Perioperative Antibiotic Use in Sleep Surgery: Clinical Relevance
Perioperative Antibiotic Use in Sleep Surgery: Clinical Relevance
Perioperative Antibiotic Use in Sleep Surgery: Clinical Relevance
Otolaryngology–
Head and Neck Surgery
G
Abstract lobally, 900 million adults aged .30 years are diag-
Objective. Upper airway surgery is a common therapeutic nosed with obstructive sleep apnea (OSA), of whom
approach recommended for patients with obstructive sleep 425 million have moderate to severe disease.1 In the
apnea (OSA) to decrease disease burden. We aimed to evalu- United States, 23.5 million people remain undiagnosed in
ate the effect of perioperative antibiotic prescription on com- conservative estimates.2 The detrimental impact of OSA is
plication rates. not confined to health-related burdens but extends to affect
society and the economy, with an estimated annual loss of $65
Study Design. Retrospective cohort (national database). to $165 billion in the United States alone.3
Setting. Tertiary referral center. Uvulopalatopharyngoplasty (UPPP) was the most common
treatment for OSA until the proposal of continuous positive
Methods. This is a retrospective study of a large national health airway pressure (CPAP).4 Although medical management is
care insurance database (Truven MarketScan) from 2007 to currently the gold standard, long-term CPAP compliance may
2015. Subjects diagnosed with OSA who had uvulopalatophar- be an issue and has been reported to be as low as 17%.5 Other
yngoplasty (UPPP) were included and stratified in single versus options, such as oral appliances, expanded compliance to
multilevel surgery. Other variables included smoking, age, sex, medical management, but effectiveness is still limited by
antibiotic prescription, and comorbidities based on the Elixhau- patient tolerability and long-term side effects, including per-
ser index. Evaluated outcomes were postoperative bleeding, manent malocclusion and temporomandibular joint discom-
intubation, pneumonia, superficial surgical site infection, tra- fort. Lack of compliance on a nightly basis has shown to
cheostomy, and hospital readmission. A multivariate regression result in reemergence of sleepiness, neurobehavioral deficits,
model was created to assess each complication. and personality changes.6
Results. A total of 5,798,528 subjects received a diagnosis of Surgical management of OSA has been challenged owing
OSA, of which 39,916 were .18 years old and underwent to possible complications and partial benefit, yet randomized
UPPP, either alone or with additional procedures. The mean trials evaluating UPPP efficacy showed a 60% reduction in
age was 43 years, and 73.4% were male. Antibiotic prescription apnea-hypopnea index.7 Additionally, sleep surgery reduced
was associated with less bleeding in UPPP alone, UPPP with associated systemic diseases when compared with CPAP.8
nasal surgery, and UPPP with nasal and tongue surgery (P \ This population is prone to increased risk of perioperative
.001, P \.001, and P = .006, respectively). It was also associated complications and may require rigorous care.9-11
with a lower prevalence of surgical site infection, pneumonia, Multilevel anatomic obstruction prevails in subjects with
tracheostomy, intubation, and hospital readmission (P \ .001). sleep-disordered breathing and is reported in 90% of patients
On a multivariate model, antibiotic prescription was significantly with OSA.12,13 As a result, the concept of multilevel sleep sur-
associated with a decreased rate of complications. gery was introduced in the early 1990s and gained popular-
ity,13 though with possibly higher complication rates.14,15
Conclusions. Although former studies recommended against
the use of antibiotics after tonsillectomy, our results suggest 1
Division of Sleep Surgery, Department of Otolaryngology–Head and Neck
that antibiotic prescription after UPPP for OSA was associ- Surgery, School of Medicine, Stanford University, Stanford, California, USA
ated with less bleeding, surgical site infection, pneumonia, 2
Department of Biostatistics, Universidade Estadual de Maringá, Maringá,
intubation, tracheostomy, and hospital readmission 30 days Brazil
*
postoperatively. These authors contributed equally to this article.
This article was presented at the AAO-HNSF 2021 Annual Meeting & OTO
Experience; October 3-6, 2021; Los Angeles, California.
Keywords
obstructive sleep apnea, sleep surgery, antibiotics, bleeding, Corresponding Author:
Mohamed Abdelwahab, MD, MS, Division of Sleep Surgery, Department of
uvulopalatopharyngoplasty Otolaryngology–Head and Neck Surgery, School of Medicine, Stanford Uni-
versity, 801 Welch Road, Stanford, CA 94304, USA.
Received March 21, 2021; accepted September 7, 2021. Email: mwahab@stanford.edu
2 Otolaryngology–Head and Neck Surgery
Therefore, we aimed to evaluate modifiable risk factors that procedures according to the additional procedures performed
can help control complication rates in palatal and multilevel on the same day. Antibiotic prescription in each group was
sleep surgery. evaluated 14 days before and 7 days after the procedure. Com-
A multitude of studies have evaluated the role of antibio- plication rates in the 30-day postoperative period were com-
tics in preventing complications following tonsillectomy in pared against patients who received antibiotics in the
children and adults, with controversial results. It is important perioperative period and those who did not.
to understand that unjustified antibiotic prescription can be Exclusion criteria were upper airway procedures that did
associated with increased cost and risk to patients’ general not include UPPP, such as skeletal procedures, maxilloman-
health.16 Among these risks is the development of antibiotic- dibular advancement, distraction osteogenesis maxillary
resistant bacteria, leaving current antibiotics futile.17 Addi- expansion, or isolated genioglossus advancement (GGA).31,32
tional risks include gastrointestinal symptoms, diabetes, obe- Antibiotic prescription was considered when it was given
sity, and bronchial asthma.18-20 prior to the complication day, to define the actual benefit or
In children, current American Academy of Otolaryngology hazard from its use.
guidelines recommend against antibiotic prescription.21,22 In Antibiotic classes were identified by the National Drug
adults there is less clarity, with a trend toward not recom- Codes into the following classes: cephalosporins, macrolides,
mending antibiotics. Interestingly, these studies evaluated penicillins, tetracyclines, and miscellaneous antibiotics
only tonsillectomy indicated for recurrent tonsillitis.23-25 (including clindamycin). The primary outcome was the rate of
Antibiotic prescription following tonsillectomy (or UPPP) for postoperative bleeding and secondary bleeding (a day after
OSA has not been highlighted in most of these trials.23 More- surgery) up to 30 days. Secondary outcomes were superficial
over, OSA has shown to be a comorbidity associated with a surgical site infection (SSI; defined as a diagnosed infection
higher incidence of cardiovascular, endocrine, and neurovas- at least a day after the surgery date), intubation, tracheostomy,
cular disorders.26-28 Studies also showed an increase in com- pneumonia, and hospital readmission (occurring a day after
plications following surgical interventions in subjects with the surgery date) in the same period. SSI was based on the
comorbid OSA.29 diagnosis by the reporting clinicians. Other comorbid diag-
While serious complications after UPPP are uncommon, noses that may affect these outcomes or antibiotic prescription
respiratory complications30 were reported to be 1.1% and were identified through ICD-9 codes. Moreover, we studied
bleeding as low as 0.3%; as such, it could be hypothesized the pattern of antibiotic prescription across years and its
that previous antibiotic clinical trials did not capture a suffi- potential correlation with other variables. Supplementary
cient sample size to show the actual benefit and hazard. In this demographics and comorbidities evaluated herein include
study, we evaluate patterns of antibiotic prescription among age, sex, and tobacco use a year before surgery, as well as a
sleep surgery procedures in a large national database and diagnosis of autoimmune or immunodeficiency disorders,
assess risk factors that may correlate to increased complica- hypertension, and diabetes mellitus prior to surgery (Supple-
tion rates. mental Table S1, available online). These comorbid diagnoses
were identified with the van Walraven modification of the
Methods Elixhauser index, which includes comorbidities into a numeri-
Study Design cal score that is closely associated with mortality in the acute
This is a retrospective study encompassing adults with OSA setting.33,34
(18 years old) who underwent UPPP alone or as part of an
upper airway multilevel surgical treatment. Data were derived Data Source
from the IBM Truven MarketScan Research and Medicare The IBM Truven MarketScan Research and Medicare Data-
Databases for January 1, 2007, to December 31, 2015. Institu- bases capture person-specific clinical utilization, expendi-
tional review was not required at the Stanford University tures, and enrollment across inpatient, outpatient, medication
School of Medicine, where data collection and analysis were prescription, and carve-out services.35
carried out, since this study involved only deidentified com-
mercially available data. The study aimed to evaluate the pat- Statistical Analysis
terns of antibiotic prescription in single and multilevel soft Univariate descriptive statistics were calculated for demo-
tissue sleep surgery and its correlation with postoperative graphic variables and comorbidities. A logistic regression
complications. model was estimated to examine factors associated with each
postoperative complication and the rates of antibiotic pre-
Study Population scription. A stepwise multivariate logistic regression (P \ .05
Patients with an OSA diagnosis were identified per the fol- for inclusion) was used to identify relevant patient character-
lowing ICD-9 codes (International Classification of Diseases, istics for inclusion in the final model. A multinominal logistic
Ninth Revision): 327.23, 327.20, 327.29, 780.51, 780.53, and regression was used to evaluate the P value and odds ratio
780.57 as a primary and/or a component diagnosis. Patients (OR) of complication rates between subjects receiving anti-
with a record of palatopharyngoplasty (421.45) and ton- biotics and not among the multilevel surgery groups. Analysis
sillectomy (428.26) with an OSA diagnosis were included. of variance was used to calculate the difference in complica-
Patients were then divided into single and multilevel surgery tions among the durations of antibiotic prescription. P values
Abdelwahab et al 3
Results
Demographics
A total of 5,798,528 subjects had a diagnosis of OSA in the
analyzed period of January 2007 to December 2015, account-
ing for only 4.34% of those in the database with an OSA diag-
nosis. An overall 41,495 surgical procedures met inclusion
criteria. After exclusion of subjects \18 years old, 39,916 sur-
gical procedures in 38,631 subjects met the inclusion criteria.
There was a small subset that had a revision procedure (3%).
The mean age on the surgery date was 43 years, following
a normal distribution with a 73.4% male predominance
(Figure 1). Complication rates are shown in Figure 2. Demo-
graphic data are shown in Table 1. Two trends were observed:
decreased antibiotic prescription and increased complication
rates over a 9-year period.
Figure 2. Complication numbers and percentage. SSI, surgical site
Outcomes infection.
The impact of antibiotics on complications is shown in
Table 2, where antibiotic prescription was associated with
lower complication rates (P \ .001). Antibiotic prescription (OR = 1.43, P = .003; OR = 1.63, P = .005). Patients under-
was associated with less bleeding in UPPP alone (P \ .001), going nasal surgery with UPPP were associated with less
UPPP with nasal surgery (P \ .001), and UPPP with nasal bleeding (OR = 0.78, P \ .001). Smoking showed a trend of
and tongue surgery (P = .006). Secondary bleeding encom- increased overall bleeding (OR = 1.14, P = .09).
passed 91.24% of total bleeding events, which represented Antibiotics were associated with less superficial SSI (OR =
5.03% of the subjects (Figure 3). The univariate regression 0.65, P \ .001). Subjects with additional GGA, nasal surgery,
model created for evaluating potential risk factors and out- or tongue base procedures were associated with higher rates
comes is shown in Table 3. of SSI (OR = 1.95-5.63, P \ .05). Antibiotic utilization was
The multivariate regression model for complications and associated with significantly less immediate postoperative tra-
hospital readmission is shown in Supplemental Table S2 cheostomy, as well as tracheostomy after the day of surgery,
(available online). Rates of postoperative and secondary shown in Supplemental Table S3 (available online; OR =
bleeding were calculated. Intriguingly, advanced age was 0.02, P \ .001; OR = 0.11, P = .04, respectively). Although
associated with fewer bleeding events (OR = 0.97, P \ .001), the clinical relevance of having an SSI after tonsillectomy
while males had higher rates (OR = 1.88, P \ .001). Antibio- may be questionable, we find the high numbers noted after
tics were associated with fewer overall postoperative and multilevel surgery interesting.
secondary bleeding events (OR = 0.57, P \ .001; OR = 0.64, Multilevel surgery including tongue procedures was asso-
P \ .001, respectively). Immune deficiency was associated ciated with higher tracheostomy rates, such as UPPP with
with a higher risk of postoperative and secondary bleeding tongue procedures (OR = 3.01, P = .046), UPPP with tongue
4 Otolaryngology–Head and Neck Surgery
2007 2008 2009 2010 2011 2012 2013 2014 2015 % No.
Sex
Male 75.34 74.89 75.31 73.85 73.88 74.53 72.89 72.09 69.55 74.10 29,577
Female 24.66 25.11 24.69 26.15 26.12 25.47 27.11 27.91 30.45 25.90 10,339
UPPP 1 surgery 100 100 100 100 100 100 100 100 100 100 39,916
UPPP alone 43.98 45.03 46.58 46.67 48.82 48.49 50.53 53.00 55.34 47.80 19,080
Nose 46.28 44.77 45.70 45.94 44.46 44.93 41.97 41.56 37.73 44.45 17,742
Tongue 3.38 3.44 2.73 2.99 2.91 3.35 3.75 2.78 3.25 3.15 1258
GGA 1.22 1.58 1.17 1.11 0.61 0.33 0.72 0.42 0.07 0.91 363
Nose and GGA 0.99 1.07 0.40 0.56 0.13 0.27 0.25 0.27 0.36 0.51 202
Nose and tongue 3.94 4.06 3.29 2.70 3.01 2.57 2.75 1.91 3.25 3.11 1240
Tongue and GGA 0.11 0.05 0.11 0.04 0.06 0.06 0.03 0.06 0 0.07 26
Nose, tongue, and GGA 0.09 0 0.02 0 0 0 0 0 0 0.01 5
Antibiotic 55.70 57.88 57.53 58.45 53.94 48.80 46.56 43.70 50.72 53.59 21,391
Surgery with complications 5.37 7.04 7.07 6.77 6.99 8.15 7.53 9.17 8.80 7.23 2887
Total complication events 5.46 7.37 7.23 6.96 7.08 8.39 7.78 9.41 9.09 7.44 2968
Bleeding 3.83 4.85 4.83 4.68 5.01 5.47 4.97 6.72 6.49 5.03 2008
Secondary bleeding 3.47 4.48 4.35 4.15 4.59 5.04 4.47 6.20 6.20 4.59 1832
Superficial SSI 1.10 1.65 1.54 1.59 1.52 2.12 1.92 2.12 1.88 1.68 669
Pneumonia 0.27 0.58 0.55 0.50 0.36 0.53 0.69 0.33 0.29 0.48 190
Tracheostomy 0.02 0.12 0.19 0.08 0.11 0.16 0.11 0.18 0.07 0.12 49
Intubation 0.23 0.18 0.11 0.12 0.08 0.10 0.08 0.06 0.36 0.13 52
Hospital readmission 42.79 49.56 54.43 58.55 58.51 59.14 58.64 57.32 57.86 54.82 1314
Risk factor 4.35 4.93 8.66 9.06 9.56 10.51 8.94 10.62 10.89 8.31 21,883
Smoking 25.88 33.44 37.41 41.86 41.13 42.09 41.89 41.07 42.86 38.09 3317
Hypertension 7.94 10.52 11.29 12.84 12.43 12.24 13.31 13.38 12.91 11.68 15,205
Diabetes 14.16 15.55 17.83 18.91 18.93 17.77 17.03 15.86 15.95 17.07 4664
Autoimmune 0.86 1.33 2.26 2.77 3.79 3.85 4.31 5.08 5.56 2.97 6815
Immune deficiency 2.89 3.07 3.00 3.22 3.69 3.68 3.42 3.60 3.10 3.29 1184
Abbreviations: GGA, genioglossus advancement; SSI, surgical site infection; UPPP, uvulopalatopharyngoplasty.
and nasal procedures (OR = 3.42, P = .026), UPPP with GGA and immune deficiency disorders were associated with
(OR = 12.96, P \ .001), and UPPP with GGA and nasal pro- increased odds of hospital readmission (OR = 1.35, P = .001;
cedures (OR = 17.74, P \ .001). Hypertension was also a risk OR = 1.64, P \ .001; OR = 1.25, P = .001; OR = 1.44,
factor (OR = 2.6, P = .003) P = .007).
Male sex was significantly associated with lower pneumo- Furthermore, when antibiotics prescribed across the years
nia rates (OR = 0.7, P = .03). Antibiotic prescription was sig- were evaluated, the most prescribed were penicillin (29.69%),
nificantly associated with less postoperative pneumonia followed by cephalosporine (14.08%) then macrolide and ery-
(overall, OR = 0.37, P \ .001; after the day of surgery, OR = thromycin (5.2%), as shown in Table 4. Table 5 shows the
0.42, P \ .001; Supplemental Tables S2 and S3, respectively, multivariate model for variables associated with antibiotic
available online). Additional nasal surgical procedures had prescription. Patients of male sex and older age tend to receive
higher rates of pneumonia (OR = 1.44, P = .01) as well as more antibiotics (OR = 1.06, P = .012; OR = 1.003, P \ .001,
combined nasal surgery with GGA (OR = 4.58, P = .02). respectively). Subjects undergoing multilevel sleep surgery
Antibiotics were associated with less overall reintubation were more likely to receive antibiotics; this also applied to
(OR = 0.37, P \ .001) while diabetes showed an OR of 2.97 those undergoing additional nasal procedures (OR = 1.48,
(P \ .001). Male sex and antibiotics were associated with P \ .001) tongue procedures (OR = 1.17, P = .006), GGA
lower readmission rates (OR = 0.87, P = .03; OR = 0.37, P \ (OR = 2.48, P \ .001), nasal and tongue procedures (OR =
.001, respectively). As shown in Supplemental Table S2 1.47, P \ .001), and nasal procedures with GGA (OR = 2.38,
(available online), most multilevel surgery was associated P \ .001). Surprisingly, subjects with diabetes, smoking his-
with a higher OR of hospital readmission (OR = 1.29-5.54, P tory, and hypertension received fewer antibiotics (OR = 0.89,
\ .05). Similarly, diabetes, hypertension, and autoimmune P \ .001; OR = 0.86, P \ .001; OR = 0.93, P = .002).
Abdelwahab et al 5
Bleeding
Antibiotic 4.3 3.4 5.0 2.8 3.0 6.3 2.9 25.0 3.8 816
No antibiotic 7.3 5.3 6.9 3.7 6.0 0.0 4.8 0.0 6.4 1192
P value \.001 \.001 .131 .549 .006 .381 ..999 \.001
Odds ratio 0.556 0.591 0.696 0.752 0.467 0.511 1.000 0.555
Bleeding per group, % 6 4 6 3 4 4 3 20 5
Superficial SSI
Antibiotic 0.6 2.3 0.9 2.4 1.9 6.3 0.7 0.0 1.5 320
No antibiotic 0.7 3.4 1.7 5.5 1.5 0.0 4.8 0.0 1.9 349
P value .174 \.001 .185 .029 .708 .039 ..999 \.001
Odds ratio 0.784 0.623 0.509 0.354 1.182 0.128 0.751
Pneumonia
Antibiotic 0.2 0.3 0.5 0.4 0.3 0.0 0.7 0.0 0.3 60
No antibiotic 0.6 0.8 0.7 0.0 1.2 0.0 3.2 0.0 0.7 130
P value \.001 \.001 .554 .344 .044 .136 \.001
Odds ratio 0.371 0.351 0.638 0.400 0.222 0.191 0.377
Tracheostomy
Antibiotic 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1
No antibiotic 0.2 0.2 0.7 2.8 0.8 0.0 3.2 0.0 0.3 48
P value .002 \.001 .111 .003 .059 .051 ..999 ..999 \.001
Odds ratio 0.044 1.000 1.000 0.017
Intubation
Antibiotic 0.1 0.1 0.0 0.0 0.1 0.0 0.7 0.0 0.1 16
No antibiotic 0.2 0.2 0.7 0.0 0.2 0.0 0.0 0.0 0.2 36
P value .005 .103 .111 .735 .776 .742 \.001
Odds ratio 0.269 0.494 0.671 1.000 1.000 0.363
Hospital readmission
Antibiotic 1.9 1.9 1.5 2.0 2.4 12.5 4.3 0.0 1.9 417
No antibiotic 3.9 5.4 7.5 15.6 7.7 0.0 29.0 0.0 4.8 897
P value .001 .001 .001 .001 .001 .001 0 \.001
Odds ratio 0.483 0.331 0.187 0.144 0.281 0.123 1.000 0.381
Total 19,080 17,742 1,258 363 1,240 26 202 5 39,916b
a
The denominator is the number of surgical procedures of the same level that had or did not have antibiotics. Shading indicates a significant P value. Blank
cells indicate not applicable.
b
Antibiotic, n = 21,391; no antibiotic, n = 18,525.
Discussion
In this study, from 39,916 sleep surgery procedures performed
from 2007 to 2015, 53.59% had postoperative antibiotics.
Figure 3. Frequency of bleeding events by day in the 30-day interval While complications were not common, postoperative anti-
after uvulopalatopharyngoplasty for obstructive sleep apnea. biotic prescription was associated with significantly reduced
6 Otolaryngology–Head and Neck Surgery
2007 2008 2009 2010 2011 2012 2013 2014 2015 Total Bleeding
Cephalosporin 16 15 14 15 14 13 13 12 14 14 3.1
Macrolide 1 erythromycin 5 5 6 6 5 5 4 4 4 5 3.1
Penicillin 31 32 32 33 30 27 25 24 28 30 3.4
Tetracycline 0 0 0 0 1 0 0 1 0 0 1.7
Miscellaneous 3 3 2 3 3 3 3 3 3 3 3.4
Quinolone 2 3 2 2 2 1 1 2 1 2 1.8
Sulfonamide 1 1 1 1 1 1 1 1 1 1 2.6
No antibiotics 43 41 41 40 45 50 52 55 48 45 6.4
a
Column percentages may add to .100%, as some had .1 antibiotic class prescribed.
complication rates in single and most multilevel sleep surgery bleeding (risk ratios = 0.49 and 0.9, respectively); however,
(Tables 2 and 3). Duration of antibiotic use was not signifi- this review excluded concomitant tonsillectomies with
cantly associated with reduced rates of bleeding or complica- UPPP.23 Additionally, most of these studies involved either
tions. Interestingly, OSA prevalence in this database (insured children or children and adults, and only 2 included adults.
subjects) was 4.34%, which is quite low when compared with Mann et al had just 8 adults in the antibiotic group.24 O’Reilly
a population-based study (24% in males and 9% in et al conducted the sole study that was composed exclusively
females).37 of adults, finding no justification for perioperative antibio-
Postoperative bleeding has been reported to range from 2% tics.25 Moreover, indications for tonsillectomy were reported
to 40%, with 8% requiring hospital readmission and 3% for recurrent tonsillitis or were not specified; in other words,
undergoing a surgical intervention.30,38-40 We found the aver- OSA was not considered in this review.
age postoperative bleeding incidence to be 5.03%, of which In some trials, antibiotics were not beneficial,23,41,42 while
8.76% occurred the day of surgery (reactionary or primary). in others they were.43,44 Antibiotic prescriptions for tonsillec-
Among subjects prescribed antibiotics, the incidence of sec- tomies in children is not advocated in recent guidelines,21
ondary bleeding was 3.13% versus 6.25% in those not receiv- although adult guidelines remain unclear.
ing antibiotics. In a Cochrane systematic review involving Reduced bacterial load in the tonsillar fossa/surgical site is
1035 patients from 10 clinical trials evaluating the effect of hypothesized to be the main hypothesis for the lower rate of
antibiotics on posttonsillectomy bleeding, data showed no sig- complications. Randomized trials suggesting that antibiotic
nificant reduction in postoperative significant or overall prescription in children was supported by microbiology tests
Abdelwahab et al 7
2007 2008 2009 2010 2011 2012 2013 2014 2015 Total
Hospital readmission 2.9 3.1 3.0 3.2 3.7 3.7 3.4 3.6 3.1 3.3
With bleeding on same day 10 11 13 7 10 12 9 13 5 11
With superficial SSI on same day 0 1 0 1 1 2 2 1 2 1
With tracheostomy on same day 0 1 0 0 0 0 0 0 0 0
With intubation on same day 0 0 1 0 0 0 1 0 2 0
With bleeding over total bleeding 8 7 8 5 8 8 6 7 2 7
With SSI over total superficial SSI 0 2 0 1 1 3 3 1 4 2
With tracheostomy over total tracheostomy 0 14 0 0 0 0 0 0 0 2
With intubation over total intubation 0 0 14 0 0 0 33 0 20 6
showed a 50% reduction of viridans streptococci. Antibiotics Subjects undergoing additional nasal surgery were associ-
were associated with reduced anaerobic bacterial load post- ated with an increased incidence of pneumonia. This is poten-
operatively as compared with those not receiving antibiotics, tially explained by nasal edema, packing, or splints after nasal
who were more likely to grow Enterococcus (40%).43 Amoxi- surgery, which could abolish the protective effect of nasal
cillin clavulanate was shown to reduce pain (prescribed pain breathing in the early postoperative period. This includes
medication) and fasten return to normal diet. In a randomized reduction in nitric oxide, which promotes the ventilation per-
placebo-controlled trial performed on 101 adults (mean, 21.7 fusion ratio and is a potent pulmonary vasodilator.52 Addi-
years), Timentin/Augmentin was associated with a smaller tionally, inhibition of the nasal-ventilator reflex by bypassing
bacterial load without increased Candida growth when com- the nose can result in a decrease in minute and spontaneous
pared with the nonantibiotic group, with significant reduction ventilation.53
in mouth odor, earlier return to physical activity, earlier toler- Although tracheostomy occurred in 0.12% of surgical
ance to a regular diet, and a trend to experience less pain.45 cases, multilevel sleep surgery, particularly with tongue sur-
For antibiotics prescribed from 2007 to 2015, there was a gery or GGA (with and without nasal surgery), was associated
trend toward decreased prescription. This can be explained by with higher tracheostomy rates. We suggest that subjects
current evidence and guidelines that called for reducing the undergoing tongue surgery complain of more odynophagia,
use of postoperative antibiotics after tonsillectomy.46 It is resulting in stagnation of saliva over the wound bed (with a
understandable that this call is to avoid indiscriminate use of subsequent higher bacterial load). This could result in more
antibiotics, which increases the risk of developing antibiotic- inflammation and swelling that can compromise the retrolin-
resistant strains, rendering these medications ineffective.17 gual airway.
Therefore, it is important for otolaryngologists and sleep sur- The results, however, should be cautiously considered.
geons to understand appropriate, evidence-based indications They capture only those with health care insurance and may
for antibiotic use. However, we hypothesized herein that an not reflect the overall incidence of complications. Analysis in
recent increase in complications among those with OSA could this study is limited to the quality of data input by clinicians,
be attributed to this trend. resulting in the possibility that available ICD-9 codes do not
The incidence of SSI is estimated to be 2% to 5%.47 The reflect patients’ true prior-year diagnoses accurately. Being a
Centers for Disease Control and Prevention define SSIs by the database study, this work includes variables that could not
following clinical criteria: necessarily be evaluated, such as compliance of antibiotics
and other comorbidities, including the use of anticoagulant
Purulent exudate draining from a surgical site medications. Identification of differences between those who
Positive culture obtained from a surgical site closed did and did not have complications can be challenging.
initially To our knowledge, this study is the first evaluating antibio-
Surgeon’s diagnosis of infection tic prescription patterns after sleep surgery, based on a large
Surgical site that requires reopening48 population-based database, as compared with previous studies
that could be limited in power. Most studies were single site
The guidelines of the Centers for Disease Control and Pre- and did not account for OSA as a comorbidity or for the dis-
vention and the World Health Organization do not indicate section performed in palatopharyngoplasty procedures. With
the use of prophylaxis antibiotics for clean and clean- the evolution in OSA surgical management, future research
contaminated wounds.49,50 However, its utilization needs to should be directed to randomized trials investigating antibio-
be better evaluated for procedures in the bacteria-ridden aero- tic prescription for this population.
digestive tract, which can have a higher risk of surgical site
contamination. In a review of nasal and oculoplastic proce- Conclusion
dures in .290,000 subjects, Olds et al comparably found that Although former studies recommended against antibiotics
antibiotic prescriptions were associated with less superficial after tonsillectomy, evidence is sparse on hazards and benefits
SSI (OR = 0.144) and deep SSI (OR = 0.254), although guide- for its prescription for UPPP among adults with OSA. Our
lines do not recommend antibiotics for rhinoplasty.51 results suggest a potential protective role of antibiotic pre-
Further comorbidities should be considered when deter- scription in minimizing complication rates, including less
mining whether to prescribe antibiotics after sleep surgery bleeding (postoperative and secondary), SSI, pneumonia,
procedures. In this study, diabetes was associated with an intubation, tracheostomy, and hospital readmission 30 days
increased risk of intubation and overall hospital readmission. postoperatively.
Hypertension was associated with a higher risk of tracheost-
omy and increased rate of hospital readmission. Immune defi- Author Contributions
ciency was associated with higher rates of postoperative and Mohamed Abdelwahab, original conception, manuscript drafting
secondary bleeding and hospital readmission. Surprisingly, and revision, design, data collection, clinical treatment, interpreta-
patients with diabetes and hypertension were actually less tion and analysis, final approval, agreement to be accountable for
likely to receive postoperative antibiotics than the remainder all aspects of the work; Sandro Marques, manuscript drafting and
of the cohort. revision, study design and data interpretation, final approval,
Abdelwahab et al 9
agreement to be accountable for all aspects of the work; Isolde 11. Dart RA, Gregoire JR, Gutterman DD, Woolf SH. The associa-
Previdelli, study design, data collection, statistical analysis and tion of hypertension and secondary cardiovascular disease with
interpretation, final approval and revision, agreement to be accoun- sleep-disordered breathing. Chest. 2003;123(1):244-260.
table for all aspects of the work; Robson Capasso, original con- 12. Abdullah VJ, Van Hasselt CA. Video sleep nasendoscopy. In:
ception, study design, clinical treatment, final approval, and Surgical Management of Sleep Apnea and Snoring. Taylor &
manuscript drafting and revision, agreement to be accountable for
Francis; 2005:143-154.
all aspects of the work.
13. Riley RW, Powell NB, Guilleminault C. Obstructive sleep apnea
Disclosures syndrome: a review of 306 consecutively treated surgical
Competing interests: None. patients. Otolaryngol Head Neck Surg. 1993;108(2):117-125.
Sponsorships: None. 14. Friedman JJ, Salapatas AM, Bonzelaar LB, Hwang MS, Fried-
man M. Changing rates of morbidity and mortality in obstructive
Funding source: None.
sleep apnea surgery. Otolaryngol Head Neck Surg. 2017;157(1):
ORCID iD 123-127.
15. Mickelson SA, Hakim I. Is postoperative intensive care monitor-
Mohamed Abdelwahab https://orcid.org/0000-0002-2588-8355
ing necessary after uvulopalatopharyngoplasty? Otolaryngol
Supplemental Material Head Neck Surg. 1998;119(4):352-356.
Additional supporting information is available in the online version 16. Porco TC, Gao D, Scott JC, et al. When does overuse of antibio-
of the article. tics become a tragedy of the commons? PLoS One. 2012;7(12):
e46505.
References 17. Laxminarayan R, Duse A, Wattal C, et al. Antibiotic resistance—
1. Benjafield AVK, Ayas N, Eastwood PR, Heinzer RC, Ip MS. the need for global solutions. Lancet Infect Dis. 2013;13(12):
Global Prevalence of Obstructive Sleep Apnea in Adults: Esti- 1057-1098.
mation Using Currently Available Data. Abstract presented at: 18. Cox LM, Blaser MJ. Antibiotics in early life and obesity. Nat
American Thoracic Society 2018 International Conference; May Rev Endocrinol. 2015;11(3):182-190.
18-23, 2018; San Diego, CA. 19. Mahana D, Trent CM, Kurtz ZD, et al. Antibiotic perturbation of
2. American Academy of Sleep Medicine. Hidden Health Crisis the murine gut microbiome enhances the adiposity, insulin resis-
Costing America Billions: Underdiagnosing and Undertreating tance, and liver disease associated with high-fat diet. Genome
Obstructive Sleep Apnea Draining Healthcare System. Frost & Med. 2016;8(1):48.
Sullivan; 2016. 20. Stokholm J, Sevelsted A, Bonnelykke K, Bisgaard H. Maternal
3. Harvard Medical School, Division of Sleep Medicine. The Price propensity for infections and risk of childhood asthma: a registry-
of Fatigue: The Surprising Economic Costs of Unmanaged Sleep based cohort study. Lancet Respir Med. 2014;2(8):631-637.
Apnea. McKinsey & Co; 2010. 21. Baugh RF, Archer SM, Mitchell RB, et al. Clinical practice
4. Fujita S, Conway W, Zorick F, Roth T. Surgical correction of guideline: tonsillectomy in children. Otolaryngol Head Neck
anatomic azbnormalities in obstructive sleep apnea syndrome: Surg. 2011;144(1):S1-S30.
uvulopalatopharyngoplasty. Otolaryngol Head Neck Surg. 1981; 22. Mitchell RB, Archer SM, Ishman SL, et al. Clinical practice
89(6):923-934. guideline: tonsillectomy in children (update)—executive sum-
5. Weaver TE, Grunstein RR. Adherence to continuous positive mary. Otolaryngol Head Neck Surg. 2019;160(2):187-205.
airway pressure therapy: the challenge to effective treatment. 23. Dhiwakar M, Clement WA, Supriya M, McKerrow W. Antibio-
Proc Am Thorac Soc. 2008;5(2):173-178. tics to reduce post-tonsillectomy morbidity. Cochrane Database
6. Weaver TE, Maislin G, Dinges DF, et al. Relationship between Syst Rev. 2012;12:CD005607.
hours of CPAP use and achieving normal levels of sleepiness 24. Mann EA, Blair EA, Levy AJ, Chang A. Effect of topical anti-
and daily functioning. Sleep. 2007;30(6):711-719. biotic therapy on recovery after tonsillectomy in adults. Otolar-
7. Browaldh N, Nerfeldt P, Lysdahl M, Bring J, Friberg D. SKUP3 yngol Head Neck Surg. 1999;121(3):277-282.
randomised controlled trial: polysomnographic results after uvu- 25. O’Reilly BJ, Black S, Fernandes J, Panesar J. Is the routine use
lopalatopharyngoplasty in selected patients with obstructive of antibiotics justified in adult tonsillectomy? J Laryngol Otol.
sleep apnoea. Thorax. 2013;68(9):846-853. 2003;117(5):382-385.
8. Ibrahim B, de Freitas Mendonca MI, Gombar S, Callahan A, 26. Punjabi NM, Newman AB, Young TB, Resnick HE, Sanders
Jung K, Capasso R. Association of systemic diseases with surgi- MH. Sleep-disordered breathing and cardiovascular disease: an
cal treatment for obstructive sleep apnea compared with continu- outcome-based definition of hypopneas. Am J Respir Crit Care
ous positive airway pressure. JAMA Otolaryngol Head Neck Med. 2008;177(10):1150-1155.
Surg. 2021;147(4):329-335. 27. Ferini-Strambi L, Lombardi GE, Marelli S, Galbiati A. Neurolo-
9. Connolly LA. Anesthetic management of obstructive sleep gical deficits in obstructive sleep apnea. Curr Treat Options
apnea patients. J Clin Anesth. 1991;3(6):461-469. Neurol. 2017;19(4):16.
10. Johnson JT, Braun TW. Preoperative, intraoperative, and post- 28. Wang X, Bi Y, Zhang Q, Pan F. Obstructive sleep apnoea and
operative management of patients with obstructive sleep apnea the risk of type 2 diabetes: a meta-analysis of prospective cohort
syndrome. Otolaryngol Clin North Am. 1998;31(6):1025-1030. studies. Respirology. 2013;18(1):140-146.
10 Otolaryngology–Head and Neck Surgery
29. Chopra T, Marchaim D, Lynch Y, et al. Epidemiology and out- 42. Gil-Ascencio M, Castillo-Gomez CJ, Palacios-Saucedo Gdel C,
comes associated with surgical site infection following bariatric Valle-de la OA. Antibiotic prophylaxis in tonsillectomy and its
surgery. Am J Infect Control. 2012;40(9):815-819. relationship with postoperative morbidity. Acta Otorrinolarin-
30. Kezirian EJ, Weaver EM, Yueh B, et al. Incidence of serious gol Esp. 2013;64(4):273-278.
complications after uvulopalatopharyngoplasty. Laryngoscope. 43. Colreavy MP, Nanan D, Benamer M, et al. Antibiotic prophy-
2004;114(3):450-453. laxis post-tonsillectomy: is it of benefit? Int J Pediatr Otorhino-
31. Abdelwahab M, Poomkonsarn S, Ren X, et al. A comprehensive laryngol. 1999;50(1):15-22.
strategy for improving nasal outcomes after large maxilloman- 44. Abdelhamid AO, Sobhy TS, El-Mehairy HM, Hamid O. Role of
dibular advancement for obstructive sleep apnea. Facial Plast antibiotics in post-tonsillectomy morbidities: a systematic
Surg Aesthet Med. Published online July 21, 2021. doi:10.1089/ review. Int J Pediatr Otorhinolaryngol. 2019;118:192-200.
fpsam.2020.0569 45. Grandis JR, Johnson JT, Vickers RM, et al. The efficacy of peri-
32. Abdelwahab M, Yoon A, Okland T, Poomkonsarn S, Gouveia C, operative antibiotic therapy on recovery following tonsillectomy
Liu SY-C. Impact of distraction osteogenesis maxillary expan- in adults: randomized double-blind placebo-controlled trial.
sion on the internal nasal valve in obstructive sleep apnea. Oto- Otolaryngol Head Neck Surg. 1992;106(2):137-142.
laryngol Head Neck Surg. 2019;161(2):362-367. 46. Patel PN, Jayawardena ADL, Walden RL, Penn EB, Francis DO.
33. van Walraven C, Austin PC, Jennings A, Quan H, Forster AJ. A Evidence-based use of perioperative antibiotics in otolaryngol-
modification of the Elixhauser comorbidity measures into a ogy. Otolaryngol Head Neck Surg. 2018;158(5):783-800.
point system for hospital death using administrative data. Med 47. Anderson DJ, Podgorny K, Berrios-Torres SI, et al. Strategies to
Care. 2009;47(6):626-633. prevent surgical site infections in acute care hospitals: 2014
34. Elixhauser A, Steiner C, Harris DR, Coffey RM. Comorbidity update. Infect Control Hosp Epidemiol. 2014;35(suppl 2):S66-
measures for use with administrative data. Med Care. 1998; S88.
36(1):8-27. 48. Mangram AJ, Horan TC, Pearson ML, Silver LC, Jarvis WR;
35. IBM MarketScan Research Databases for health services Centers for Disease Control and Prevention (CDC) Hospital
researchers. Published 2019. https://www.ibm.com/downloads/ Infection Control Practices Advisory Committee. Guideline for
cas/6KNYVVQ2 prevention of surgical site infection, 1999. Am J Infect Control.
36. R Core Team. R: A Language and Environment for Statistical 1999;27(2):97-132.
Computing. R Foundation for Statistical Computing; 2019. 49. Berrios-Torres SI, Umscheid CA, Bratzler DW, et al. Centers for
37. Young T, Palta M, Dempsey J, Skatrud J, Weber S, Badr S. The Disease Control and Prevention guideline for the prevention of
occurrence of sleep-disordered breathing among middle-aged surgical site infection, 2017. JAMA Surg. 2017;152(8):784-791.
adults. N Engl J Med. 1993;328(17):1230-1235. 50. World Health Organization. Global Guidelines for the Preven-
38. Wei JL, Beatty CW, Gustafson RO. Evaluation of posttonsillect- tion of Surgical Site Infections. World Health Organization;
omy hemorrhage and risk factors. Otolaryngol Head Neck Surg. 2016.
2000;123(3):229-235. 51. Olds C, Spataro E, Li K, Kandathil C, Most SP. Postoperative
39. Evans AS, Khan AM, Young D, Adamson R. Assessment of sec- antibiotic use among patients undergoing functional facial plas-
ondary haemorrhage rates following adult tonsillectomy—a tele- tic and reconstructive surgery. JAMA Facial Plast Surg. 2019;
phone survey and literature review. Clin Otolaryngol Allied Sci. 21(6):491-497.
2003;28(6):489-491. 52. Haight JS, Djupesland PG. Nitric oxide (NO) and obstructive
40. van der Meulen J. Tonsillectomy technique as a risk factor for sleep apnea (OSA). Sleep Breath. 2003;7(2):53-62.
postoperative haemorrhage. Lancet. 2004;364(9435):697-702. 53. McNicholas WT, Coffey M, Boyle T. Effects of nasal airflow on
41. Guerra MM, Garcia E, de Mendoncxa Pilan RR, Rapoport PB, breathing during sleep in normal humans. Am Rev Respir Dis.
Campanholo CB, Martinelli EO. Antibiotic use in post- 1993;147(3):620-623.
adenotonsillectomy morbidity: a randomized prospective study.
Braz J Otorhinolaryngol. 2008;74(3):337-341.