Вы находитесь на странице: 1из 6

International Journal of Surgery 54 (2018) 170–175

Contents lists available at ScienceDirect

International Journal of Surgery

journal homepage: www.elsevier.com/locate/ijsu

Original Research

Perioperative antibiotic prophylaxis in open tracheostomy: A preliminary T

randomized controlled trial
Pichit Sittitraia,∗, Chatmanee Siriwittayakornb
Department of Otolaryngology, Faculty of Medicine, Chiang Mai University, Chiang Mai, 50200, Thailand
Otolaryngology Unit, Nakornping Hospital, Chiang Mai, 50180, Thailand


Keywords: Background: The efficacy of perioperative antibiotic prophylaxis for prevention of wound infection in open
Tracheostomy tracheostomy has been minimally studied and remains controversial.
Surgical wound infection Methods: A preliminary double-blind, randomized, placebo-controlled trial was conducted. A total of 159 pa-
Antibiotic prophylaxis tients who underwent open tracheostomy were enrolled, and 88 patients were excluded because of lack of desire
to participate in research, emergency condition, administration of other antibiotics, immunocompromise, or
cervical skin infection. The remainings were randomly assigned to an antibiotic group or a control group.
Another 11 patients were excluded after the randomization due to intraoperative contamination, death from the
underlying disease, receiving other antibiotics, or lost to the follow-up. A total of 30 patients in each group were
qualified for analysis. In the antibiotic group, clindamycin was intravenously administered 30 min before the
incision and every 8 h after the operation for 24 h. In the control group, an equal volume of sterile saline was
Results: Wound infection developed in 2 patients (6.7%) in the antibiotic group and 7 patients (23.3%) in the
control group (p = 0.08). In multivariate analysis, smoking and previous neck irradiation were the significant
risk factors for wound infection (p = 0.042 and 0.019, respectively). The mean length of hospital stay after
tracheostomy in patients with and without wound infection were 17 ± 2 days and 4 ± 2 days, respectively
(p = 0.013).
Conclusion: The result of this preliminary study reveals that antibiotic prophylaxis reduced tracheostomy wound
infection rate from 23.3% to 6.7% although it was not statistically significant. However, wound infection may
lead to serious complications and prolonged postoperative length of hospital stay, and therefore proper perio-
perative antibiotics should be considered in patients who are not receiving other antibiotics, and particularly in
patients with risk factors for wound infection.

1. Introduction tracheostomy still has a main role for emergency situations, difficult
neck anatomy, previous tracheostomy, patients with comorbidities, and
Tracheostomy is one of the most common operations performed in failure of percutaneous tracheostomy [1].
the hospital [1,2]. It has been done as an emergency or elective pro- The incidence of complications associated with open tracheostomy
cedure [1,2]. Initially, tracheostomy was performed for upper airway has been reported between 5% and 40% which include hemorrhage,
obstruction and later the indications were extended to include re- apnea, adjacent tissue injury, tube displacement, subcutaneous em-
spiratory failure, prolonged endotracheal intubation and ventilation, physema, pneumothorax, and infection [5–7]. The incidence of infec-
pulmonary toilet, and as an adjunct to surgery [1–3]. Open or surgical tion following elective open tracheostomy has been reported up to 33%
tracheostomy has been widely accepted as the standard of care for a [8–10]. Tracheotomy is considered a clean-contaminated wound as the
long period of time and mostly performed in the operating room [1]. procedure enters the upper aerodigestive tract which has bacterial co-
Since 1985, when percutaneous dilatational tracheostomy was in- lonization and may also become contaminated by the bacterial flora of
troduced by Ciaglia [4], the procedure has been rapidly adopted and the skin especially in urgent or emergency tracheostomy [5,9]. Wound
has increased in popularity because of its safety, convenience, and infection may produce serious complications such as tracheitis, med-
technical advantages of bedside procedure [1]. However, open iastinitis, clavicular osteomyelitis, and necrotizing fasciitis [5–7].

Corresponding author.
E-mail addresses: psittitrai@yahoo.com (P. Sittitrai), chatmaneesiri@gmail.com (C. Siriwittayakorn).

Received 17 January 2018; Received in revised form 6 April 2018; Accepted 24 April 2018
Available online 02 May 2018
1743-9191/ © 2018 IJS Publishing Group Ltd. Published by Elsevier Ltd. All rights reserved.
P. Sittitrai, C. Siriwittayakorn International Journal of Surgery 54 (2018) 170–175

However, perioperative antibiotics to prevent surgical site infection in 30 min before the surgical incision and every 8 h after the operation for
tracheostomy have not been generally considered because of minimal 24 h for a total of 4 doses. An equal volume of sterile saline was ad-
evidence of their effectiveness [11,12]. ministered to the control group in the same manner.
The purpose of this randomized controlled trial was to clarify
whether antibiotic prophylaxis for open tracheostomy would reduce 2.4. Follow-up
infection rate and to identify risk factors for tracheostomy wound in-
fection. Surgical site infection is classified as wound infection which occurs
within 30 days after the operation [14]. The wounds were carefully
2. Materials and methods examined every day for 30 days by surgeons who were blinded to the
randomization. Patients who were discharged from the hospital earlier
2.1. Study design than day 30 would be instructed to record wound characteristics, and
follow-up at day 30 after the operation. The criteria of wound infection
The preliminary prospective randomized controlled trial was con- included: 1) purulent drainage, 2) identification of organisms, 3) in-
ducted in the Department of Otolaryngology, Faculty of Medicine, be- flammation accompanied by fever or tenderness, 4) abscess or other
tween February 2015 and January 2017. The study protocol was per- evidence of infection around the wound detected on gross anatomical
formed in compliance with the guidelines for Good Clinical Practice and or histopathological exam, or imaging test, or 5) diagnosis of a wound
the Declaration of Helsinki and approved by the institutional ethics infection by the surgeon [14]. Wound discharge was sent for aerobic
committee. All the subjects signed an informed consent form prior to bacterial culture and sensitivity. Anaerobic bacterial culture was not
enrollment. done because the tracheostomy wound is an open wound and no pus
The inclusion criteria were patients between 18 and 85 years of age collection in the deep tissues was detected in any patients. The infected
who had open tracheostomy performed with the following indications: wound was managed with intensive local wound dressing, daily tra-
1) upper airway obstruction from tumor, airway edema or stenosis, cheostomy tube change, and antibiotics change based on the culture
neck or maxillofacial trauma, or foreign body, 2) prolonged en- results.
dotracheal intubation, and 3) facilitating pulmonary toilet. Selected demographic data (sex, age, history of smoking, diabetes
The exclusion criteria were patients who had immunosuppressive mellitus, previous neck irradiation, preoperative albumin level, pre-
conditions, allergy to clindamycin, diagnosis of cervical skin infection operative hemoglobin level), indication for tracheostomy, and pre-
or tracheobronchopulmonary infection, administration of other anti- operative length of hospital stay were documented and compared be-
biotics within 7 days before tracheostomy or within 14 days after tra- tween study groups. The outcomes including infection rate, causative
cheostomy, and patients who were deceased or unavailable for follow- organism, duration of postoperative hospitalization, and surgical com-
up in the 30 days after tracheostomy. plication were recorded.
The randomization process was carried out using the CONSORT
statement 2010 checklist for randomized controlled clinical trials. 2.5. Statistical analysis
Patients were assigned in a double-blinded method to either an anti-
biotic group or a control group. A randomized block design by a com- Statistical analyses were performed using SPSS (version 22) for
puter-generated list was used to allocate patients into the two groups. Windows (IBM Corporation, Armonk, NY, USA). Continuous variables
The surgery team, the surgeon who instructed patients regarding the were compared using Student's t tests or Mann–Whitney U tests, and
intravenous injection and the surgeon who evaluated the wound were categorical variables were compared using chi-square tests or Fisher's
all blinded to the allocation group. In addition, the patients were not exact tests as appropriate. For analysis of risk factors, the chi-square
informed about the contents of injection. association test was used. Then, a logistic regression model was out-
lined based on a stepwise forward method testing the significant vari-
2.2. Surgical technique ables at univariate and multivariate analysis. The risk measure used was
the odds ratio (OR), and its respective 95% confidence interval was
Tracheostomy was generally performed in the operating room under calculated. Statistical significance was considered for p values less than
general anesthesia or monitored anesthesia care. The patient was po- 0.05.
sitioned with the neck extended. An aqueous solution of 10% povidone-
iodine was used for skin disinfection before the operation. A 3–4 cm 3. Results
transverse skin incision was made midway between the sternal notch
and cricoid cartilage or two fingerbreadths above the sternal notch. The There were 159 eligible patients during the study period. Before
strap muscles were separated along midline and retracted laterally. The randomization, 88 patients were excluded because of lack of desire to
thyroid isthmus was retracted superiorly, inferiorly, or divided. A participate in research, emergency condition, administration of other
cruciate incision was made at the second and third tracheal rings. Two antibiotics, immunocompromise, or cervical skin infection. The re-
stay sutures were used to secure the tracheal opening. The plastic cuffed maining 71 patients were randomly assigned to either antibiotic group
tracheostomy tube was inserted. The skin incision was not sutured (n = 36) or control group (n = 35). Six patients in the antibiotic group
unless it was larger than 5 cm. The tracheostomy tube tape was placed and five patients in the control group were excluded after randomiza-
and secured around the neck. A 4x4-inch sterile gauze was cut to allow tion because of intraoperative contamination, death due to underlying
placement under the tracheostomy tube. In postoperative care, the diseases, receipt of other antibiotics during the postoperative period,
wound and surrounding skin were cleaned with 70% ethyl alcohol and and loss to follow-up. Therefore, data from 30 patients in each group
the sterile gauze was changed twice daily. were analyzed (Fig. 1).
There were no statistically significant intergroup differences in de-
2.3. Intervention mographic parameters and indication for tracheostomy which con-
firmed the correct selection of the subjects and randomization of the
Because the common pathogens responsible for tracheostomy groups and enabled further analyses (Table 1). Because of the exclusion
wound infection were reported as Staphylococcus spp. and Streptococcus criteria of no other perioperative antibiotics, the only indication for
spp [5,11,13], the selected antibiotic in the study was clindamycin. tracheostomy in this study was upper airway obstruction from tumor or
Patients in the antibiotic prophylaxis group received 100 mL of sterile stenosis.
saline with 600 mg of clindamycin by continuous intravenous infusion Wound infection was detected in 9 patients with the overall

P. Sittitrai, C. Siriwittayakorn International Journal of Surgery 54 (2018) 170–175

Fig. 1. Flowchart of participants in randomized controlled trial of perioperative antibiotic prophylaxis versus placebo in open tracheostomy.

Table 1
Comparison of demographic data between the study groups.
Variables Antibiotic group Control group p value
N = 30 (%) N = 30 (%)

Age, year (mean ± SD) 49.9 ± 11.9 48.8 ± 9.5 0.66

M 23 (76.7) 25 (83.3) 0.75
F 7 (23.3) 5 (16.7)
Cause of airway obstruction
Laryngohypopharyngeal cancer 20 (66.7) 19 (63.3) 0.72
Thyroid cancer 2 (6.7) 4 (13.3)
Tracheal tumor 1 (3.3) 2 (6.7)
Laryngotracheal stenosis 7 (23.3) 5 (16.7)
Yes 5 (16.7) 6 (20) 1.0
Diabetes mellitus
Yes 4 (13.3) 3 (10) 1.0
Previous neck irradiation
Yes 2 (6.7) 4 (13.3) 0.67
Albumin level, g/dl (mean ± SD) 3.5 ± 0.1 3.6 ± 0.2 0.46
Hemoglobin level, g/dl (mean ± SD) 12.9 ± 0.6 12.7 ± 0.5 0.17
Preoperative hospitalization
< 6h 1 (3.3) 4 (13.3) 0.5
6–24 h 14 (46.7) 11 (36.7)
24–72 h 12 (40) 11 (36.7)
> 72 h 3 (10) 4 (13.3)

P. Sittitrai, C. Siriwittayakorn International Journal of Surgery 54 (2018) 170–175

Table 2
Comparison of outcomes between the study groups.
Outcomes Antibiotic group Control group p value
N = 30 (%) N = 30 (%)

Yes 2 (6.7) 7 (23.3) 0.08
Time to infection, day (mean ± SD) 3.0 ± 1.4 3.7 ± 0.9 0.42
Postoperative hospitalization, day (mean ± SD) 5.1 ± 3.9 7.1 ± 5.9 0.13
Yes 3 (10) 5 (16.7) 0.71
Yes 2 (6.7) 1 (3.3) 1.0

Table 3 Table 4
Univariate analysis of risk factors for tracheostomy wound infection. Details of preoperative hospitalization duration and bacterial pathogens in nine
patients who developed wound infection.
Variables Infection No infection p value
N = 9 (%) N = 51 (%) Patient group Preoperative hospitalization Pathogen
duration (h)
Yes 2 (22.2) 28 (54.9) 0.18 Antibiotic > 72 Pseudomonas aeruginosa
Age, year Klebsiella pneumoniae
< 60 8 (90) 42 (82.4) 0.45 Antibiotic > 72 Pseudomonas aeruginosa
≥ 60 1 (10) 9 (17.6) Proteus mirabilis
Sex Control <6 Staphylococcus aureus
M 9 (100) 39 (76.5) 0.99 Control <6 Staphylococcus aureus
F 0 12 (23.5) Streptococcus viridans
Cause of airway obstruction Control <6 Streptococcus viridans
Laryngohypopharyngeal cancer 8 (88.9) 31 (60.8) 0.99 Control <6 Streptococcus pyogenes
Thyroid cancer 0 6 (11.8) 0.99 Control > 72 Pseudomonas aeruginosa
Tracheal tumor 1 (11.1) 2 (3.9) 1.0 Proteus mirabilis
Laryngotracheal stenosis 0 12 (23.5) 0.99 Streptococcus viridans
Smoking Control > 72 Klebsiella pneumoniae
Yes 8 (88.9) 7 (13.7) < 0.001 Proteus mirabilis
Diabetes mellitus Control > 72 Proteus mirabilis
Yes 5 (55.6) 2 (3.9) 0.09
Previous radiation
Yes 6 (66.7) 1 (2) < 0.001
bacterial infection were hospitalized for less than 6 h (Table 4). The
Albumin level, g/dl
< 3.5 7 (77.8) 6 (11.8) 0.001 mean length of hospital stay after the operation in patients with and
≥ 3.5 2 (22.2) 45 (88.2) without wound infection were 17 ± 2 days and 4 ± 2 days, respec-
Hemoglobin level, g/dl tively which was significantly different (p = 0.013).
< 12.5 1 (11.1) 13 (25.5) 0.36
Regarding the bacteriology results, the isolated Gram-positive bac-
≥ 12.5 8 (88.9) 38 (74.5)
Preoperative hospitalization
teria were Staphylococcus aureus, Streptococcus viridans, and
< 6h 3 (33.3) 2 (3.9) 0.81 Streptococcus pyogenes while the isolated Gram-negative bacteria were
6–24 h 0 25 (49) 0.33 Pseudomonas aeruginosa, Klebsiella pneumonia, and Proteus mirabilis.
24–72 h 0 23 (45.1) 0.99 Other complications noted in the study were minor bleeding and
> 72 h 6 (66.7) 1 (1.9) 0.99
pneumothorax, with no significant difference noted between groups for
either complication (Table 2). Minor bleeding was observed in eight
infection rate of 15%, and two patients were in the antibiotic group patients, three in the antibiotic group and five in the control group. All
(infection rate, 6.7%) and seven patients in the control group (infection patients with bleeding were detected within two days after the opera-
rate, 23.3%) (p = 0.08). Characteristics of wound infection were de- tion, and were treated successfully with packing. All three patients with
tected as cellulitis with turbid discharge or with purulent discharge. All pneumothorax were detected during or within one hour of the opera-
wound infection occurred during day two to day five after the operation tion, and were managed with tube drainage. There was no record of any
(Table 2). major complication or perioperative mortality related to the procedure.
In univariate analysis, eight variables were analyzed to determine
the risk factors for wound infection (Table 3). History of smoking,
4. Discussion
previous neck irradiation, and albumin level < 3.5 g/dl were associated
with the development of infection (p < 0.001, p < 0.001, and
Complications of open tracheostomy are consistent, in terms of type,
p = 0.001, respectively). When using multivariate analysis, only
frequency, and severity, with wound infection is second in incidence
smoking (p = 0.042, 95% CI = 22, [1.1–46.7]) and previous neck ir-
only to bleeding [6,15]. Exposure to contaminated oral secretions, in-
radiation (p = 0.019, 95% CI = 31.5, [1.7–65.9]) had a significant ef-
fected sputum, and the colonizing flora of skin provide favorable con-
fect on wound infection while albumin level did not (p = 0.051).
ditions for tracheotomy infection [5–7,9]. Wound infection includes
Both infected patients in the antibiotic group were hospitalized
local tissue inflammation, cellulitis, and abscess which may progress to
longer than 72 h (6 days, and 7 days) before having tracheostomy and
the serious conditions such as mediastinitis and necrotizing fasciitis
the causative organisms were Gram-negative bacteria. Of the seven
infected patients in the control group, three patients who had either
Antibiotic prophylaxis initiated prior to skin incision and continued
pure Gram-negative (2 patients) or mixed Gram-positive and Gram-
over the period of 24 h after surgery has been documented to reduce
negative (1 patient) bacterial infection were hospitalized longer than
surgical site infection in clean-contaminated head and neck surgery
72 h preoperatively. The other four patients with pure Gram-positive
[16]. Clindamycin was selected for antibiotic prophylaxis in this study

P. Sittitrai, C. Siriwittayakorn International Journal of Surgery 54 (2018) 170–175

because of its effectiveness in preventing head and neck surgical site Study code: 2556-01852.
infection [16,17] and because the causative organisms of tracheostomy ID: 01852.
wound infection are considered to be Staphylococcus spp. and Strepto-
coccus spp. [5,11,13]. Funding
All nine patients with wound infection in this study developed the
infection within five days after the operation with an overall infection None.
rate of 15%. This is consistent with the rates reported in a meta-analysis
(range, 0%–33.3%) [9]. The wound infection rate was reduced from Author contribution
23.3% (seven patients) in the control group to 6.7% (two patients) in
the antibiotic group although it was close, it was not statistically sig- Pichit Sittitrai designed the study, developed the methodology,
nificant (p = 0.08). collected the data, performed the analysis, and wrote the manuscript.
It should be noted that the four infected patients in the control Chatmanee Siriwittayakorn designed the study, collected the data,
group with Gram-positive bacterial infection were hospitalized less and wrote the manuscript.
than 6 h before the operation and all of them were sensitive to clin-
damycin. The other five infected patients in both groups who were Conflicts of interest
hospitalized longer than 72 h had either pure Gram-negative or mixed
Gram-positive and Gram-negative organisms, and these pathogens are a The authors declare no competing financial interests exist.
common cause of hospital-acquired infection. According to this finding,
the causative organisms in patients who were recently hospitalized Research registration Unique Identifying Number (UIN)
before tracheostomy were most likely to be Gram-positive bacteria,
while the possible causative organisms in patients who were hospita- UIN3575.
lized longer than 72 h were either pure Gram-negative or mixed Gram-
positive and Gram-negative bacteria. Guarantor
The risk factors for wound infection in head and neck surgery have
been reported to include smoking, alcohol, albumin, hemoglobin, dia- Pichit Sittitrai.
betes mellitus, previous neck irradiation, chemotherapy, and poor oral Saisaward Chiyasate.
hygiene [18]. In this study, smoking, previous neck irradiation, and Department of Otolaryngology, Faculty of Medicine, Chiang Mai
albumin level < 3.5 g/dl were identified as factors increasing risk of University.
wound infection in univariate analysis while only smoking and previous
radiotherapy were revealed as risk factors in multivariate analysis. Author disclosure statement
Based on these findings, the use of antibiotic prophylaxis to prevent
tracheostomy wound infection should be considered in high risk pa- No competing financial interests exist.
tients who have history of smoking or previous neck irradiation. Based
on the culture results in the study, antibiotics which include Gram- Acknowledgments
positive bacterial coverage such as clindamycin, antistaphylococcus
penicillins, or first-generation cephalosporins are the suitable anti- The authors would like to thank Dr. for assistance in statistical
biotics in patients who are hospitalized within 6 h prior to the tra- analysis.
cheostomy and antibiotics which cover both Gram-positive and Gram-
negative bacteria such as clindamycin combined with third-generation References
cephalosporin are the preferred choice in patients who are hospitalized
longer than 72 h. [1] K.P. Alfonso, M.R. Kaufman, E.V. Dressler, et al., Otolaryngology consultation
There are limitations in the study. First, this is a single-center ran- tracheostomies and complex patient population, Am. J. Otolaryngol. 38 (2017)
domized study with a rather small sample size. Second, our study did [2] L. Muallem-Kalmovich, J. Pitaro, A. Asaly, et al., Open tracheostomy training: a
not include emergency or urgent tracheostomy which may have a high nationwide survey among Otolaryngology-Head and Neck Surgery residents, Eur.
rate of wound infection without antibiotic prophylaxis because of the Arch. Oto-Rhino-Laryngol. 274 (2017 Nov) 4035–4042.
[3] D. Goldenberg, A. Golz, A. Netzer, H.Z. Joachims, Tracheotomy: changing indica-
inability to obtain informed consent for the study with this specific si- tions and a review of 1,130 cases, J. Otolaryngol. 31 (2002) 211–215.
tuation. [4] P. Ciaglia, R. Firsching, C. Syniec, Elective percutaneous dilatational tracheostomy.
In conclusion, this preliminary study reveals that the tracheostomy A new simple bedside procedure; preliminary report, Chest 87 (1985 Jun) 715–719.
[5] D. Goldenberg, E.G. Ari, A. Golz, et al., Tracheotomy complications: a retrospective
wound infection rate was reduced from 23.3% to 6.7% with perio-
study of 1130 cases, Otolaryngol. Head Neck Surg. 123 (2000) 495–500.
perative clindamycin although it was not statistically significant. [6] R.C. Wang, P.W. Perlman, S.M. Parnes, Near-fatal complications of tracheotomy
However, with the possible serious complications and prolonged post- infections and their prevention, Head Neck 11 (1989) 528–533.
[7] A. Cipriano, M.L. Mao, H.H. Hon, et al., An overview of complications associated
operative length of hospital stay, antibiotic prophylaxis should be
with open and percutaneous tracheostomy procedures, Int J Crit Illn Inj Sci. 5
considered in patients who are currently not receiving other antibiotics, (2015) 179–188.
and particularly in patients who are at risk of wound infection such as a [8] S. Briggs, J. Ambler, D. Smith, A survey of tracheostomy practice in a cardiothoracic
history of smoking or previous neck irradiation. Perioperative anti- intensive care unit, J. Cardiothorac. Vasc. Anesth. 21 (2007) 76–80.
[9] A. Delaney, S.M. Bagshaw, M. Nalos, Percutaneous dilatational tracheostomy versus
biotics which cover Gram-positive bacteria should be selected in pa- surgical tracheostomy in critically ill patients: a systematic review and meta-ana-
tients who are hospitalized less than 6 h prior to the tracheostomy and lysis, Crit. Care 10 (2) (2006) R55 https://doi.org/10.1186/cc4887.
antibiotics which cover both Gram-positive and Gram-negative bacteria [10] E. Cheng, W.E. Fee Jr., Dilatational versus standard tracheostomy: a meta-analysis,
Ann. Otol. Rhinol. Laryngol. 109 (2000) 803–807.
for patients who are hospitalized longer than 72 h. With the trend to- [11] H. Hagiya, H. Naito, S. Hagioka, et al., Effects of antibiotics administration on the
ward decreasing tracheostomy wound infection rate with perioperative incidence of wound infection in percutaneous dilatational tracheostomy, Acta Med.
antibiotics, this warrants further investigation in a larger study. Okayama 68 (2014) 57–62.
[12] P. De Leyn, L. Bedert, M. Delcroix, et al., Tracheotomy:clinical review and guide-
lines, Eur. J. Cardio. Thorac. Surg. 32 (2007) 412–421.
Ethical approval [13] G. Cardone, M. Lepe, Tracheostomy: complications in fresh postoperative and late
postoperative settings, Clin. Pediatr. Emerg. Med. 11 (2010) 122–130.
[14] A.J. Mangram, T.C. Horan, M.P. Pearson, et al., Guideline for prevention of surgical
The study was approved by the institutional ethics committee of site infection, 1999. Hospital infection control practices advisory committee, Infect.
Faculty of Medicine, Chiang Mai University.

P. Sittitrai, C. Siriwittayakorn International Journal of Surgery 54 (2018) 170–175

Control Hosp. Epidemiol. 20 (1999) 250–278. of prophylactic antibiotic matter? J. Laryngol. Otol. 122 (2008) 403–408.
[15] P. Hazard, C. Jones, J. Benitone, Comparative clinical trial of standard operative [17] W.R. Carroll, D. Rosenstiel, J.R. Fix, et al., Three-dose vs extended-course clin-
tracheostomy with percutaneous tracheostomy, Crit. Care Med. 19 (1991) damycin prophylaxis for free-flap reconstruction of the head and neck, Arch.
1018–1024. Otolaryngol. Head Neck Surg. 129 (2003) 771–774.
[16] S.A. Liu, K.C. Tung, J.Y. Shiao, et al., Preliminary report of associated factors in [18] H. Coskun, L. Erisen, O. Basut, Factors affecting wound infection rates in head and
wound infection after major head and neck neoplasm operations–does the duration neck surgery, Otolaryngol. Head Neck Surg. 123 (2000) 328–333.