Deep Neck Infections (DNI) are defined as suppurative infectious processes of deep visceral spaces of the neck that usually originates as soft tissue fasciitis and may lead to an abscess.1 Direct extension of an upper aerodigestive infection through fascial planes is the most common cause. DNI are a frequent emergency in Otolaryngology which can be life-threatening as it may lead to airway obstruction, mediastinitis or jugular vein thrombosis.2 The aim of this study is to review different factors that may be the predisposing ones to an increase of infection risk and may have an important role in prognosis.

We performed a retrospective study of patients diagnosed of cervical infection who were admitted in the emergency room of our hospital from January 2005 to December 2015. We excluded patients with superficial skin infections, limited intraoral infections (such us dental phlegmon or abscess) and cervical necrotizing fasciitis. Finally, 330 patients were enrolled in our study.

Although peritonsillar infections are not truly DNI, we decided to include them in our review because of its high incidence and sometimes coexistence with other deep neck space infection.

We used excel and SPSS to perform statistical analysis and Pearson X2 test were calculated to obtain p-values.

In our review pharyngotonsillar infections were the most common cause of peritonsillar and DNI. This result is consistent with some studies in the literature,1–4 although for the majority, odontogenic infections are the main cause, especially in studies carried out in Asia and Eastern Europe.5–7 This may be related to different oral hygiene conditions between different countries. Although peritonsillar infections are not strictly DNI we chose to consider them in our review, as well as other studies did,2,6,8 because in many cases it was the start of a proper DNI or because it had severe complications as a DNI can have. If we quantified just strict DNI we had a population of 91 parapharyngeal and 11 retropharyngeal infections in 10 years. In patients with DNI is more common to find cases who had not had a tonsillectomy, this may be explained as tonsils have an increased bacterial load living within crypts.7

We would like to enhance that we found an increase in DNI incidence in the second period studied (2011–2015); this could be due to an aging population or the fear to over prescribe antibiotics and develop resistant microorganisms. In fact, 3 out of 4 people had not taken any medication prior to the emergency consult.

In this study we found that systemic comorbidities like diabetes mellitus3,4,9,10 or hepatopathy and allergy to penicillins are common in cases of DNI who suffer complications or require ICU stay. DM results in a defect of polymorphonuclear neutrophil function, cellular immunity and complement activation. Consequently, hyperglycemia and high glycosylated hemoglobin are predictors of worse prognosis,10 due to it, our diabetic patients (including those unknown who present with high blood glucose concentration while hospitalization) were studied by the Endocrinology department. The prevalence of penicillin allergy in our review (5.5%) was lower than the global population one (10%–12%),11 however it was much higher in patients who required ICU stay (2 out of 5) or who suffered complications (3 out of 13).

S. viridans was the most common pathogen in our population, as well as in other studies.3,6,12 We did not find Klebsiella pneumoniae in our environment, which differs from studies in Asia.4,8,10 They usually find a high prevalence of this microorganism, specially in diabetic patients.13 We had two ways of obtaining material for culture, either a FNA in the consult or a sample obtained during surgical drainage. Sometimes none of them could be performed. Some patients had a severe trismus which hindered the FNA (it was realized in 227/330 cases). On the other hand, 245 patients (74.24%) received surgical drainage, which is the best moment to take a sample of the infected material, but it was not always possible, as in some cases the material obtained from the infected area was not enough or was not in suitable conditions (purulent material mixed with blood or saliva). Besides, even when the sample was enough cultures were not always positive. This may be explained by antibiotics taken prior to sample extraction or an incorrect sample management (e.g. anaerobes in an aerobic environment).

According to the treatment, we confirm what most studies have already said. Every patient received antibiotics and corticosteroids.1–11,14 Surgical drainage still is the option when medical treatment is not enough, when there is already a well formed hypodense area with margins well defined or an air/fluid level or signs of complications such as mediastinitis or involvement of multiple regions.15

Complications may appear as a consequence of extension of the infection through neck spaces. Mediastinitis and airway obstruction were the most common ones as previous studies have shown before.1–3 In cases of mediastinitis, thoracic surgeons performed the drainage in the same surgical time as our team did a cervicotomy. Traqueostomy was needed in a lower percentage than other studies,3 around 1% like some Indian review.6 The use of corticosteroids decreases tissue edema and the probability of pus gush into the airway while endotracheal intubation, making the procedure safer and more successful.16,17

Even there has been an increase in DNI incidence, mortality remains low as it have been previously shown in other studies.17–19

DNI are still common and can develop serious complications. Immunocompromised patients with systemic comorbidities are susceptible of worse prognosis. In spite of the increase in DNI, mortality has decreased thanks to multidisciplinary attention and improvements in imaging techniques or antibiotics and surgery, which have enabled an earlier diagnosis and treatment.

The authors declare no conflicts of interest.