Page 30 - Revista Portuguesa - SPORL - Vol 61. Nº4
P. 30
Methods assessed with the Kolmogorov-Smirnov
This retrospective observational study included Test. Associations of different variables
adult patients who underwent surgery due to with mediastinitis were analyzed using the
DNI in the otorhinolaryngology department of Chi-square Test or Fischer’s Exact Test for
a tertiary hospital center from January 2015 to categorical variables, Student’s T Test for
December 2021. Exclusion criteria were patients normally distributed continuous variables,
with an isolated peritonsillar abscess and and Mann-Whitney U Test for non-normally
those with a malignant head and neck tumor. distributed continuous variables. Multivariate
All patients were treated according to the logistic regression analysis was performed
department’s protocol for DNI. Every patient by taking into consideration age, etiology,
underwent blood investigations and contrast laboratory parameters (neutrophil count,
enhanced computed tomography (CT) prior lymphocyte count, and NLR), CT parameters
to the surgery. The type of surgery (intra-oral (highest dimensions, infra-hyoid extension,
or external drainage) was selected according and presence of gas), location, and laterality
to the clinical and imaging examination to assess the predictive factors for progression
results, and purulent exudates were collected to mediastinitis. All statistical analyses were
in every case for microbiological assessment. conducted using the software IBM® SPSS®
Antibiotic therapy was initiated empirically Statistics version 27, and associations were
with ceftriaxone (1 g/IV/q.12.h.) and clindamycin considered significant when p<0.05.
(600 mg/IV/q.6.h.), and was modified, if
required, after antimicrobial susceptibility Results
tests. Drains were placed in every patient Epidemiology of DNI
who underwent external drainage and daily A total of 165 patients with DNI were included.
lavages were performed with Microdacyn®. The mean age at presentation was 48.5 ± 16.6
The following data were retrospectively years; there was a male predominance (n=106;
analyzed: patients’ demographic variables 64.2%); and 23 patients (13.9%) had diabetes
(age, sex) and clinical history (DNI etiology, mellitus (DM). The median symptom duration
previous antibiotic therapy, symptom until otorhinolaryngological evaluation was 5
duration, and other diseases), blood tests at (1–30) days, and 105 (63.6%) patients had taken
admission (C-reactive protein [CRP], leucocyte antibiotics previously, among whom 52.4%
count, lymphocyte count, neutrophil count, were undergoing treatment with amoxicillin
and neutrophil to lymphocyte ratio [NLR]), with clavulanic acid. Most patients (n=134;
purulent exudate cultures, and contrast 81.2%) underwent external cervical drainage,
enhanced CT scan (involved spaces, highest while the remaining underwent intra-oral
dimensions, infra-hyoid extension, and drainage (n=31; 18.2%). After surgery, the
presence of gas), treatment, hospitalization median hospitalization days were 10.5 (1–133)
days and complications of DNI. Furthermore, days, and 40 (24.2%) patients required re-
patients who developed mediastinitis were intervention. The reported complications
compared with patients with isolated DNI are shown in Figure 1. The most common
to assess the prognostic factors related to complication was mediastinitis in 20 (12,1%)
mediastinitis. A descriptive analysis of the patients. The most frequent etiology was
patients’ characteristics was performed by odontogenic infection (n=86, 52.1%), while the
using absolute and relative frequencies for second most common was tonsillitis (n=42;
categorical variables, mean and standard 25.5%). The etiologies are shown in Figure 2.
deviation for normally distributed continuous Furthermore, the most frequently involved
variables, and median and range for non- neck space was the submandibular space
normally distributed continuous variables. (n=100; 60.6%) followed by the parapharyngeal
Normality of continuous variables was space (n=69; 41.8%), as illustrated in Figure 3.
368 Revista Portuguesa de Otorrinolaringologia - Cirurgia de Cabeça e Pescoço
