Buscar en
Acta Otorrinolaringológica Española
Toda la web
Inicio Acta Otorrinolaringológica Española Deep Neck Infection: Review of 286 Cases
Journal Information
Vol. 63. Issue 1.
Pages 31-41 (January - February 2012)
Vol. 63. Issue 1.
Pages 31-41 (January - February 2012)
Original article
DOI: 10.1016/j.otoeng.2012.01.006
Full text access
Deep Neck Infection: Review of 286 Cases
Infecciones cervicales profundas. Revisión de 286 casos
Pablo Santos Gorjóna,??
Corresponding author

Corresponding author.
, Pedro Blanco Pérezb, Ana Cristina Morales Martínc, Juan Carlos del Pozo de Diosb, Santiago Estévez Alonsob, María Isabel Calle de la Cabanillasb
a Servicio de ORL, Hospital Nuestra Señora de Sonsoles, Ávila, Spain
b Servicio de ORL y PCF, Hospital Universitario de Salamanca, Salamanca, Spain
c Servicio de Anestesiología y Reanimación, Hospital Nuestra Señora de Sonsoles, Ávila, Spain
Article information
Full Text
Download PDF
Figures (4)
Show moreShow less
Tables (7)
Table 1. Location and Age at Onset.
Table 2. Increase in Frequency of DNI between 1999–2004 and 2004–2009.
Table 3. Concomitant and Enabling Diseases.
Table 4. Origin and Hospital Stay of DNI Cases (No.=286).
Table 5. DNI Complications and Mortality (No.=286).
Table 6. Prior Treatment, Diagnostic Methods and Types of Surgery in DNI (No.=286).
Table 7. Hospital Antibiotic Treatment in the DNI (No.=286).
Show moreShow less

Deep neck infections (DNI) are potentially lethal processes and are susceptible to severe complications. This study shows an increment of the incidence and investigated the cause. More than 30% of cases are idiopathic, but they are commonly related to a dental or oropharyngeal infection.

Material and methods

We present 286 consecutive cases in a retrospective 11-year study. We analysed the clinical picture, antecedents and concurrent diseases, and diagnostic and therapeutic approaches that could be related to developing a DNI.


A major increase in DNI incidence in our setting was seen in the last few years. The major complications were mediastinitis, septic shock with disseminated intravascular coagulation syndrome, necrotising fasciitis and acute respiratory failure. A lower cranial nerve palsy that develops into dysphagia and leukoencephalopathy is the most frequent sequela. We found 10% of aspiration pneumonia. Mortality in adults was 5.9% and in children 6.2%.


Deep neck infections constitute a medical and surgical emergency. Severe complications may arise in a short time. We must be vigilant to alarm symptoms such as dyspnoea, stridor, pain in the floor of the mouth, sialorrhoea, trismus, etc.

Improvements in antibiotic therapy, diagnostic imaging and critical patient support modalities have decreased mortality and there is a better prognosis, with complications being identified and treated earlier.

Deep neck infection
Peritonsillar abscess
Ludwig angina

Las infecciones cervicales profundas (ICP) son afecciones potencialmente letales y susceptibles de complicaciones graves. En este estudio se constata el aumento de la incidencia y se trata de buscar las causas. Más del 30% son idiopáticas, pero suelen relacionarse con infecciones dentarias y orofaríngeas.

Material y métodos

Se presentan 286 ICP consecutivas en un estudio retrospectivo de los últimos 11 años en nuestro centro y se analizan antecedentes y enfermedades concomitantes y aspectos diagnósticos y terapéuticos de los pacientes que pueden influir en el desarrollo de las ICP.


Se objetiva un repunte de la incidencia de las ICP en nuestro medio. Las complicaciones más frecuentes fueron mediastinitis, shock séptico con coagulación intravascular diseminada, fascitis necrotizante y compromiso respiratorio agudo. Las secuelas más frecuentes observadas fueron la disfagia por parálisis de pares craneales bajos y neuroencefalopatía. Se produjeron un 10% de neumonías aspirativas. La mortalidad por ICP en los adultos es del 5,9% y en los niños del 6,2%.


Las ICP debe ser consideradas una urgencia médicoquirúrgica. Son susceptibles de complicaciones graves en cuestión de horas. Debemos estar atentos a la aparición de síntomas de alarma como: disnea, estridor, dolor a la palpación en el suelo de la boca, sialorrea, trismus, etc.

El avance en los diversos procedimientos diagnósticos (imagen, microbiología) y terapéuticos (antibióticos, cirugía), así como la mejor atención del paciente crítico (UCI- REA), han sido decisivos en el diagnóstico y tratamiento precoz, identificar y tratar las complicaciones y mejorar el pronóstico y la mortalidad de los pacientes.

Palabras clave:
Infección cervical profunda
Absceso periamigdalino
Angina de Ludwig
Full Text

Deep neck infections (DNI) are diseases of great importance due to their relative frequency and their serious complications. They include peritonsillar, masseteric, pterygopalatine maxillary, retropharyngeal, parapharyngeal, submandibular, parotid and floor of mouth abscesses (Ludwig's angina).1 These DNI develop in virtual spaces limited by the aponeurosis with low resistance to the spread of infection.2 Infections anywhere in the upper aerodigestive tract (UAT) can spread into these spaces quickly and easily, causing serious complications such as mediastinitis or necrotising fasciitis.3,4

It is sometimes difficult to find the origin of DNI because the primary source of infection may precede it by weeks.4 They are usually polymicrobial infections caused by both aerobic and anaerobic bacteria as well as gram positives.5 Anaerobic bacteria and streptococci are involved in retropharyngeal and peritonsillar abscesses, while staphylococci and mixed flora predominate in submandibular cases.6 It is necessary to investigate risk factors such as infections, foreign bodies, trauma, immunosuppression and addiction to intravenous drugs. Concomitant diseases such as congenital cysts and fistulas, TB, diabetes mellitus, HIV, tumours, deficiency states and so on should also be taken into consideration.7,8 Possible warning signs include dyspnoea, stridor, increased cervical volume, tenderness of the mouth floor upon palpation, sialorrhoea and bulging of the pharyngeal wall.9,10

Plain radiography can identify an odontogenic source of infection or its spread to the thorax (mediastinum or lung parenchyma), with the lateral neck projection being the most useful for this purpose.11 Cervical ultrasound can be used serially and may provide information on the location, size, relationship with neighbouring structures and content of the lesion. The gold standard is CT with intravenous contrast, but MRI is also used on some occasions.12,13

Medical and surgical treatments should be instituted early, taking the location of the DNI and any involvements produced into account to select the most effective therapeutic modality and approach.14,15 Complications in DNI can be very serious and interrelated, as in the case of dyspnoea due to obstruction of the airway, which can be aggravated by the rupture of the abscess during intubation manoeuvres, or trismus, which conditions an impossible orotracheal intubation. Septic shock, jugular vein thrombosis, disseminated intravascular coagulation (DIC), paralysis of cranial nerves and possible rupture of the carotid are all extremely serious complications that justify the importance of this pathology.14

In this study we analysed the clinical aspects of DNI in our environment, especially their incidence and severity, comparing their different locations and differentiating between paediatric and adult DNI.

Material and Methods

We performed a descriptive clinical study over a total of 11 years retrospectively between January 1999 and December 2009. We reviewed the records of 297 patients with DNI who required admission for medical or surgical treatment. We excluded 11 patients whose clinical histories did not contain sufficient data to complete the study. The final sample therefore consisted of 286 patients, 238 adults and 48 children (<14 years).

We analysed the following clinical variables grouped into:

  • (a)

    Background: age, gender, frequency, risk factors (infections, comorbidity) and service or centre of origin.

  • (b)

    Present illness: location, previous treatments, allergies to antibiotics, length of hospital stay, complications, sequelae and mortality.

  • (c)

    Diagnostic procedures: only clinical, microbiology, imaging tests and their types.

  • (d)

    Treatment: medical or surgical, antibiotic therapy (drugs, doses, routes, duration) and surgical technique.

We used the statistical software SPSS 12.0 for Windows to perform the statistical analysis.


The mean age was 37.2 years, with a standard deviation ±17 and range between 22 days and 86 years. As for gender, the ratio in adults was 1.2/1 for males (129 males and 109 females), being slightly higher in children (1.4/1; 28 males and 20 females).

Out of the 286 patients in the sample, there were 238 adults (83.2%) and 48 children (16.8%). The location of the DNI was variable according to age (Table 1).

Table 1.

Location and Age at Onset.

Location  Adults No. (%)  Children No. (%)  Total No. (%) 
Peritonsillar  137 (57.5)  8 (6.0)  145 (50.6) 
Submandibular  35 (57.6)  23 (47.9)  58 (20.3) 
Parotid  23 (8.0)  0 (0)  23 (8.0) 
Parapharyngeal  11 (4.6)  6 (8.0)  17 (5.9) 
Retropharyngeal  7 (2.9)  9 (18.8)  16 (5.6) 
Masseteric  11 (4.6)  0 (0)  11 (4.6) 
Pterygomaxillary  9 (3.8)  0 (0)  9 (3.8) 
Ludwig's angina  5 (2.1)  2 (4.2)  7 (2.4) 
Total  238 (83.2)  48 (16.8)  286 (100) 

If we establish 2 consecutive periods of 5.5 years, we note that there was a rise in frequency in all locations in the second year, with 53.8% of cases (129 in total) in adults and 29 children of the total, or 58.3% of paediatric cases (Table 2). Among the analysed risk factors were previous infectious processes (60%) and diabetes mellitus (24.8%). However, a large group of patients (38.5%) presented no comorbidity, previous infections or trauma. In adults, comorbidity due to diabetes mellitus was associated to 4/7 (57.14%) retropharyngeal abscesses and 30/137 (22%), peritonsillar abscesses, while no paediatric abscess associated this disease. In total, 3 of the 9 paediatric retropharyngeal abscesses (33%) presented a history of trauma/foreign body, but only 2 of the 16 adult retropharyngeal abscesses had this same background. In peritonsillar abscesses, it was possible to establish a relation with pharyngotonsillar infections in 91/137 adults (66%), with this figure being somewhat lower in children (4/8 cases; 50%). Odontogenic infections were associated with submandibular abscesses in 21/35 adults who presented them (60%) and in 11/23 children (48%) (Table 3). Finally, among the risk factors, we should note that 18% of patients (52) in our series had undergone adenotonsillectomy.

Table 2.

Increase in Frequency of DNI between 1999–2004 and 2004–2009.

Location  Period 1999–2004aPeriod 2004b–2009Total
  Adults  Children  Adults  Children  Adults  Children 
  No. (%)  No. (%)  No. (%)  No. (%)  No. (%)  No. (%) 
Peritonsillar  65 (47.4)  3 (37.5)  72 (52.2)  5 (62.5)  137 (57.5)  8 (6.0) 
Submandibular  15 (52.9)  10 (43.5)  20 (57.1)  13 (56.3)  35 (57.6)  23 (47.9) 
Parotid  11 (47.8)  0 (0.0)  12 (52.2)  0 (0.0)  23 (8.0)  0 (0) 
Parapharyngeal  4 (26.4)  2 (33.3)  7 (63.6)  4 (66.7)  11 (4.6)  6 (8.0) 
Retropharyngeal  3 (42.9)  3 (37.5)  4 (57.1)  5 (62.5)  7 (2.9)  8 (18.8) 
Masseteric  5 (45.5)  0 (0.0)  6 (54.5)  0 (0.0)  11 (4.6)  0 (0) 
Pterygomaxillary  4 (44.5)  0 (0.0)  5 (55.5)  0 (0.0)  9 (3.8)  0 (0) 
Ludwig's angina  2 (40.0)  1 (50.0)  3 (60.0)  1 (50.0)  5 (2.1)  2 (4.2) 
Total  110 (46.2)  19 (41.7)  129 (53.8)  28 (58.3)  238 (83.2)  48 (16.8) 

First semester.


Second semester.

Table 3.

Concomitant and Enabling Diseases.

Pathologies  Adults No. (%)  Children No. (%)  Total No. (%) 
Pharyngotonsillitis  112 (47.0)  12 (26.0)  124 (63.0) 
Odontogenic  69 (29.1)  4 (7.2)  73 (36.3) 
Others  24 (10.0)  5 (10.0)  29 (20.0) 
Total UAT infections  205 (86.1)  21 (43.7)  172 (60.0) 
Diabetes mellitus  7 (2.4)  0 (0)  7 (2.4) 
AH  63 (22.0)  0 (0)  63 (22.0) 
Dyslipidemia  84 (29.4)  0 (0)  84 (29.4) 
Malnutrition  6 (8.6)  1 (0.8)  7 (2.4) 
Total metabolic diseases  160 (67.2)  1 (0.8)  161 (68.0) 
Haematological  2 (0.8)  2 (4.2)  4 (0.8) 
HIV  2 (0.8)  0 (0)  2 (0.4) 
Transplant  1 (0.4)  0 (0)  1 (0.2) 
Oncological  6 (2.5)  1 (2.1)  7 (2.4) 
Total immunosuppression  11 (4.6)  3 (6.25)  14 (4.9) 
Lesions/trauma  4 (1.2)  4 (8.3)  7 (2.4) 
Cervical surgery  16 (8.6)  0 (0)  16 (5.8) 
Foreign bodies/choking  3 (1.0)  0 (0)  3 (1.2) 
Total UAT trauma  25 (15.1)  4 (8.3)  29 (10.0) 
Cysts/fistulas  4 (1.2)  1 (2.1)  5 (1.7) 
Tuberculosis  5 (2.1)  1 (2.1)  6 (2.1) 
Tumours  6 (2.5)  0 (0.0)  6 (2.1) 
Total local causes  15 (6.3)  2 (4.2)  17 (5.9) 
Absence of comorbidity  84 (29.4)  21 (43.7)  75 (38.5%) 

The origin of patients mainly corresponded to the Emergency Service of the hospital (245 cases; 89.2%) (Table 4). The location of the abscess also conditioned the referral of patients from other centres. Parapharyngeal and retropharyngeal abscesses and Ludwig's angina were the most commonly derived from other centres, while this percentage was very low in peritonsillar abscesses (6 cases). More than half of patients with descending necrotising mediastinitis were referred to our centre for joint assessment by ENT and thoracic surgery (9 cases; 56.2%) (Table 5).

Table 4.

Origin and Hospital Stay of DNI Cases (No.=286).

Location  Peritonsillar No. (%)  Submandibular No. (%)  Parotid No. (%)  Parapharyngeal No. (%)  Retropharyngeal No. (%)  Masseteric No. (%)  Pterygomaxillary No. (%)  Ludwig's Angina No. (%) 
Emergency Service  132 (46.1)  49 (17.2)  20 (6.9)  10 (3.5)  11 (3.9)  9 (3.4)  5 (1.6)  4 (1.4) 
Other centres  8 (2.8)  8 (2.8)  3 (1.3)  7 (2.5)  5 (1.7)  0 (0.0)  2 (0.6)  3 (1.3) 
Other services  5 (1.7)  1 (0.4)  0 (0)  0 (0)  0 (0)  2 (0.6)  2 (0.6)  0 (0.0) 
Stay, days  2.67±1  6.04±2  8.22±4  24.24±15  15.02±6  3.89±2  6.95±2  9.15±4 
Table 5.

DNI Complications and Mortality (No.=286).

Location  Peritonsillar No. (%)  Submandibular No. (%)  Parotid No. (%)  Parapharyngeal No. (%)  Retropharyngeal No. (%)  Masseteric No. (%)  Pterygomaxillary No. (%)  Ludwig's Angina No. (%) 
Sepsis and DIC  3 (1.0)  3 (1.0)  7 (2.4)  20 (7.0)  13 (4.5)  2 (0.7)  1 (0.3)  6 (2.1) 
Mediastinitis  2 (0.7)  1 (0.3)  1 (0.3)  6 (2.1)  5 (1.7)  0 (0)  0 (0)  2 (0.7) 
Jugular vein thrombosis  0 (0)  0 (0)  0 (0)  2 (0.7)  1 (0.3)  0 (0)  0 (0)  0 (0) 
Necrotising fasciitis  0 (0)  0 (0)  0 (0)  3 (1.0)  0 (0)  0 (0)  0 (0)  0 (0) 
Pneumonia  0 (0)  6 (2.1)  9 (3.2)  8 (2.8)  5 (1.7)  0 (0)  2 (0.7)  2 (0.7) 
Cranial nerve paralysis  11 (3.8)  7 (2.4)  0 (0)  0 (0)  2 (0.7) 
Othersa  1 (0.3)  2 (0.7)  1 (0.3)  3 (1.0)  2 (0.7)  0 (0)  0 (0)  3 (1.0) 
Tracheotomy  0 (0)  4 (2)  3 (1.0)  11 (3.9)  8 (2.8)  0 (0)  0 (0)  6 (2.0) 
Deaths  0 (0)  0 (0)  2 (0.7)  7 (2.4)  4 (1.4)  1 (0.3)  0 (0)  3 (1.0) 

Neurological deficits, severe anaemia, reactions to drugs.

Current Disease

The location of DNI is highly variable, as can be seen in Table 1. The most common site in adults is peritonsillar (57.5%), followed by submandibular (14.7%); in children, it is submandibular (48%), followed by retropharyngeal (18.7%). There were no DNI in the parotid, pterygomaxillary and masseteric regions in the paediatric age group (Table 1).

Previous treatments received by patients consisted of antibiotics in 59.4% (170 cases), although the type of antibiotic, dosage and administration time of these treatments can only be considered correct in 63.5% (Table 6). A total of 28 patients were allergic to penicillin and its derivatives (9.8%).

Table 6.

Prior Treatment, Diagnostic Methods and Types of Surgery in DNI (No.=286).

Location  Peritonsillar No. (%)  Submandibular No. (%)  Parotid No. (%)  Parapharyngeal No. (%)  Retropharyngeal No. (%)  Masseteric No. (%)  Pterygomaxillary No. (%)  Ludwig's Angina No. (%) 
Prior treatment  95 (33.2)  30 (10.4)  8 (2.8)  13 (4.5)  11 (3.9)  5 (1.7)  6 (2.1)  5 (1.7) 
Correct compliancea  60 (20.9)  21 (7.3)  5 (1.7)  8 (2.8)  6 (2.1)  3 (1.0)  4 (1.1)  4 (1.5) 
Beta-lactam allergy (total 28)  15 (5.2)  2 (0.7)  5 (1.9)  1 (0.3)  3 (1.0)  0 (0)  0 (0)  1 (0.3) 
CT  1 (0.3)  32 (10.3)  6 (2.1)  16 (5.6)  0 (0)  8 (2.8)  9 (3.2)  7 (2.4) 
Ultrasound  3 (1.0)  21 (7.3)  9 (3.0)  12 (3.4)  13 (3.7)  3 (1.0)  1 (0.3)  2 (0.7) 
MRI  0 (0)  0 (0)  3 (1.0)  3 (1.0)  2 (0.7)  0 (0)  0 (0)  0 (0) 
No imaging tests  140 (48.9)  21 (7.3)  7 (2.3)  1 (0.3)  0 (0)  0 (0)  0 (0)  0 (0) 
Drainage with local anaesthesia  125 (43.7)  5 (1.9)  19 (6.5)  0 (0)  1 (0.3)  0 (0)  0 (0)  0 (0) 
Drainage with sedation  18 (6.3)  15 (5.2)  2 (0.7)  0 (0)  1 (0.3)  1 (0.3)  0 (0)  0 (0) 
Drainage with general anaesthesia  2 (0.7)  38 (13.3)  2 (0.7)  1 (4.2)  14 (3.9)  1 (0.3)  9 (3.2)  7 (2.3) 

In type antibiotic, time and dose.

The time of admission was highly variable, depending on the severity of the DNI. It was short (2.7±1 days) in peritonsillar abscesses, while it increased for the retropharyngeal (15±4 days), being even higher in the parapharyngeal (24.2±8 days), largely due to the more severe complications they cause. The stay also increased in relation to the associated condition. Diabetic patients were discharged with a mean delay of 4 days and DNI severity was also conditioned (6 deaths in diabetics). In immunocompromised patients, discharge was delayed 5 days on average due to comorbidity and multidisciplinary treatment received, with 7 deaths in this group. The remaining comorbidities (hypertension, dyslipidemia, etc.) did not lengthen the stay or act as indicators of increased mortality (Table 4).

Patients with severe DNI required hospitalisation in ICU and prolonged intubation. Some of them required a temporary tracheotomy concomitant with drainage surgery (9.8%), while others, after remaining with sedatives and intubated for more than 14 days, underwent deferred tracheotomy (2.1%). As can be seen in Table 5, the more severe infections (parapharyngeal and retropharyngeal abscesses) required the highest number of tracheotomies. Severe cases also had the highest rate of complications: sepsis (9.4%), DIC (9.8%), descending necrotising mediastinitis (5.9%), jugular vein thrombosis (1.1%) and severe necrotising fasciitis (1%). The most commonly reported sequelae in severe patients were dysphagia from cranial nerve paralysis (9.8%) and neuro-encephalopathy (4.2%). Both are risk factors for the development of aspiration pneumonia, which occurred in 10.5% of cases. The remaining complications are presented in Table 5.

In adults, DNI mortality was 5.9% and, in children, 6.2%. The DNI with the highest mortality were Ludwig's angina (43%) and parapharyngeal abscesses (41%). In children, both locations had a 50% mortality (1 death due to Ludwig's angina and 3 due to parapharyngeal abscesses). In adults the mortality was lower, 20% for Ludwig's angina (1 case) and 36.4% for parapharyngeal abscess (4 cases). It is noteworthy that all paediatric patients who died were syndromic cases or with base conditions, with evolution being fulminant (48h) and fatal despite appropriate medical and surgical treatment. Fulminant multiple organ failure and sepsis occurred in all cases of paediatric death, while in adults the cause of death was generally mediastinitis, necrotising fasciitis, sepsis and respiratory infection (bronchopneumonia) of aspiration type (Table 5).

Diagnostic Procedures

A culture was performed in 186 patients (65.0%). In 131 cases (70.4%), the result was polymicrobial flora. There were 18.8% negative (35) and contaminated cultures (20 of the total; 1.1%).

Imaging techniques were performed according to the location of the DNI. No imaging studies were conducted in 140 patients with peritonsillar abscess (130 adults and 10 children). In contrast, all patients diagnosed with Ludwig's angina required a CT scan. Plain radiography was useful in certain cases (Fig. 1). The most commonly used technique, even in paediatric ages, was CT (88 patients; 30%) and the second was ultrasound (55 patients; 19.2%), which was performed in children serially in the most severe locations (Figs. 2 and 3). MRI was rarely used (8 patients; 0.3%) (Table 6).

Figure 1.

Lateral radiograph of the neck showing an increase of soft tissues at the retropharyngeal level and a cervical crushing of C5. The patient suffered a tuberculous discitis complicated by a retropharyngeal abscess.

Figure 2.

Chest CT (axial section) showing a density increase in superior mediastinal fat planes. In addition to the cervical approach, this patient required mediastinal debridement by thoracotomy, as well as placement of drainage for subsequent washes.

Figure 3.

Cervical CT (axial and sagittal reconstruction) showing a right retropharyngeal abscess measuring 2.9cm×3.5cm×1.9cm (approximate volume: 10–12cc), located on the right side and reaching the midline, which also compressed and displaced the airway at the level of the oropharynx. Multiple bilateral laterocervical adenopathies, larger and more numerous, can be observed on the right side.


Preference for one technique or another was determined according to location. Parotid abscess was more often diagnosed with ultrasound and only in some cases was this supplemented with CT (26%). Submandibular abscesses were diagnosed by ultrasound alone in 33% and were completed by CT in 55% (Table 6).


Once patients were admitted, the most widely used antibiotic was amoxicillin–clavulanate (190 patients; 66.4%) in monotherapy or with metronidazole (68 cases; 23.7%). Next were third-generation cephalosporins (54 cases; 19.1%) and piperacillin (55 patients; 19.2%), in monotherapy or in combination with tazobactam (4.2%), especially in critically ill patients. Clarithromycin was used in outpatient treatment for the patients who were allergic to penicillin, which was replaced most often by quinolones (ciprofloxacin/levofloxacin).

The need for intravenous therapy was determined by the evolution of the disease. For peritonsillar abscesses, it was maintained throughout the hospital stay, followed by outpatient oral treatment until completing 10 days. In the remaining DNI, intravenous treatment was maintained until the resolution of symptoms (clinical improvement, apyrexia, analytical normality, negative blood cultures, etc.).

When the symptoms did not show good evolution and did not improve despite antibiotic treatment, we proceeded to use surgery and place a drainage, which helped to evacuate purulent material and made it possible for us to clean the wound periodically with antiseptic solution and hydrogen peroxide (Fig. 4). In paediatric DNI, we performed serial ultrasound scans every 24–48h and, if necessary, a control CT scan prior to surgical drainage of the abscess when it was greater than 1.5–2cm in diameter.

Figure 4.

Surgical steps in the drainage of a neck abscess. Firstly, suppuration and necrosis areas become evident after an extensive cervicotomy (A). Secondly, cellulo-aponeurotic spaces are opened and exposed (B). Subsequently, necrotic tissue is excised (C). Finally, anatomical compartments are washed and drainage is placed for subsequent washes (D).


Logically, the anaesthesia used for DNI drainage depended on the infection location and type of patient. In peritonsillar abscesses, the procedure was most often performed under topical and/or local anaesthesia for adults. However, 18 patients required sedation and 2 of them even required general anaesthesia. These patients were noncompliant or suffered a concomitant illness (mental retardation, dementia or psychosis). This situation was reversed in the paediatric ages. Except for children older than 10 years who collaborated in the implementation of drainage (43%), they required sedation for surgery in 22% of cases and general anaesthesia in 35%. In the latter group, 2 tonsillectomies had to be carried out in the same surgical procedure (Table 6).

As evidenced in Table 6, the remaining DNI studied required general anaesthesia in most cases.


The reported incidence of DNI is around 9–12/100000 inhabitants/year.1 Our sample, adjusting roughly to the total population, would be around 15/100000 inhabitants/year. It must be kept in mind that our centre receives referrals from other hospitals, especially for the more severe DNI cases.

In the study we differentiated adult DNI (83.2%) from paediatric cases (16.8%) (Table 1) because we were unable to find publications comparing the frequencies of these pathologies between these age segments.

Anglo-Saxon countries have witnessed an increase in DNI in recent years, especially in children under 5 years, with an incidence around 2/100000 inhabitants/year.3,4 This fact was also reflected in our study. The present series was collected over the past 11 years. When comparing the first period of time with the second, we can observe a peak in the frequency of these infections: 60% of the adult DNI and 68% of the paediatric cases were diagnosed in the second period (Table 2). This may be due to a bias in the selection of patients, given that our hospital is a referral centre for severe DNI and paediatric cases because it has a paediatric ICU service and a thoracic surgery department.

Our series also showed an increase in parapharyngeal abscesses, because 70% of them came from other centres. In addition, 56% of DNI complicated by mediastinitis were referred from other hospitals.

With respect to peritonsillar abscesses, the figures considered in the aforementioned studies put their incidence at around 5/100000 inhabitants/year,3,4 while in our series it was over 50% of all DNI, with an even higher extrapolated incidence, close to 7 patients/100000 inhabitants/year. Therefore, we should seek other causes for this increase, such as self-medication, poor therapeutic compliance, inappropriate antibiotics, an increased number of immunocompromised patients or patients with comorbidities (diabetics, transplanted patients, haematology, oncology, HIV), increase in neck surgeries (ENT, general and maxillofacial surgery, traumatology), along with a possible increase in young patients allergic to penicillin. Nevertheless, the idiopathic cases of DNI represented 38.5% in our series (Table 3). Other facts to consider in this increase of DNI are the smaller number of tonsillectomies performed in recent years and the fact that, in general, odontostomatological problems are still not a high priority during oral check-ups, despite affecting 60% of the population considered “healthy”.16

In our DNI series, 18.1% of the patients had undergone adenotonsillectomy. We noted that the more severe DNI cases took place among diabetic and immunocompromised patients (Table 3). In fact, most of the deceased suffered associated comorbidities (88.2%; 15 patients). Among the adult patients studied, we found oropharyngeal and odontogenic infectious backgrounds in 47.0% and 29.1% of cases, respectively. However, less than 1% had a history of choking with foreign bodies, prior surgery or UAT trauma. These data differ from those obtained from the paediatric population, whose histories contained, proportionately, more cases of oropharyngeal (26.1%) than odontogenic (7.2%) infection. In children there were no cases of choking on foreign bodies, prior surgery or trauma. Up to 10% of patients in our series, both adults and children, had a history of infection of the salivary glands, paranasal sinuses and ears. Less than 1% of adults acknowledged being addicted to drugs and having recently used needles on the major neck vessels (Table 3).

Once all these adjustments concerning possible causes of DNI have been made to our series, we can consider that nearly 22% of DNI in adults and 58% in the paediatric population are idiopathic, with no clear trigger. Other authors put idiopathic causes of adult DNI at between 18% and 20%, whereas they place it at 35%–40% in the paediatric ages. Consequently, it appears that our data are consistent with these values.1,5

The reasons why we obtained such high figures of idiopathic aetiology, especially in the paediatric age group, could be that it is difficult in children to locate the initial pain symptoms or infection in the UAT. There were patients in a severe condition referred from other centres with advanced disease state in whom it was not possible to determine the original source. In children, the symptoms may be latent until some characteristic signs appear, such as stiff neck, torticollis, trismus or even dyspnoea due to airway collapse.

There are clinical characteristics that can serve as a guide to identify DNI. This is the case of retropharyngeal space infections, which often occur between 2 and 5 years of age and have torticollis as a characteristic sign.15,17 In our series, the age was in that interval in the 9 children with that location. Of the 7 adults suffering retropharyngeal abscesses, only 1 had a history of foreign body (spine), while 2 had undergone cervical surgery (cyst and vertebral fixation). In the remaining 4 cases, there was no history of invasive techniques (esophagoscopy, endotracheal intubation, trauma or adjacent infections), considered typical by other authors.18

Among adult parapharyngeal abscesses (11 cases), the majority had previous associated oropharyngeal infections, extending from the molars or the pterygomandibular space, as has also been pointed out by other authors.18,19 However, there was no evidence of this background in 5 of the 6 paediatric cases, suggesting some other type of infection, such as the progression of adenitis or perhaps that some cases had been incorrectly identified as parapharyngeal. In fact, 1 case presented trismus, an uncharacteristic sign in this location, although it can occur in advanced stages if the abscess spreads into the pterygomandibular region.20

Pterygomandibular space abscesses presented trismus and dysphagia, with previous infection in the parapharyngeal and retromaxillary regions being confirmed in all cases. Infections in this space were described after injections to block the mandibular nerve or from pericoronitis of the third inferior molars.1,19

All masseteric abscesses in our series presented infections of the oral space or pericoronitis. The main characteristic was inflammation of the masseter muscle and tumefaction of the mandibular angle and ramus with intense trismus, logically.20

Ludwig's angina is a cellulitis that spreads quickly to the cervical region and frequently appears with airway obstruction. In all, 6 of the 7 patients in our series required a tracheostomy to ensure airway patency for surgical intervention (Table 5).

The germs involved in DNI were similar in our study to those published in other series.2,4,5 In this microbiological diagnosis, it is necessary to emphasise some peculiarities. For peritonsillar abscess, it is not common to obtain cultures. Ludwig's angina or masseteric abscesses are usually produced by anaerobic bacteria, streptococci and fusospirochetes.21 However, there is a predominance of polymicrobial growths in DNI cultures, but the number of negative or contaminated cultures is not negligible, as has been published.4,5 This last fact may be due to mistakes during sample collection, as occurs with anaerobes when the sample is taken using a syringe containing air or if the culture is not made immediately after being taken.

With respect to imaging studies, laterocervical radiography is generally carried out unless there is severe respiratory involvement. This requires the patient to remain in cervical extension and inspiration, to avoid false increases of the retropharyngeal space.11 In children, the retropharyngeal space is considered as pathological when it is over 7mm in C2 or 14mm in C6 or there is gas and fluid-gas levels.

Cervical ultrasonography provides useful information about the location, size, relationship with neighbouring structures and contents of the lesion. It helps to locate the abscess, even during surgery, and carry out its puncture. This technique can be carried out serially in children to avoid extensive exposure to radiation.

The most interesting technique, due to the information it provides, is CT with contrast of the cervical and thoracic regions, with a sensitivity of 64%–100% and a specificity of 45%–82%.2,9 CT distinguishes between abscesses, boils or cellulitis and better defines the characteristics of the lesions, although it is sometimes advisable to perform it with sedation in children and uncooperative patients. It can detect the presence of gas (anaerobic germs) and is useful in necrotising fasciitis.12

Although rarely performed, an MRI can be used to obtain 3-dimensional images, perform reconstructions and delineate soft tissue involvement.

Regarding treatment, we can say, based on our experience that the abscesses rarely resolve spontaneously if there is a collection of over 3 cc. The latest reviews recommend an initial antibiotic treatment for 48h for DNI under 2cm without respiratory involvement and in phlegmonous stage, although these are non-randomised, observational studies.11,13 Patients should be closely monitored during this period to detect complications, such as dyspnoea if the lesion increases or pneumonia from pus aspiration if it breaks.

Medical treatment as discussed is usually intravenous for 2–3 weeks. As in other studies consulted,22,23 any of the antibiotics listed in Table 7 can be used empirically until an antibiogram is obtained. These authors use essentially the same therapeutic arsenal as in our field.

Table 7.

Hospital Antibiotic Treatment in the DNI (No.=286).

Location  Peritonsillar No. (%)  Submandibular No. (%)  Parotid No. (%)  Parapharyngeal No. (%)  Retropharyngeal No. (%)  Masseteric No. (%)  Pterygomaxillary No. (%)  Ludwig's Angina No. (%) 
Amoxicillin and clavulanate1  60 (21.0)  32 (11.2)  20 (7.0)  4 (2.0)  6 (2.1)  7 (2.4)  4 (2.0)  2 (0.7) 
Amoxicillin and clavulanate+metronidazole1  20 (7.0)  22 (7.7)  5 (1.7)  1 (0.3)  3 (1.0)  2 (0.7)  1 (0.3)  1 (0.3) 
Cephalosporins1  6 (2.1)  1 (0.3)  2 (0.7)  6 (2.1)  2 (0.7)  0 (0)  0 (0)  1 (0.3) 
Piperacillin–tazobactam1  0 (0)  2 (0.7)  0 (0)  2 (0.7)  3 (1.0)  1 (0.3)  1 (0.3)  3 (1.0) 
Quinolones1  20 (7.0)  1 (0.3)  1 (0.3)  0 (0)  2 (0.7)  0 (0)  0 (0)  0 (0) 

Dose: amoxicillin–clavulanate (2g/8h); penicillin G sodium (2 million IU/4h); ertapenem (1g/24h); metronidazole (500mg/8h); ceftriaxone (100mg/kg). The alternatives are clindamycin (600mg/8h)+moxifloxacin (400mg/12h) or levofloxacin (500mg/12h). Imipenem (1g/8h)+amikacin (15mg/kg/day) are recommended in immunocompromised patients.

Surgical drainage by transoral approach may be indicated in cases without improvement, provided that the abscess is peritonsillar or retropharyngeal (the latter only when it is accessible, usually in very young children). An alternative approach is external lateral cervicotomy, as proposed by various authors.22 We consider that a transoral approach can be attempted in cases of retropharyngeal abscesses in children, provided they are medial to the larger vessels and located in the inflammatory ring of the retropharyngeal lymph node, without free pus in the cervical fascial spaces. However, our view in adult subjects is to use a cervical entry approach always. As mentioned previously, in all other locations, we prefer the external cervical approach.

Another disagreement with other authors is the need to carry out tonsillectomy in the same surgical act as drainage.21 We performed only 2 hot tonsillectomies associated with transoral drainage of tonsillar abscesses, but always following the recommendations in the consensus document of the Spanish ENT and Paediatrics Societies. This defines that the only absolute criterion for tonsillectomy in DNI is a second peritonsillar abscess, along with the rest of criteria established by number of infections/year.24

A technical comment that we deem important in external surgical approaches or in those with extension to thorax and mediastinum is the need to allow good drainage in the surgical wound, even approaching only its edges, without closing it completely. This serves to make daily debridement and postoperative washings through this drainage possible.

Finally, we must note that, despite the increase in DNI at our centre, we have not observed a rise in mortality, as indicated by most authors consulted, who present figures below 10%.1,5,8 In our series, mortality was 4.9% in adults and 6.2% in children. Adult DNI patients who died tended to suffer multiple associated conditions and a more advanced age (75% of deaths occurred in patients >65). All deceased children were syndromic patients with other base conditions. None of the children presented jugular vein thrombosis; only 10% suffered a septic state, 5% DIC and 2% acute respiratory distress syndrome requiring ICU admission.


Deep neck infections are serious conditions that should be considered as medical and surgical emergencies. They are susceptible to serious complications within hours (mediastinitis, septic shock with DIC, necrotising fasciitis and acute respiratory involvement). They are associated with oropharyngeal and odontogenic infections, although 34% are idiopathic. Special attention should be paid to diabetic patients who have undergone any intervention in the head and neck area, as they are susceptible to more severe infections and a higher number of complications. We must also pay attention to the occurrence of alert symptoms including dyspnoea, stridor, pain on palpating the floor of the mouth, sialorrhoea, trismus, etc.

The most frequent DNI is peritonsillar abscess, followed by submandibular, parotid, parapharyngeal, retropharyngeal, masseteric and pterygomaxillary abscesses and, finally, Ludwig's angina. In recent years we have witnessed a rise in their incidence but mortality (5%–6%) has decreased thanks to a comprehensive patient approach with multidisciplinary teams. Progress made in the various diagnostic procedures (imaging, microbiology) and treatment (antibiotics, surgery), and an improvement in the care of critically ill patients (ICU) have been instrumental in early diagnosis and treatment, identification and treatment of complications and in the improvement of prognosis for patients.

Conflict of Interests

The authors have no conflict of interest to declare.


The authors wish to thank the Documentation Service for the valuable assistance provided for this work.

F. Vieira, S.M. Allen, R.M. Stocks, J.W. Thompson.
Deep neck infection.
Otolaryngol Clin North Am, 41 (2008), pp. 459-483
F.W. Craig, J.E. Schunk.
Retropharyngeal abscess in children: clinical presentation, utility of imaging, and recurrent management.
Pediatrics, 111 (2003), pp. 1394-1398
C.M. Philpott, D. Selvadurai, A.R. Barnejee.
Paediatric retropharyngeal abscess.
J Laryngol Otol, 118 (2004), pp. 919-926
N.M. Abdel-Haq, A. Harahsheh, B.I. Asmar.
Retropharyngeal abscess in children: the emerging role of group A beta haemolytic streptococcus.
E. Martín Campagne, F. Del Castillo Martín, M.M. Martínez López, C. Borque de Andrés, M.I. De José Gómez, M.J. García de Miguel, et al.
Abscesos periamigdalinos y retrofarígeos: estudio de 13 años.
An Pediatr, 65 (2006), pp. 32-36
J.C. Inman, M. Rowe, M. Ghostine, T. Fleck.
Pediatric neck abscesses, changing organisms and empiric therapies.
Laryngoscope, 118 (2008), pp. 2111-2114
L.C. Dawes, R. Bova, P. Carter.
Retropharyngeal abscess in children.
J Surg, 72 (2002), pp. 417-420
J. Hasegawa, H. Hidaka, M. Tateda, T. Kudo, S. Sagai, M. Miyazaki, et al.
An analysis of clinical risk factors of deep neck infection.
Auris Nasus Larynx, 10 (2010), pp. 356-361
M. Sakaguchi, S. Sato, S. Asawa, K. Taguchi.
Computerized tomographic findings in peritonsillar abscess and cellulitis.
J Laryngol Otol, 113 (1999), pp. 229-232
N.C. Page, E.M. Bauer, J.E. Lieu.
Clinical features and treatment of retropharyngeal abscess in children.
Otolaryngol Head Neck Surg, 138 (2008), pp. 300-306
R.H. Haug, R.T. Wible, J. Lieberman.
Measurement standards for the prevertebral region in the lateral soft-tissue radiograph of the neck.
J Oral Maxillofac Surg, 49 (1991), pp. 1149-1151
H.C. Chao, C.H. Chiu, S.J. Lin, T.Y. Lin.
Colour Doppler ultrasonography of retropharyngeal abscess.
J Otolaryngol, 28 (1999), pp. 138-141
T.M. Osborn, L.A. Assael, R.B. Bell.
Deep space neck infection: principles of surgical management.
Oral Maxillofac Surg Clin North Am, 20 (2008), pp. 353-365
H. Daya, S. Lo, B.C. Papsin, A. Zachariasova, H. Murray, J. Pirie, et al.
Retropharyngeal and parapharyngeal infections in children: the Toronto experience.
Int J Pediatr Otorhinolaryngol, 69 (2005), pp. 81-86
L. Chang, H. Chi, N.C. Chiu, F.Y. Huang, K.S. Lee.
Deep neck infections in different age groups of children.
J Microbiol Immunol Infect, 43 (2010), pp. 47-52
T. Treasure, W. Hughes, J. Bennett.
Cervical necrotizing fasciitis originating with a periapical infection.
J Am Dent Assoc, 141 (2010), pp. 861-866
S.E. Shefelbine, A.A. Mancuso, B.J. Gajewski, H. Ojiri, S. Stringer, J.D. Sedwick.
Pediatric retropharyngeal lymphadenitis: differentiation from retropharyngeal abscess and treatment implications.
Otolaryngol Head Neck Surg, 136 (2007), pp. 182-188
B. Al-Sabah, H. Bin Salleen, A. Hagr, J. Shoi-Rosen, J.J. Manoukian, T.L. Tewfik.
Retropharyngeal abscess in children: 10-year study.
J Otolaryngol, 33 (2004), pp. 352-355
C. Page, A. Biet, R. Zaatar, V. Strunski.
Parapharyngeal abscess: diagnosis and treatment.
Eur Arch Otorhinolaryngol, 265 (2008), pp. 681-686
D.G. Balatsouras, G.M. Kloutsos, D. Protopapas, S. Korres, C. Economou.
Submasseteric abscess.
J Laryngol Otol, 115 (2001), pp. 68-70
J. Heavey, N. Gupta.
Images in clinical medicine, Ludwig's angina.
N Engl J Med, 359 (2008), pp. 1501
M.B. Tom, D.H. Rice, M.D. Roberson.
Surgical management of retropharyngeal space infection in children.
Laryngoscope, 111 (2001), pp. 1413-1422
J. Peter, C. Green, M. Hoelscher, P. Mwaba, A. Zumla, K. Dheda.
The diagnosis of tuberculosis: current approaches, clinical applicability, and new developments.
Curr Opin Pulm Med, 16 (2010), pp. 262-270
M. Moya, T. Sacristán, A. Blanco, J. Cervera, L.M. Gil Carcedo.
Indicaciones de amigdalectomía y adenoidectomía en el niño y adolescentes.
An Esp Pediatr, 47 (1997), pp. 12-13

Please cite this article as: Santos Gorjón P, et al. Infecciones cervicales profundas. Revisión de 286 casos. Acta Otorrinolaringol Esp. 2012;63:31–41.

Copyright © 2011. Elsevier España, S.L.. All rights reserved
Article options
es en pt

¿Es usted profesional sanitario apto para prescribir o dispensar medicamentos?

Are you a health professional able to prescribe or dispense drugs?

Você é um profissional de saúde habilitado a prescrever ou dispensar medicamentos

es en pt
Política de cookies Cookies policy Política de cookies
Utilizamos cookies propias y de terceros para mejorar nuestros servicios y mostrarle publicidad relacionada con sus preferencias mediante el análisis de sus hábitos de navegación. Si continua navegando, consideramos que acepta su uso. Puede cambiar la configuración u obtener más información aquí. To improve our services and products, we use "cookies" (own or third parties authorized) to show advertising related to client preferences through the analyses of navigation customer behavior. Continuing navigation will be considered as acceptance of this use. You can change the settings or obtain more information by clicking here. Utilizamos cookies próprios e de terceiros para melhorar nossos serviços e mostrar publicidade relacionada às suas preferências, analisando seus hábitos de navegação. Se continuar a navegar, consideramos que aceita o seu uso. Você pode alterar a configuração ou obter mais informações aqui.