palatal procedures exhibited the strongest and supraglottoplasty exhibited the weakest association.

Keywords: Complication, Morbidity, Obstructive sleep apnea, Surgery

1. Introduction

Sleep disorders are a group of chronic conditions that manifest themselves in difficulty initiating or maintaining sleep and in nonrestorative sleep (Beydoun et al., 2016; Gamaldo et al., 2016). Obstructive sleep apnea (OSA) is a serious and potentially life-threatening sleep disorder. Falling along the spectrum of sleep disordered breathing (SDB) (Choi et al., 2015), it is characterized by frequent episodes of upper airway (soft palate and/or tongue base) collapse, through intermittent closure of the pharyngeal (nasal, oral, hypo) airway during sleep, which can lead to hypoxemia, hypercapnia, arousal, sleep fragmentation and sleep disruption (Gillis et al., 2016; Strohl et al., 2016). Whereas simple snoring results from minimal airway obstruction, OSA results from complete or partial airway obstruction (Goessler et al., 2007).

The obesity epidemic and population aging have rendered SDB, including OSA, a growing public health concern (Eastwood et al., 2011). OSA is an under-appreciated and under-diagnosed, yet highly prevalent sleep disorder, affecting 2–4% of the general population and 37–40% of heavy snorers (Hamans et al., 2010; Pavelec et al., 2016; Strohl et al., 2016). The U.S./European prevalence of OSA is 2–11%, with estimated healthcare costs of $34 billion (Pavelec et al., 2016). It is estimated that 7–18 million people in the United States are affected by OSA, with the highest prevalence rate among men between the ages of 40 and 65 years (Goessler et al., 2007). If left untreated, OSA could lead to excessive daytime sleepiness, fatigue, headache, poor motor/visual skills, motor vehicle accidents, occupational accidents, lost productivity, neurocognitive impairment, hypertension, stroke, ischemic heart disease, coronary heart disease, arrhythmias, congestive heart failure, metabolic syndrome, type 2 diabetes, depression, impotence, decreased quality of life, and death (Gillis et al., 2016; Heidsieck et al., 2016).

Although numerous treatment modalities have been introduced for moderate-to-severe OSA, these are often tailored to the needs of the individual patient (Goessler et al., 2007). The “standard of care” or “first-line” treatment is continuous positive airway pressure (CPAP), which provides a “pneumatic splint” that holds the airway open (Gillis et al., 2016; Pavelec et al., 2016). Safety and effectiveness of CPAP treatment are well-established through its ability to reduce daytime sleepiness and cardiovascular complications (Choi et al., 2013). However, tolerance and long-term adherence to CPAP as prescribed (at least 4 h/night for 5 nights/week) is sub-optimal, with an estimated non-adherence rate of 30–50% (Choi et al., 2015; Friedman et al., 2016; Heidsieck et al., 2016; Murphey et al., 2016; Pavelec et al., 2016).

Alternative treatment options for patients who cannot tolerate CPAP include non-surgical treatments such as behavioral modifications (weight loss, avoidance of alcohol and sedative medicine, body position training for sleep) and dental/oral appliances, as well as single-level or multiple-level surgical treatments with or without implantable devices (Hamans et al., 2010). Surgical treatments are designed to increase the dimensions of the upper airway by addressing redundant tissue structures (soft palate, uvula, tongue, tonsils) that may be causing airway obstruction, stiffening the pharyngeal wall and/or increasing muscle tone, without compromising normal functions such as breathing, speaking or swallowing (Goessler et al., 2007; Pavelec et al., 2016). Laser-assisted uvulopalatoplasty (LAU), uvulopalatopharyngoplasty (UPPP), maxillomandibular advancement (MMA), tracheostomy, and the Pillar Implant and Upper Airway Stimulation are examples of surgical procedures that are utilized to treat patients diagnosed with OSA (Choi et al., 2013; Heidsieck et al., 2016; Murphey et al., 2016; Pavelec et al., 2016).

Distinct types of surgical procedures (nasal surgery, palatal procedure, tongue base/hypopharynx, maxillomandibular advancement, tracheostomy, supraglottoplasty, transoral robotic assist) are often applied simultaneously to individual patients depending on their specific healthcare needs (Pathak et al., 2015). Unlike CPAP, daily use of equipment is not needed, and non-adherence is therefore not an issue for most of these surgical procedures (Pavelec et al., 2016). However, since sham surgery is often not acceptable as a control group, and since clinical endpoints such as the apnea-hypopnea index (AHI) score may not be the only relevant measure, the long-term success of these surgical procedures remains equivocal (Goessler et al., 2007; Choi et al., 2013). Surgical procedures are often painful, invasive and irreversible and can carry significant morbidity risks, including complications such as infections, hemorrhage, edema, velopalatal insufficiency, thickened secretion, foreign body sensation, speech and swallowing dysfunction (Gillespie et al., 2011; Gillis et al., 2016; Pavelec et al., 2016).

To date, specific complications accompanying surgical procedures performed in adult patients diagnosed with SDB have been identified in the context of small clinical studies that focus on few types of surgical procedures at a time. At the national level, the safety profile of surgical treatments for SDB including OSA has been evaluated using a limited number of perioperative complications among adult (Capobianco et al., 2014) and pediatric (Allareddy et al., 2016) populations. Surgical treatments in the context of hospitalized adult patients diagnosed with SDB have not been characterized in terms of broadly defined complications, namely wound healing, infectious, cardiovascular, respiratory, gastrointestinal, urinary or other types of complications. The purpose of this cross-sectional study is two-fold: [1] to describe the demographic and clinical characteristics of hospitalized adult U.S. patients who had undergone surgical treatments for SDB; and [2] to analyze the relationship between surgical treatments for SDB and a composite measure of surgical complications in a nationally representative sample of hospital discharges among U.S. adults from the 2002–2012 Nationwide Inpatient Sample.

2. Materials and methods

2.1. Data source

The Agency for Healthcare Research and Quality (AHRQ) Healthcare Cost and Utilization Project (HCUP) is a federal–state–industry partnership comprising multiple databases and software tools including the Nationwide Inpatient Sample (NIS), the largest publicly available all-payer database in the United States that can provide national estimates of health care utilization, access, charges, quality and outcomes. NIS consists of a 20% stratified probability sample of community hospitals from the State Inpatient Databases which include all inpatient data currently contributed to HCUP. In 2012, NIS became a sample of discharge records from all HCUP-participating hospitals, rather than a sample of hospitals of which all discharges were kept. The sampling frame of hospitals (or hospital discharges) was divided into strata using five hospital characteristics, namely, ownership/control, bed size, teaching status, urban/rural location and U.S. region. NIS samples were selected with probabilities proportionate to the number of hospitals (or hospital discharges) in each stratum. Between 2002 and 2012, NIS data were collected from ~1000 hospitals annually and the number of states covered by NIS has grown from 35 to 44. The NIS contains clinical, non-clinical and resource use data elements typical of a discharge abstract, including hospital-level and patient-level characteristics. The original research project was approved by an institutional review board with appropriate informed consent procedures in accordance with principles outlined by the Declaration of Helsinki.

2.2. Study population and sample

We performed secondary analyses using a sub-sample from the 2002–2012 NIS data. Eligibility criteria were applied to the 2002–2012 NIS data to define the study population. Hospital discharges were included in the study if they corresponded to adult patients (≥18 years of age) who were diagnosed with conditions broadly referred to as SDB (ICD-9-CM diagnostic codes: 327.23, 327.20, 327.24, 327.29, 780.50, 780.51, 780.52, 780.53, 780.55, 780.56, 780.57, 786.09) using 15 (2002–2008) to 25 (2009–2012) diagnostic (DX) variables in NIS (Griffin et al., 2013; Mokhlesi et al., 2013a, b; Jean et al., 2015; Pathak et al., 2015; Allareddy et al., 2016; Becerra et al., 2016; Gamaldo et al., 2016; Smith et al., 2017) and had undergone at least one of the following types of surgical procedures: nasal surgery (septoplasty (ICD-9-CM procedure codes: 21.5, 21.88); rhinoplasty (ICD-9-CM procedure codes: 21.84, 21.85, 21.86, 21.87)), palatal procedure (palatal surgery (ICD-9-CM procedure codes: 27.64, 27.69, 27.72, 27.73, 27.79, 29.39, 29.4); tonsillectomy (ICD-9-CM procedure codes: 28.2, 28.3); adenoidectomy (ICD-9-CM procedure codes: 28.6); palatal implant/Pillar (ICD-9-CM procedure codes: 27.64)), tongue base/hypopharynx (tongue radiofrequency/midline glossectomy (ICD-9-CM procedure codes: 25.1, 25.2, 25.59, 25.94, 25.99); genioglossus advancement, genioplasty or tongue stabilization (ICD-9-CM procedure codes: 76.63, 76.64, 76.68); lingual tonsillectomy (ICD-9-CM procedure codes: 28.5); hyoid suspension (ICD-9-CM procedure codes: 83.02), maxillomandibular advancement (ICD-9-CM procedure codes: 76.43, 76.46, 76.61, 76.62, 76.65, 76.66), tracheostomy (ICD-9-CM procedure codes: 31.1, 31.2, 31.29), supraglottoplasty (ICD-9-CM procedure codes: 31.69), transoral robotic assist (ICD-9-CM procedure codes: 17.41,17.44, 17.49) using 15 procedure (PR) variables in NIS (Ishman et al., 2014; Pathak et al., 2015). Hospital discharges were excluded from the study if they corresponded to patients diagnosed with head-and-neck neoplasms (ICD-9-CM diagnostic codes: 140–149, 160–169) (Capobianco et al., 2014). ICD-9-CM codes were selected on the basis of previously published studies using AHRQ data. The resulting study sample consists of 33,679 out of 87,039,711 hospital discharges from the 2002–2012 NIS (Fig. 1).

Fig. 1.

Fig. 1.

Open in a new tab

Flowchart for selection of study sample.

2.3. Exposure variables

Hospital discharges were classified according to whether they corresponded to patients who had undergone a specific type of surgical procedure, accounting for sample size limitations. Accordingly, the following dichotomous (‘yes’, ‘no’) exposure variables were defined: ‘nasal surgery’, ‘palatal procedure’, ‘tongue base/hypopharynx’, ‘maxillomandibular advancement’, ‘tracheostomy’, ‘supraglottoplasty’ and ‘transoral robotic assist’.

2.4. Outcome variables

In-hospital complications and morbid conditions were defined as dichotomous (‘yes’, ‘no’) variables based on ICD-9-CM diagnostic codes previously identified in a validation study (Tuinen et al., 2005). To achieve adequate sample size, we grouped the originally identified classes of complications into seven broadly defined types: ‘wound healing’, ‘infections’, ‘cardiovascular’, ‘respiratory’, ‘gastrointestinal’, ‘urinary’ and ‘other’. Furthermore, a composite (‘yes’ or ‘no’) measure of surgical complications was defined for hospital discharges based on presence or absence of at least one of these ICD-9-CM diagnostic codes (Table 1).

Table 1.

ICD-9-CM diagnostic coding for types of complications.

Type of complication ICD-9-CM diagnostic code
Wound healing 01.23, 03.02, 06.02, 34.03, 35.95, 39.49, 54.12, 54.61, 28.7, 39.41, 39.98, 49.95, 57.93, 60.94, 29.51, 31.61, 33.41, 33.43, 42.82, 44.61, 46.71, 46.75, 48.71, 50.61, 51.91, 55.81, 56.82, 57.81, 58.41, 69.41, 53.04, 569.83, 575.4, 576.3
Infections 038.0, 038.10, 038.11, 038.19, 038.3, 790.7, 038.40–038.9, 481, 482.0, 482.1, 482.2, 482.30, 482.31, 482.32, 482.39, 482.40, 482.41, 482.49, 482.89, 482.9, 483.8, 485, 486, 008.45, 320.3, 320.82, 321.3, 421.0, 421.1, 421.9, 424.90, 424.91, 424.99, 510.0, 510.9, 513.0, 513.1, 519.01, 536.41, 569.61, 590.10, 590.11, 590.80, 590.9, 595.0, 595.9, 599.0, 670.00, 670.02, 670.04, 682.3, 682.4, 683, 958.3
Cardiovascular 410.00, 410.01, 410.10, 410.11, 410.20, 410.21, 410.30, 410.31, 410.40, 410.41, 410.50, 410.51, 410.60, 410.61, 410.70, 410.71, 410.80, 410.81, 410.90, 410.91, 423.0, 427.5, 429.4, 415.11, 415.19, 453.8, 458.2, 451.11–451.9
Respiratory 495.7, 507.0, 512.1, 514, 518.0, 518.4, 518.5, 518.81, 518.82, 519.02, 519.09, 519.1, 519.2
Gastrointestinal 530.82, 535.01, 535.11, 535.21, 535.41, 535.51, 535.61, 536.2, 536.3, 532.00–532.21, 536.40, 536.42, 536.49, 564.2, 564.3, 564.4, 569.60, 569.62, 569.69, 787.01–787.03,787.91
Urinary 584.5–584.9,598.2
Other 996.00–996.99, 998.11–998.13, 999.0–999.9, 998.2, E870.0-E870.9, 998.4, 998.7, E871.0-E876.9, E878.0-E878.9
Open in a new tab

2.5. Covariates

The hypothesized relationships between exposure and outcome variables were examined before and after adjustment for selected socio-demographic covariates: age (years) defined as a continuous variable; sex (‘male’, ‘female’); race/ethnicity (‘White’, ‘Black’, ‘Hispanic’, ‘Asian/Pacific Islander’, Other/Unknown’); OSA diagnosis (‘yes’, ‘no’) based on a specific ICD-9-CM diagnostic code of ‘327.23’ and obesity diagnosis (‘yes’, ‘no’) based on ICD-9-CM diagnostic codes of ‘278.00’ and ‘V85.3x’. Age was further categorized as (‘18–29′, ‘30–39′, ‘40–49′, ‘50–59′, ‘60–69′, ‘70+’) years for descriptive purposes.

2.6. Statistical analyses

All statistical analyses were conducted using the SAS version 9.4 software (SAS Institute, Cary, NC, USA), taking complex survey design into account. Descriptive statistics included mean (±standard error) for continuous variables and frequencies with percentages for categorical variables. Bivariate associations were examined using design-based Chi-square and F-tests, as appropriate. Univariate and multivariate logistic regression models were constructed to estimate crude and adjusted odds ratios (cOR and aOR) with 95% confidence intervals (Cl), taking age, sex, race/ethnicity, OSA and obesity diagnosis into consideration. Two-sided statistical tests were conducted at an alpha level of 0.05.

3. Results

Out of 33,679 hospital discharges, 67.2% corresponded to male patients, 56.4% to patients of white ethnicity, 55.8% to patients specifically diagnosed with OSA and 11% to obese patients. Furthermore, the mean (±standard error) age of the study sample was 50.4 (±0.08) years, with nearly half of all patients between 40 and 59 years of age. Palatal procedure (44%), including palatal surgery (39.6%) and tonsillectomy (30.8%), was the most frequently observed and supraglottoplasty (0.5%) was the least frequently observed surgical procedure type. Nearly 42% of hospital admissions corresponded to patients who had experienced any type of complication, of which respiratory (29.3%) and infectious (23.9%) complications were the most frequently observed.

As shown in Table 2, patients assigned the OSA-specific ICD-9-CM code of ‘327.23’ were significantly older, more frequently female or of black ethnicity and experienced consistently higher prevalence of surgical complications than their counterparts. Specific surgical procedures (palatal implant/pillar, tongue radiofrequency/midline glossectomy, lingual tonsillectomy, hyoid suspension, tracheostomy, transoral robotic assist) were significantly more prevalent among patients assigned the OSA-specific ICD-9-CM code of ‘327.23’, while others were significantly more prevalent among their counterparts (septoplasty, palatal surgery, tonsillectomy). In all, 3.9% of hospitalized patients in the study sample (4.6% of patients with OSA diagnosis and 2.9% of patients without OSA diagnosis, P < 0.0001) died in-hospital.

Table 2.

Distribution of study sample according to patient characteristics, surgical treatments and complications by obstructive sleep apnea (327.23) diagnosis–Nationwide Inpatient Sample, 2002–2012a.

Total (n = 33679)
Mean ± SEM or %		 ICD-9-CM: 327.23 (n = 18704)
Mean ± SEM or %		 No ICD-9-CM: 327.23 (n = 14975)
Mean ± SEM or %		 P value
Age (yr)
Mean ± SEM 50.4 ± 0.08 51.4 ± 0.1 49.3 ± 0.1 <0.0001
 18–29 8.1 7.5 8.9 <0.0001
 30–39 16.2 14.8 17.9
 40–49 24.2 23.1 25.5
 50–59 23.9 24.4 23.4
 60–69 16.9 18.6 14.8
 70+ 10.7 11.7 9.6
Sex <0.0001
 Male 67.2 65.9 68.7
 Female 32.7 33.9 31.2
 Unknown 0.2 0.1 0.2
Race/ethnicity <0.0001
 White 56.4 56.9 55.8
 Black 11.6 13.3 9.4
 Hispanic 7.7 8.2 7.1
 Asian/Pacific Islander 1.9 1.8 1.9
 Other/Unknown 22.4 19.7 25.8
Obesity Diagnosis <0.0001
Yes 11.0 12.6 8.9
No 88.9 87.4 91.0
Surgical Procedures
 Nasal surgery: 23.9 21.5 27.0 <0.0001
 Septoplasty 23.5 21.0 26.5 <0.0001
 Rhinoplasty 0.8 0.9 0.7 0.1
 Palatal procedure: 44.0 39.0 50.4 <0.0001
 Palatal surgery 39.6 34.6 46.0 <0.0001
 Tonsillectomy 30.8 27.7 34.7 <0.0001
 Adenoidectomy 0.4 0.5 0.4 0.3
 Palatal implant/Pillar 0.1 0.2 0.04 0.0007
 Tongue base/Hypopharynx: 10.9 15.4 14.3 0.01
 Tongue radiofrequency/Midline glossectomy 10.9 11.2 10.5 0.03
 Genioglossus advancement, genioplasty or tongue stabilization 3.2 3.1 3.3 0.4
 Lingual tonsillectomy 0.5 0.6 0.3 <0.0001
 Hyoid suspension 2.08 2.4 1.6 <0.0001
 Maxillomandibular advancement: 3.9 4.0 3.8 0.3
 Tracheostomy: 40.9 45.8 34.7 <0.0001
 Supraglottoplasty: 0.5 0.4 0.6 0.1
 Transoral robotic assist: 3.7 4.3 3.0 <0.0001
Surgical Complications
 Wound healing 1.0 0.8 1.2 0.003
 Infectious 23.9 29.9 16.4 <0.0001
 Cardiovascular 5.7 7.3 3.7 <0.0001
 Respiratory 29.3 34.9 22.4 <0.0001
 Gastrointestinal 5.5 7.2 3.5 <0.0001
 Urinary 11.9 17.2 5.2 <0.0001
 Other 10.9 12.3 9.4 <0.0001
 Any 41.8 48.6 33.3 <0.0001
In-hospital mortality rate 3.9 4.6 2.9 <0.0001
Open in a new tab

SEM = standard error of the mean.

a

Numbers within cells represent mean ± standard error, percentages, or P values; P values are outcomes of design-based F-tests and Chi-square tests, as appropriate.

Table 3 presents surgical treatments according to patient characteristics in the study sample. Out of 33,679 hospital discharges, 24,310 had one type, 7,976 had two types, 1,334 had three types and 59 had four types of surgical treatments. It is worth noting that certain types of surgical procedures (e.g. nasal and palatal procedures; tongue-base/hypopharynx, MMA and palatal procedures; supraglottoplasty and tongue-base/hypopharynx procedures) often clustered together, whereas a consistently negative relationship was observed between tracheostomy and other types of surgical procedures. Patients who had undergone tracheostomy, supraglottoplasty or transoral robotic assist were significantly older, whereas patients who had undergone nasal surgery, palatal procedure, tongue-base/hypopharynx or MMA procedures were significantly younger than those who did not. Similarly, sex and race/ethnicity distributions varied according to use of specific types of surgical procedures, with the exception of MMA and transoral robotic assist, which did not differ significantly according to sex and supraglottoplasty, which did not differ significantly according to sex or race/ethnicity. Further analyses suggested that specific surgical procedures were more frequently performed at larger bed-size hospitals (tongue/hypopharynx, MMA, tracheostomy), whereas others (nasal surgery, palatal procedure, transoral robotic assist) were more frequently performed at smaller bed-size hospitals.

Table 3.

Surgical treatments for sleep disordered breathing according to patient characteristics in study sample–Nationwide Inpatient Sample, 2002–2012a.

Nasal Surgery
 Palatal Procedure
 Tongue base/Hypopharynx
 Maxillomandibular Advancement
 Tracheostomy
 Supraglottoplasty
 Transoral robotic assist:
y/n Mean ± SEM or % y/n Mean ± SEM or % y/n Mean ± SEM or % y/n Mean ± SEM or % y/n Mean ± SEM or % y/n Mean ± SEM or % y/n Mean ± SEM or %
Age (yr)
Mean ± SEM 46.2 ± 0.1/51.8 ± 0.09 43.7 ± 0.1/55.6 ± 0.1 46.7 ± 0.2/51.1 ± 0.1 42.1 ± 0.4/50.8 ± 0.08 57.2 ± 0.1/45.7 ± 0.09 54.4 ± 1.0/50.4 ± 0.08 59.2 ± 0.3/50.1 ± 0.08
P value <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 0.0001 <0.0001
 18–29 2.1/6.0 5.4/2.7 1.1/6.9 0.7/7.3 1.5/6.5 0.02/8.1 0.02/8.1
 30–39 5.5/10.7 11.2/4.9 3.0/13.1 0.8/15.4 3.1/13.1 0.05/16.1 0.05/16.1
 40–49 7.4/16.8 13.8/10.4 4.8/19.4 1.2/22.9 6.8/17.3 0.1/24.0 0.1/24.0
 50–59 5.5/18.4 9.2/14.8 3.8/20.1 0.9/23.1 10.7/13.3 0.1/23.8 0.1/23.8
 60–69 2.5/14.4 3.5/13.4 1.6/15.4 0.3/16.6 10.5/6.4 0.1/16.8 0.1/16.8
 70+ 10.7/0.9 1.0/9.7 0.6/10.2 0.05/10.7 8.2/2.6 0.07/10.7 0.5/99.5
P value <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 0.007 <0.0001
Sex
 Male 18.8/48.5 33.1/34.1 10.8/56.4 2.6/64.6 24.1/43.1 0.3/66.9 2.3/64.9
 Female 5.2/27.5 10.9/21.8 4.0/28.7 1.3/31.4 16.8/15.9 0.1/32.6 1.4/31.4
P value <0.0001 <0.0001 <0.0001 0.60 <0.0001 0.32 0.32
Race
 White 13.5/42.9 23.3/33.1 8.9/47.5 2.7/53.7 23.3/33.1 0.3/56.1 2.6/53.8
 Black 1.5/10.1 4.1/7.5 0.8/10.8 0.1/11.4 6.7/4.9 0.04/11.5 0.2/11.3
 Hispanic 1.9/5.8 3.8/3.9 0.9/6.8 0.2/7.6 3.0/4.7 0.01/7.7 0.2/7.5
 Asian/Pacific Islander 0.4/1.4 1.2/0.7 0.4/1.5 0.09/1.8 0.4/1.4 0.009/1.9 0.05/1.8
 Other/Unknown 6.6/15.8 11.7/10.8 3.9/18.5 0.9/21.6 7.4/15.0 0.1/22.3 0.6/21.9
P value <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 0.06 <0.0001
Obesity diagnosis
 Yes 2.6/8.4 5.2/5.8 1.7/9.3 0.4/10.7 3.9/7.0 0.07/10.9 0.5/10.5
 No 21.3/67.7 38.8/50.1 13.2/75.8 3.6/83.4 36.9/52.1 0.4/88.6 3.2/85.8
P value 0.0003 <0.0001 0.08 0.01 <0.0001 0.3 0.0002
Open in a new tab

SEM = standard error of the mean; y/n = yes/no.

a

Numbers within cells represent mean ± standard error, percentages, or P values; P values are outcomes of design-based F-tests and Chi-square tests, as appropriate.

Table 4 presents complications and morbid conditions according to patient characteristics. Patients who experienced any of the broadly defined types of surgical complications were significantly older than those who did not. Except for ‘other’ types of surgical complications which did not exhibit sex disparities, males were significantly more likely than females to experience complications after surgical procedures. Racial and ethnic disparities in surgical complications were consistently observed except for wound healing.

Table 4.

Surgical complications according to patient characteristics–Nationwide Inpatient Sample, 2002–2012a.

Wound Healing
 Infectious
 Cardiovascular
 Respiratory
 Gastrointestinal
 Urinary
 Other
 Any
y/n Mean ± SEM or % y/n Mean ± SEM or % y/n Mean ± SEM or % y/n Mean ± SEM or % y/n Mean ± SEM or % y/n Mean ± SEM or % y/n Mean ± SEM or % y/n Mean ± SEM or %
Age (yr)
Mean ± SEM 57.2 ± 0.8/50.4 ± 0.08 57.7 ± 0.2/48.1 ± 0.09 58.9 ± 0.3/49.9 ± 0.08 56.8 ± 0.1/47.8 ± 0.09 57.3 ± 0.3/50.0 ± 0.08 59.4 ± 0.2/49.2 ± 0.08 54.4 ± 0.2/49.9 ± 0.08 56.2 ± 0.1/46.3 ± 0.09
P value <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001
 18–29 0.04/8.0 0.8/7.2 0.1/7.9 1.2/6.8 0.3/7.8 0.2/7.9 0.6/7.4 1.9/6.2
 30–39 0.09/16.0 1.8/14.4 0.4/15.8 2.5/13.7 0.4/15.8 0.7/15.4 1.3/14.9 3.8/12.3
 40–49 0.2/23.9 3.8/20.3 0.9/23.3 5.0/19.0 0.8/23.3 1.7/22.4 2.0/22.1 7.4/16.8
 50–59 0.2/23.7 6.0/17.9 1.4/22.5 7.4/16.6 1.4/22.6 3.0/20.9 2.8/21.1 10.6/13.3
 60–69 0.2/16.7 6.2/10.7 1.6/15.3 7.4/9.5 1.4/15.5 3.5/13.5 2.5/14.5 10.1/6.8
 70+ 0.2/10.5 5.2/5.6 1.3/9.4 5.8/4.9 1.2/9.5 2.7/8.0 1.8/8.9 7.9/2.8
P value <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001
Sex
 Male 0.6/66.7 13.3/53.9 3.5/63.7 17.6/49.7 3.1/64.1 7.1/60.1 7.3/59.9 24.9/42.3
 Female 0.4/32.3 10.7/22.0 2.1/30.6 11.8/20.9 2.4/30.3 4.8/27.9 3.7/29.0 16.9/15.8
P value 0.002 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 0.17 <0.0001
Race
 White 0.6/55.8 13.8/42.6 3.3/53.1 16.7/39.7 3.4/53.1 6.9/49.5 6.5/49.9 23.9/32.5
 Black 11.4/56.4 4.1/7.5 0.9/10.6 4.7/6.8 0.9/10.6 2.3/9.2 1.5/10.1 6.6/4.9
 Hispanic 0.07/7.6 2.0/5.7 0.4/7.3 2.4/5.3 0.5/7.3 0.9/6.8 0.8/6.9 3.2/4.6
 Asian/Pacific Islander 1.1/1.8 0.3/1.6 0.07/1.8 0.4/1.5 0.1/1.8 0.1/1.7 0.2/1.7 0.5/1.3
 Other/Unknown 0.2/22.3 3.7/18.7 0.9/21.5 5.2/17.3 0.7/21.7 1.6/20.8 2.0/20.4 7.6/14.8
P value 0.11 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001
Obesity
 Yes 1.1 10.9 2.2/8.8 0.6/10.5 3.1/7.9 0.6/10.4 1.0/9.9 1.2/9.8 4.3/6.7
 No 0.88/88.1 21.7/67.2 5.1/83.8 26.2/62.7 4.9/84.0 10.9/78.1 9.8/79.2 37.5/51.4
P value 0.002 <0.0001 0.08 0.2 0.36 <0.0001 0.9 0.0001
Open in a new tab

SEM = standard error of the mean; y/n = yes/no.

a

Numbers within cells represent mean ± standard error, percentages, or P values; P values are outcomes of design-based F-tests and Chi-square tests, as appropriate.

Table 5 presents the results of logistic regression models with OR and 95% CI for the association between surgical treatments and a composite measure of surgical complications, before and after adjustment for covariates. Multivariate logistic regression models suggested a significantly positive association between any complication and specific procedures, namely, palatal procedure (aOR = 12.69,95% CI: 11.91,13.53), nasal surgery(aOR = 6.47,95% CI: 5.99, 6.99), transoral robotic assist (aOR = 5.06, 95% CI: 4.34–5.88), tongue base/hypopharynx (aOR = 4.24, 95% CI: 3.88, 4.62), MMA (aOR = 3.24,95% CI: 2.74–3.84), supraglottoplasty (aOR = 2.75, 95% CI: 1.81, 4.19), and a significantly protective association between any complication and tracheostomy (aOR = 0.033, 95% CI: 0.031, 0.035).

Table 5.

Logistic regression models (unadjusted and adjusted) with odds ratios and 95% confidence intervals for surgical treatments as predictors of a composite measure of surgical complications in patients diagnosed with sleep disordered breathing according to obstructive sleep apnea status–Nationwide Inpatient Sample, 2002–2012.

Overall Composite Measure of Surgical Complications: OR (95% CI)
Unadjusted Adjusteda
Nasal Surgery 7.98 (7.42, 8.57) 6.47 (5.99, 6.99)
Palatal Procedure 16.79 (15.81, 17.84) 12.69 (11.91, 13.53)
Tongue base/Hypopharynx 4.82 (4.45, 5.23) 4.24 (3.88, 4.62)
Maxillomandibular Advancement 4.45 (3.81, 5.19) 3.24 (2.74, 3.84)
Tracheostomy 0.027 (0.025, 0.028) 0.033 (0.031, 0.035)
Supraglottoplasty 2.23 (1.57, 3.18) 2.75 (1.81, 4.19)
Transoral robotic assist 2.79 (2.43, 3.21) 5.06 (4.34, 5.88)
Open in a new tab

CI = confidence interval; OR = odds ratio.

a

Multiple logistic regression models were adjusted for age (in years), sex, race, diagnosis with obstructive sleep apnea, and obesity diagnosis.

4. Discussion

In a cross-sectional study of 33,679 hospital discharges from the 2002–2012 AHRQ NIS, we evaluated the associations between specific surgical procedures and a composite measure of surgical complications among hospitalized adult U.S. patients free of head-and-neck neoplasms who were diagnosed with SDB. Palatal procedure was the most frequent and supraglottoplasty was the least frequent surgical procedure type. The most commonly observed types of surgical complications were infectious or respiratory in nature. Multivariate logistic regression models suggested that palatal procedure, nasal surgery, transoral robotic assist, tongue base/hypopharynx surgery and supraglottoplasty were associated with higher prevalence rates of surgical complications, whereas tracheostomy was associated with a lower prevalence rate of surgical complications.

Study findings can be generalized to patients admitted to U.S. hospitals who had undergone surgical procedures to treat OSA and related sleep disorders, but not to non-hospitalized patients or the general population of the United States. The estimated OSA prevalence (56%) in our study sample was substantially greater than previously reported among U.S. and European populations (2–11%), an expected finding given that all patients were hospitalized and had undergone at least one surgical procedure associated with sleep disorders (Pavelec et al., 2016). Whereas the actual OSA prevalence would be 3.2 million out of the 72.5 million, or 4.4%, the OSA prevalence among those undergoing surgeries may be underestimated because it is limited by the number of patients who underwent polysomnography or home sleep testing. Moreover, surgical complications were classified into broad categories and were less specific than those identified in previously conducted studies. As a result, our estimated surgical complication rate (42%) was considerably higher than that in the relevant literature. Current evidence pertaining to the safety profile of distinct surgical procedures among adult patients diagnosed with OSA and related sleep disorders originates from small clinical studies with limited external validity; whereas several studies reported no major, serious or long-term complications, others reported complication rates of <15% (Li and Shi, 2013; Banuchi et al., 2014). Frequently reported complications include bleeding (Plzak et al., 2013; Banuchi et al., 2014), pneumonia (Busaba, 2002), hematoma (Busaba, 2002), ulceration (Nelson and Barrera, 2007; Plzak et al., 2013), abscess formation (Stuck et al., 2002; Fischer et al., 2003), granulation tissue (Thatcher and Maisel, 2003), infection (Thatcher and Maisel, 2003), regurgitation (Liu et al., 2012), aspiration (Neruntarat, 2003), dysphagia (Neruntarat, 2003), hypesthesia (Vigneron et al., 2017), paresthesia (Liu et al., 2012), taste disturbances (Plzak et al., 2013) and revision (Thatcher and Maisel, 2003). These specific types of complications can be further grouped under our broadly-defined wound healing, infectious, respiratory, gastrointestinal as well as other types of complications. Consistent with previous studies, our study identified infectious and respiratory types of complications as being associated with most surgical procedures. Although the death rate (3.9%) may be relatively high, it is worth noting that these patients were hospitalized to undergo surgical procedures that may be performed in the outpatient setting, suggesting that they may be less healthy at baseline than their outpatient counterparts. Accordingly, the high death rate is likely attributable to the severity of the underlying condition rather than to the surgical procedure itself.

To our knowledge, this is the first study to examine broadly-defined complications of surgical procedures for SDB, including OSA, using a nationally representative sample of hospital discharges among U.S. adults. Previously, Allareddy and colleagues examined 141,599 U.S. hospitalized children ≤21 years of age (58.1% male, 26.4% OSA) who underwent tonsillectomy with (n = 116,319) or without (n = 25,280) adenoidectomy. They found an overall complication rate of 6.4%, of which postoperative pneumonia (2.3%), bacterial infections (1.4%), respiratory complications (1.3%) and hemorrhage (1.2%) were the most commonly observed complications (Allareddy et al., 2016). Our study included adult patients with considerably higher rates of surgical complications as compared to those in children. As expected, we also found that complicated surgeries corresponded to patients that were, on average, older than those with uncomplicated surgeries.

To date, a limited number of studies have focused on the impact of SDB diagnosis on surgical outcomes, including complications, among U.S. adults. Using data on 1,058,710 hospital discharges from adult patients who had undergone elective surgery (orthopedic, prostate, abdominal, and cardiovascular), Mokhlesi and colleagues examined the relationship between SDB and various outcomes, including in-hospital death, total charges, length of stay, respiratory and cardiac complications (Mokhlesi et al., 2013b). SDB was associated with decreased mortality in the orthopedic, abdominal, and cardiovascular surgery groups but had no impact on mortality in the prostate surgery group. SDB was independently associated with a small increase in length of hospital stay and total charges in the orthopedic surgery group (Mokhlesi et al., 2013b). In the abdominal and cardiovascular surgery groups, SDB was associated with a decrease in length of hospital stay and total charges in adjusted models (Mokhlesi et al., 2013b). SDB was also associated with increased odds of emergent intubation, mechanical ventilation, noninvasive ventilation, and atrial fibrillation in all four surgical categories (Mokhlesi et al., 2013b). Using data on 91,028 U.S. hospital discharges from adult patients who had undergone bariatric surgeries, Mokhlesi and colleagues examined the relationship between SDB and various outcomes, including in-hospital death, total charges, length of stay, respiratory and cardiac complications (Mokhlesi et al., 2013a). SDB was associated with decreased mortality, total charges, and length of stay and with an increased odds ratio of emergent endotracheal intubation, non-invasive ventilation and atrial fibrillation (Mokhlesi et al., 2013a). Griffin and colleagues examined OSA as a risk factor for postoperative in-hospital complications, length of stay and costs after shoulder arthroplasty. Among 22,988 U.S. patients who underwent total shoulder arthroplasty or hemiarthroplasty, 5.9% were diagnosed with OSA. Patients with OSA had similar in-hospital mortality, complication rates (including those for pulmonary embolism) and postoperative charges compared with patients without OSA (D’Apuzzo and Browne, 2012). However, OSA patients experienced shorter hospital stay as compared to non-OSA patients (Griffin et al., 2013). Finally, D’Apuzzo and Browne evaluated whether OSA diagnosis was linked to postoperative complications and costs after revision arthroplasty using data from 258,455 U.S. patients (6.4% OSA) who underwent revision total hip or total knee arthroplasty (D’Apuzzo and Browne, 2012). Multivariable logistic regression models suggested that OSA was associated with increased in-hospital mortality, pulmonary embolism, wound hematomas or seromas and increased postoperative charges (D’Apuzzo and Browne, 2012).

We identified three studies that had previously addressed surgeries for OSA and related sleep disorders in U.S. adults, but did not examine complications associated with these surgeries. Kezirian and colleagues evaluated OSA surgical volume, types, costs, and trends within inpatient and outpatient U.S. facilities using 2000, 2004, and 2006 NIS as well as 2006 State Ambulatory Surgery Databases and State Inpatient Databases from four representative states (California, New York, North Carolina, and Wisconsin) (Kezirian et al., 2010). In 2006, an estimated 35,263 surgeries were performed in inpatient and outpatient settings, including 33,087 palate, 6,561 hypopharyngeal and 1,378 MMA procedures, and the odds of undergoing isolated palate surgery were higher for younger patients (18–39 years) and patients of black ethnicity (Kezirian et al., 2010). Outpatient procedures were more common than inpatient procedures, with inpatient surgical volume declining from 2000 to 2006. In 2006, mean costs were approximately $6000 per admission (Kezirian et al., 2010). For inpatient procedures in 2004 and 2006, costs were higher for hypopharyngeal (vs. isolated palate) surgery, in rural hospitals, and for patients who were younger, with greater medical comorbidity, and with primary Medicaid coverage (Kezirian et al., 2010). Ishman and colleagues examined time trends in patterns of surgical procedures for sleep disorders using 1993–2010 data on 232,470 U.S. patients who underwent surgical procedures for SDB or OSA (Ishman et al., 2014). Sleep surgeries increased from 97,363 in 1993–2000 to 135,107 in 2001–2010, and sleep surgery in 2001–2010 was linked to increased hypopharyngeal surgery, tongue radiofrequency/midline glossectomy, hyoid suspension, nasal surgery and multilevel surgery and decreased tracheostomy compared to 1993–2000 (Ishman et al., 2014). Capobianco and colleagues assessed the relationship between surgeon/hospital volume and outcomes (mortality rate, hospital length of stay, postoperative complication rate, hospitalization charges) among 24,298 adults undergoing OSA surgery (Capobianco et al., 2014). The authors found that patients treated by low volume (1 procedure/year) surgeons experienced higher mortality rate, longer average length of stay, and higher hospitalization charges versus medium- and high-volume surgeons (2–4 procedures/year; ≥ 5 procedures/year, respectively) and patients treated at hospitals with a low volume of procedures (0–5/year) had significantly higher occurrence of oxygen desaturation, longer length of stay and higher hospitalization charges than patients treated at high-volume hospitals (≥18 procedures/year) (Capobianco et al., 2014).

This study has numerous strengths, including national representativeness, sample size as well as patient-level data elements that are harmonized across study years. Nevertheless, our study findings should be interpreted cautiously in light of several limitations. First, the unit of analysis is a hospital discharge rather than a patient, and clustering by patient identifiers could not be corrected as a result of record de-identification. Second, it is not clear whether the selected study sample consists of hospital discharges among patients who have failed CPAP, the gold standard treatment for OSA and related sleep disorders. Similarly, we cannot assume, based on the NIS data, that surgeries were performed specifically to treat SDB. This is particularly problematic for tracheostomies, which are highly prevalent in this population and may have been performed for other reasons besides SDB. Third, although AHRQhas installed a system for quality checks using internal and external validations, the use of ICD-9-CM procedure and diagnostic codes from an administrative database could lead to exposure, outcome or covariate misclassification. In addition, ICD-9-CM coding does not allow us to distinguish moderate-to-severe OSA, or to classify complications according to their severity, device- or procedure-relatedness. Fourth, the cross-sectional design does not allow us to establish temporal and causal relationships or to examine long-term outcomes beyond the period of hospitalization. In particular, the sequence of surgeries performed on individual patients cannot be ascertained from the NIS data and is an important predictor of surgical complication rates. Moreover, the outcomes identified as surgical complications may be pre-existing or comorbid conditions. In summary, surgeries may not be performed to treat SDB but rather other primary diagnoses (e.g. respiratory failure, vent dependence), and observed complications may not be the effects of surgery or new diagnoses. Fifth, the study compares hospital discharges according to exposure to seven alternative surgical treatments, and nearly one quarter of all hospital discharges consists of patients who had undergone multiple types of surgeries, making them a heterogeneous comparison group. Sixth, data on comorbid conditions as well as physical examination, laboratory tests, surgical and medication use details are lacking in the NIS, limiting our ability to adjust for relevant confounders. It is worth noting that the degree of OSA severity may affect morbidity and mortality rates, and ideally should have been controlled for at the analysis stage. Seventh, NIS is restricted to inpatient discharges, a special population that may differ in many ways from outpatient or non-institutionalized populations. Several types of surgical procedures described in this study may be undertaken in an outpatient setting, and patients treated with these surgical procedures in an inpatient setting may be more severely affected by SDB or its surgical treatment. Eighth, surgical treatment patterns for sleep disorders such as OSA have evolved over time. The observed patterns of surgical treatments in this study, which may include LAU and exclude neurostimulators, are typical of the time period ranging between 2002 and 2012, but may not reflect current practice guidelines. Finally, missing data can lead to selection bias in some situations, although it is unlikely to be a serious type of bias given the large sample size in the NIS.

5. Conclusion

In conclusion, palatal procedure was the most frequent and supraglottoplasty was the least frequent type of surgical procedure among hospitalized adult U.S. patients diagnosed with SDB. Also, infectious and respiratory types of complications were frequently observed in this patient population. The study suggested that most surgical procedures were directly associated with a composite measure of surgical complications, whereby palatal procedures exhibited the strongest and supraglottoplasty exhibited the weakest association with these complications. Due to its counterintuitive nature, the inverse association between the composite measure of surgical complications and tracheostomy among hospitalized patients needs further investigation. In fact, tracheostomies were typically performed among older patients, who generally experience more complications, but were also less likely to be performed with other types of OSA-related surgical procedures and more likely to be performed in larger hospitals. Residual confounding by uncontrolled patient and hospital characteristics may explain this study finding. Although patients undergoing tracheostomies may experience fewer complications than those undergoing other types of surgical procedures, this does not rule out the possibility that more severe (but less frequent) complications may be experienced by the former than the latter group of patients. Observed relationships need to be critically evaluated in the absence of a true connection to OSA surgery cases and of granular data needed to adjust for OSA severity and BMI grading (30/32/35 kg/m2), which may be responsible for the pattern of surgical treatments, complications and death rates.

Acknowledgements

Preparation of this manuscript was supported in part by the Intramural Research Program at the National Institute on Aging and Johns Hopkins University School of Medicine. This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Footnotes

Conflicts of interest

The authors declare that they have no conflict of interest in regard to this work.

References

  1. Allareddy V, Martinez-Schlurmann N, Rampa S, Nalliah RP, Lidsky KB, Allareddy V, et al. : Predictors of complications of tonsillectomy with or without adenoidectomy in hospitalized children and adolescents in the United States, 2001-2010: a population-based study. Clin Pediatr (Phila) 55(7): 593–602, 2016 [DOI] [PubMed] [Google Scholar]
  2. Banuchi V, Cohen JC, Kacker A: Safety of concurrent nasal and oropharyngeal surgery for obstructive sleep apnea. Ann Otol Rhinol Laryngol 123(9): 619–622, 2014 [DOI] [PubMed] [Google Scholar]
  3. Becerra MB, Becerra BJ, Teodorescu M: Healthcare burden of obstructive sleep apnea and obesity among asthma hospitalizations: results from the U.S.-based nationwide inpatient sample. Respir Med 117: 230–236, 2016 [DOI] [PubMed] [Google Scholar]
  4. Beydoun HA, Beydoun MA, Jeng HA, Zonderman AB, Eid SM: Bisphenol-a and sleep adequacy among adults in the national health and nutrition examination surveys. Sleep 39(2): 467–476, 2016 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Busaba NY: Same-stage nasal and palatopharyngeal surgery for obstructive sleep apnea: is it safe? Otolaryngol Head Neck Surg 126(4): 399–403, 2002 [DOI] [PubMed] [Google Scholar]
  6. Capobianco DM, Batilana A, Gandhi M, Shah J, Ferreira R, Carvalho E, et al. : Surgical treatment of sleep apnea: association between surgeon/hospital volume with outcomes. Laryngoscope 124(1): 320–328, 2014 [DOI] [PubMed] [Google Scholar]
  7. Choi JH, Cho JH, Chung YS, Kim JW, Kim SW: Effect of the Pillar implant on snoring and mild obstructive sleep apnea: a multicenter study in Korea. Laryngoscope 125(5): 1239–1243, 2015 [DOI] [PubMed] [Google Scholar]
  8. Choi JH, Kim SN, Cho JH: Efficacy of the Pillar implant in the treatment of snoring and mild-to-moderate obstructive sleep apnea: a meta-analysis. Laryngoscope 123(1): 269–276, 2013 [DOI] [PubMed] [Google Scholar]
  9. D’Apuzzo MR, Browne JA: Obstructive sleep apnea as a risk factor for postoperative complications after revision joint arthroplasty. J Arthroplasty 27(8 Suppl): 95–98, 2012 [DOI] [PubMed] [Google Scholar]
  10. Eastwood PR, Barnes M, Walsh JH, Maddison KJ, Hee G, Schwartz AR, et al. : Treating obstructive sleep apnea with hypoglossal nerve stimulation. Sleep 34(11): 1479–1486, 2011 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Fischer Y, Khan M, Mann WJ: Multilevel temperature-controlled radiofrequency therapy of soft palate, base of tongue, and tonsils in adults with obstructive sleep apnea. Laryngoscope 113(10): 1786–1791, 2003 [DOI] [PubMed] [Google Scholar]
  12. Friedman M, Jacobowitz O, Hwang MS, Bergler W, Fietze I, Rombaux P, et al. : Targeted hypoglossal nerve stimulation for the treatment of obstructive sleep apnea: six-month results. Laryngoscope 126(11): 2618–2623, 2016 [DOI] [PubMed] [Google Scholar]
  13. Gamaldo AA, Beydoun MA, Beydoun HA, Liang H, Salas RE, Zonderman AB, et al. : Sleep disturbances among older adults in the United States, 2002–2012: nationwide inpatient rates, predictors, and outcomes. Front Aging Neurosci 8:266, 2016 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Gillespie MB, Wylie PE, Lee-Chiong T, Rapoport DM: Effect of palatal implants on continuous positive airway pressure and compliance. Otolaryngol Head Neck Surg 144(2): 230–236, 2011 [DOI] [PubMed] [Google Scholar]
  15. Gillis E, Rampersaud C, Pease E, Buscemi P: A novel implantable device for a minimally invasive surgical treatment of obstructive sleep apnea: design and preclinical safety assessment. Nat Sci Sleep 8: 249–258, 2016 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Goessler UR, Hein G, Verse T, Stuck BA, Hormann K, Maurer JT: Soft palate implants as a minimally invasive treatment for mild to moderate obstructive sleep apnea. Acta Otolaryngol 127(5): 527–531, 2007 [DOI] [PubMed] [Google Scholar]
  17. Griffin JW, Novicoff WM, Browne JA, Brockmeier SF: Obstructive sleep apnea as a risk factor after shoulder arthroplasty. J Shoulder Elbow Surg 22(12): e6–e9, 2013 [DOI] [PubMed] [Google Scholar]
  18. Hamans E, Shih M, Roue C: A novel tongue implant for tongue advancement for obstructive sleep apnea: feasibility, safety and histology in a canine model. J Musculoskelet Neuronal Interact 10(1): 100–111, 2010 [PubMed] [Google Scholar]
  19. Heidsieck DS, de Ruiter MH, de Lange J: Management of obstructive sleep apnea in edentulous patients: an overview of the literature. Sleep Breath 20(1): 395–404, 2016 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Ishman SL, Ishii LE, Gourin CG: Temporal trends in sleep apnea surgery: 1993-2010. Laryngoscope 124(5): 1251–1258, 2014 [DOI] [PubMed] [Google Scholar]
  21. Jean rE, Gibson CD, Jean RA, Ochieng P: Obstructive sleep apnea and acute respiratory failure: an analysis of mortality risk in patients with pneumonia requiring invasive mechanical ventilation. J Crit Care 30(4): 778–783, 2015 [DOI] [PubMed] [Google Scholar]
  22. Kezirian EJ, Maselli J, Vittinghoff E, Goldberg AN, Auerbach AD: Obstructive sleep apnea surgery practice patterns in the United States: 2000 to 2006. Otolaryngol Head Neck Surg 143(3): 441–447, 2010 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Li S, Shi H: Lingual artery CTA-guided midline partial glossectomy for treatment of obstructive sleep apnea hypopnea syndrome. Acta Otolaryngol 133(7): 749–754, 2013 [DOI] [PubMed] [Google Scholar]
  24. Liu SR, Yi HL, Yin SK, Guan J, Chen B, Meng LL, et al. : Primary maxillomandibular advancement with concomitant revised uvulopalatopharyngoplasty with uvula preservation for severe obstructive sleep apnea-hypopnea syndrome. J Craniofac Surg 23(6): 1649–1653, 2012 [DOI] [PubMed] [Google Scholar]
  25. Mokhlesi B, Hovda MD, Vekhter B, Arora VM, Chung F, Meltzer DO: Sleep-disordered breathing and postoperative outcomes after bariatric surgery: analysis of the nationwide inpatient sample. Obes Surg 23(11): 1842–1851, 2013a [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Mokhlesi B, Hovda MD, Vekhter B, Arora VM, Chung F, Meltzer DO: Sleep-disordered breathing and postoperative outcomes after elective surgery: analysis of the nationwide inpatient sample. Chest 144(3): 903–914, 2013b [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Murphey AW, Baker AB, Soose RJ, Padyha TA, Nguyen SA, Xiao CC, et al. : Upper airway stimulation for obstructive sleep apnea: the surgical learning curve. Laryngoscope 126(2): 501–506, 2016 [DOI] [PubMed] [Google Scholar]
  28. Nelson LM, Barrera JE: High energy single session radiofrequency tongue treatment in obstructive sleep apnea surgery. Otolaryngol Head Neck Surg 137(6): 883–888, 2007 [DOI] [PubMed] [Google Scholar]
  29. Neruntarat C: Hyoid myotomy with suspension under local anesthesia for obstructive sleep apnea syndrome. Eur Arch Otorhinolaryngol 260(5): 286–290, 2003 [DOI] [PubMed] [Google Scholar]
  30. Pathak R, Giri S, Karmacharya P, Aryal MR: Obstructive sleep apnea syndrome and secondary polycythemia: analysis of the nationwide inpatient sample. Sleep Med 16(1): 205–206, 2015 [DOI] [PubMed] [Google Scholar]
  31. Pavelec V, Rotenberg BW, Maurer JT, Gillis E, Verse T: A novel implantable device for the treatment of obstructive sleep apnea: clinical safety and feasibility. Nat Sci Sleep 8: 137–144, 2016 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Plzak J, Zabrodsky M, Kastner J, Betka J, Klozar J: Combined bipolar radiofrequency surgery of the tongue base and uvulopalatopharyngoplasty for obstructive sleep apnea. Arch Med Sci 9(6): 1097–1101, 2013 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Smith DF, Sa T, Fenchel M, Cohen AP, Heubi C, Shott SR, et al. : Temporal trends in inpatient pediatric sleep apnea surgery: 1993 through 2010. Laryngoscope 127(5): 1235–1241, 2017 [DOI] [PubMed] [Google Scholar]
  34. Strohl MDK, Baskin MDJ, Lance MDC, Ponsky MDD, Weidenbecher MDM, Strohl BAM, et al. : Origins of and implementation concepts for upper airway stimulation therapy for obstructive sleep apnea. Respir Investig 54(4): 241–249, 2016 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Stuck BA, Maurer JT, Verse T, Hormann K: Tongue base reduction with temperature-controlled radiofrequency volumetric tissue reduction for treatment of obstructive sleep apnea syndrome. Acta Otolaryngol 122(5): 531–536, 2002 [DOI] [PubMed] [Google Scholar]
  36. Thatcher GW, Maisel RH: The long-term evaluation of tracheostomy in the management of severe obstructive sleep apnea. Laryngoscope 113(2): 201–204, 2003 [DOI] [PubMed] [Google Scholar]
  37. Tuinen MV, Elder S, Link C, Li S, Song JH, Pritchett T: Surveillance of surgery-related adverse events in Missouri using ICD-9-CM Codes In: Henriksen K, Battles JB, Marks ES, Lewin DI (eds), Advances in patient safety: from research to implementation (Volume 1: research findings); 2005, 2005 Rockville, MD: [PubMed] [Google Scholar]
  38. Vigneron A, Tamisier R, Orset E, Pepin JL, Bettega G: Maxillomandibular advancement for obstructive sleep apnea syndrome treatment: long-term results. J Craniomaxillofac Surg 45(2): 183–191, 2017 [DOI] [PubMed] [Google Scholar]

RESOURCES