Factors Creating a Need for Repeated Drainage of Deep Neck Infections

Deep neck infection (DNI) is associated with morbidity and mortality. Surgical incision and drainage (I&D) of DNI abscesses are essential. Refractory abscesses require repeat I&D. Few studies have assessed the risk factors associated with repeat I&D; here, we investigated such factors. In total, 605 patients with DNI were enrolled between July 2016 and February 2022. Of these patients, 107 underwent repeat I&D. Clinical variables were assessed. On univariate analysis, a high blood sugar level (odds ratio (OR) = 1.006, p < 0.001), the involvement of at least four neck spaces (OR = 15.44, p < 0.001), and mediastinitis (OR = 1.787, p = 0.040) were significant risk factors for repeat I&D. On multivariate analysis, a high blood sugar level (OR = 1.005, p < 0.001) and the involvement of at least four neck spaces (OR = 14.79, p < 0.001) were significant independent risk factors for repeat I&D. Patients who required repeat I&D had longer hospital stays and a higher tracheostomy rate than did other patients (both p < 0.05). The pathogens did not differ between patients who did and did not require repeat surgical I&D (all p > 0.05), but the rates of pathogen non-growth from blood cultures were 19.47% (97/498) in the group without a need for repeat I&D and 0.93% (1/107) in the group with such a need (p < 0.001). DNI can be fatal; a higher blood sugar level and the involvement of at least four neck spaces were independent risk factors for repeat surgical I&D. If at least four neck spaces are involved, we recommend controlling the blood sugar level after admission. We found significant differences in the length of hospital stay and the need for tracheostomy between groups who did and did not require repeat surgical I&D. Although the pathogens did not differ between the groups, pathogen non-growth from blood cultures was less common in the group with for repeat surgical I&D than in the group without such a need.


Introduction
Deep neck infection (DNI) is a life-threatening bacterial infection [1]. Clinical manifestations of DNI include neck swelling and redness, localized heat, and shortness of breath; some patients exhibit airway obstruction. DNI can trigger severe sepsis, esophageal perforation, necrotizing fasciitis, descending necrotizing mediastinitis, disseminated intravascular coagulation, jugular vein thrombosis, and pericarditis [2][3][4][5][6][7][8][9]. In addition to antibiotics, surgical drainage and postoperative wound irrigation are required by patients with advanced DNI [10]. Surgery requires great care; the anatomical landmarks of DNI patients are not clearly evident [11,12]. Because major blood vessels such as the carotid artery and the jugular vein are located in the cervical region, precise abscess drainage is difficult [13]. Repeat surgery is required by some patients during the hospital stay [14,15]. Therapeutic management is complicated. However, few studies have sought risk factors associated with repeat surgical drainage; here, we investigated such factors.

Materials and Methods
We retrospectively reviewed the medical records of 605 patients with DNIs who were admitted to Chang Gung Memorial Hospital (Linkou, Taiwan) between July 2016 and February 2022. DNI was diagnosed by ultrasonography and computed tomography (CT) [7]. The involved deep neck spaces of CT images were identified retrospectively by at least one radiologist (S.-C.C.) and one otorhinolaryngologist (S.-L.C.). If there were situations where the involved deep neck spaces were difficult to identify, it would be decided by team discussion. Treatment included antibiotics and surgical incision and drainage (I&D). I&D was performed when the DNI compromised the airway, the abscess was large (≥2 cm), and when the DNI did not improve after 48 h of intravenous empirical antibiotics [16]. These antibiotics were ceftriaxone 1 g/12 h and metronidazole 500 mg/8 h; they reduced the levels of aerobic and anaerobic bacteria before culture results were available [17,18]. Postoperative irrigation was performed at intervals of 4-8 h depending on disease severity [19]. Repeat I&D was defined as more than one drainage procedure during the same hospital stay. We followed the patient's clinical condition and laboratory data after the first open I&D. If the patient's symptoms and data progressed seriously within 48-72 h, we further arranged a CT exam. If there were abscesses that must be drained, the clinician would arrange a repeated open I&D ( Figure 1A,B).

Results
Demographic and clinical data are listed in

Exclusion Criteria
The exclusion criteria were a swallowed foreign body, severe cardiopulmonary disease, previous head-and-neck tumor surgery, and prior head-and-neck chemoradiotherapy. Follow-up CT was used during hospitalization to exclude patients whose abscesses were initially ineffectively drained. In total, 605 patients were enrolled, of whom 107 required repeat drainage.

Data Collection
We recorded gender, age, C-reactive protein (CRP) and blood sugar levels, diabetes mellitus (DM) status, number of involved DNI spaces, mediastinitis status, length of hospital stay, tracheostomy status, all I&D procedures, and pathogens involved.

Statistical Analysis
All data were analyzed using MedCalc software (ver. 18.6; MedCalc, Ostend, Belgium). Because the Kolmogorov-Smirnov test showed that the data were not normally distributed, we used the chi-squared test to compare categorical variables and the Mann-Whitney U test to compare continuous variables. We employed logistic regression for univariate and multivariate analyses. We engaged in forward stepwise selection, which was followed by multivariate logistic regression; all variables included in univariate analysis were entered into the final multivariate model. For all analyses, p < 0.05 was considered to indicate statistical significance.

Discussion
DNI can trigger life-threatening complications; management includes broad-spectrum intravenous antibiotics, timely surgical drainage, and airway protection (i.e., intubation and/or tracheostomy) [20]. The annual incidence of DNI is approximately 9-15/100,000; the most common etiology is dental [21,22]. Obregon-Guerrero reported that of 87 patients with DNIs, 21 (24%) required re-operation [14]. This burdens the healthcare system and increases medical costs [15,23]. Our rate of repeat I&D was 17.68%. A higher blood sugar level and the involvement of at least four spaces were independent risk factors associated with repeat I&D. Patients with such factors required longer hospital stays and a higher rate of tracheostomy than did other patients. Although the pathogens did not differ between groups, pathogen non-growth from blood was less common in the repeat I&D group (0.93%, 1/107) than in the other group (19.47%, 97/498). As shown in Table 2, a higher blood sugar level, the involvement of at least four spaces, and mediastinitis were risk factors for repeat I&D on univariate analysis. Mediastinitis can be fatal; DNI spreads along the cervical fascia and the neck spaces down to the mediastinum [24]. The mortality rate was reportedly near 40% prior to the 1990s [25]. As diagnostic methods, surgical techniques, and intensive care unit protocols improved over the past 20 years, the outcomes and mortality have also improved. Multidisciplinary approaches and comprehensive medical treatments have significantly reduced mortality [9]. We found that the mediastinitis rate did not differ between the two groups on multivariate analysis. Mediastinitis indicates a severe infection. However, mediastinitis is not necessarily associated with incomplete surgical drainage. If mediastinitis did not occur in a site that was difficult to drain or irrigate, repeat I&D would be unnecessary.
We found that a higher blood sugar level and DNI involving at least four spaces were independent risk factors for repeat I&D on both univariate and multivariate analyses. Liu et al. reported that a high preoperative blood sugar was independently correlated with persistent discharge after operation to treat DNI [26]. Moreover, early blood sugar control has been shown to reduce infectious complications [27,28]. Importantly, we found that DNI involving at least four spaces was an independent risk factor for repeat I&D. The involvement of multiple spaces indicates an advanced infection [1] and increases mortality; elderly patients tend to exhibit more involved spaces than do patients aged <18 years [29]. Multiple-space DNI is associated with severe inflammation and increases the volume of cryptic space that is difficult to irrigate. Although blood sugar level can be the underlying reason for the repeat I&D, it can also be the effect of a more serious inflammatory response due to infection or even sepsis. We still recommend controlling the blood sugar level, especially in DNI patients who exhibit involvement of multiple spaces. Table 3 shows that the length of hospital stay and tracheostomy rate were higher in the repeat I&D group than in the other group. Careful airway protection with close surveillance of respiratory function and adequate infection control are important considerations [30]. Ideally, airway specialists (otolaryngologists and anesthesiologists) should evaluate all patients early to prepare for airway security. The involvement of multiple spaces was previously identified as a risk factor for tracheostomy [31]. Patients undergoing repeat I&D should be closely monitored until airway signs and symptoms resolve [32]. Postoperative ventilation for >48 h was independently associated with unplanned re-operation [15], which is consistent with the findings of other studies: prolonged ventilation was a risk factor for re-operation and re-admission [33,34]. Older age and DM were correlated with a longer hospital stay [35]. Table 4 shows that the pathogens did not significantly differ between groups. The rate of specific pathogen non-growth from blood cultures was 19.47% (97/498) in the group that did not undergo repeat I&D, while it was only 0.93% (1/107) in the group that did undergo this procedure, which differed significantly. In addition, the overall rate of specific pathogen non-growth was 16.19% (98/605). Blood culture does not sensitively identify pathogens, especially when antibiotics have been administered [36,37].
Limitations of the Article This study has some limitations. First, the retrospective nature of the study at a single institution resulted in a certain attrition rate. Furthermore, the repeated drainage for DNI is clinical and determined by the clinician. As such, while the independent variables are logical for repeated drainage, they do not mandate the requirement for repeated drainage. The bias occurred in this mode of data acquisition.

Conclusions
A higher blood sugar level and the involvement of at least four spaces were independent risk factors for repeat I&D. In such patients, the blood sugar level should be controlled after admission. The repeat I&D group exhibited a longer hospital stay and a higher tracheostomy rate than did the other group. Although the pathogens did not differ between the two groups, specific pathogen non-growth from blood was less common in the repeat I&D group than in the other group.

Informed Consent Statement:
The data were collected retrospectively, and the patients were anonymized before data analysis. The IRB waived the need for informed consent.
Data Availability Statement: All data generated or analyzed during this study are included in this published article. The data are available on request.

CT
Computed tomography CRP C-reactive protein DNI Deep neck infection DM Diabetes mellitus I&D Incision and drainage OR Odds ratio