“Deep neck infection is a common disease in otolaryngology clinics and sometimes can result in life threatening comorbidities,” Yehai Liu, MD, PhD, explained.

The term deep neck infection (DNI) specifically refers to cellulitis or abscesses in the deep cervical fascial space, according to the researchers, and can arise from progressive infections of the tonsils, odontogenic structures, pharynx, esophagus, trachea, or salivary glands. DNIs may spread to interconnected soft tissue sites, and such spread is associated with the potential to cause severe symptoms and poor health outcomes.

“However, a prediction model for mortality associated with deep neck infections is still lacking,” Dr. Liu said. “Our work herein aimed to predict the prognosis of the disease and contribute to doctor-patient communication and understanding before treatment is initiated.”

The retrospective study, which was published in Ear, Nose & Throat Journal, aimed to identify a novel model capable of predicting treatment outcomes in DNI. Dr. Liu and colleague examined the medical records of 198 patients treated for DNIs between 2010 and 2022; among this group, 153 patients were enrolled in the study
(median age, 48.0). Logistic regression models were used to identify risk factors associated with mortality, and a predictive model was developed based on ORs for factors calculated using a multivariate regression model.

Predictive Role of Creatinine, Albumin, & Mediastinal Infections

The study team determined that being older than 50, residing in a rural area, dyspnea at the time of presentation, involvement of multiple infected sites, serum albumin levels of less than 34 g/L, renal insufficiency, mediastinitis, pulmonary infection, and septic shock were independent predictors of mortality in DNIs. Serum creatinine levels of greater than 95 μmol/L were “particularly critical” to outcomes, according to the study results.

Most patients received treatments that resulted in recovery or stabilization; 15 patients had treatment interventions that led to a worsening of the condition or death. The average age of patients who recovered was significantly younger than the average age of the patients who declined (44.9 vs 60.4). Almost all patients (N=14) who were not cured of the DNI resided in rural areas, while only about half of the patients who were cured resided in such an area. The researchers observed a significant difference in these rates between groups, according to the study results (P<0.05).

Once the risk factors for mortality were entered into a multivariate model, only a creatinine level of greater than 95 μmol/L, an albumin (ALB) level of less than 34 g/L, and mediastinal infections were retained for the development of the final predictive model. The model used two condition descriptors labeled condition 1 and
condition 2. Condition 1 described patients with creatinine levels of greater than 95 μmol/L, and condition 2 described patients with mediastinitis and ALB levels of less than 34 g/L.

The patients were then separated into three groups using these condition descriptions (Figure). The first was the recoverable group; this group did not have condition 1 or condition 2, and 98.4% of these patients survived. The high-risk group had either condition 1 or condition 2; about half of these patients eventually worsened or died. The third group, known as the fatal group, presented with both condition 1 and condition 2; two-thirds of these patients died. The area under the receiver operating characteristic curve for this model is 0.896, with an 86.67% sensitivity and a 92.03% specificity.

Directions for Future Research

“Our study revealed that mediastinitis, renal dysfunction, and hypoalbuminemia were independent predictors of mortality in patients with deep neck infection, with renal dysfunction being particularly critical to the outcomes,” Dr. Liu said.

He noted that, in the present study, “the mechanisms by which mediastinal infection, renal dysfunction, and hypoalbuminemia lead to poor prognosis have not been elucidated,” and suggested that “future research could focus on this point.”

Regarding the current work, Dr. Liu believes the results “could help practitioners recognize high risk cases.”