Head and neck swellings often present to oral and maxillofacial surgeons. It is often unclear whether a collection of pus is present. This information is essential for decisions regarding surgical intervention. Although ultrasound scanning (USS) is widely used to investigate the presence of a collection, the reliability and validity of this investigation in this context is uncertain. A retrospective review of 4000 consecutive head and neck USS reports over a 4 year period in the Cardiff University Health Board produced 43 cases in which ultrasound had been used to look for evidence of pus collection. The management and treatment outcome of these patients were reviewed and the data analysed. 36 of 43 patients had their swelling incised in theatre, and in 92% of these cases USS and clinical findings corresponded. Of the seven not taken to theatre, four were USS negative and three USS positive; in all seven cases the swelling resolved with antimicrobial therapy. Sensitivity and specificity of USS imaging to identify pus collection were very high, 96% and 82%, respectively. The evidence in this study indicates that USS is a very reliable diagnostic tool in the diagnosis of a collection as well as providing evidence that small collections of pus can resolve without surgical drainage.
Head and neck swellings are commonly encountered in oral and maxillofacial surgery. It is often unclear on clinical examination whether a collection of pus is present . Prompt identification of a collection, or otherwise, is necessary for efficient and timely surgical or medical intervention.
The use of ultrasound scanning (USS) in the head and neck is well established and is used to image the salivary glands, lymph nodes and the thyroid gland, as well as suspected abscess formation . The advantages of USS over other head and neck imaging modalities such as CT and MRI scanning, are that it is readily available, inexpensive, side effects are minimal and it can be repeated. USS guided aspiration may also be carried out, though this technique is known to be operator sensitive . Despite the potential high diagnostic yield of USS in the identification of head and neck abscesses , the reliability and validity of this investigation has not been studied in great depth.
The aim of this study was to investigate the reliability of USS in identifying collections of pus in swellings of the head or neck.
Materials and methods
4000 consecutive USS reports, over a 4 year period (2004–2008) from the University Dental Hospital, Cardiff (UDH) and the University Hospital of Wales, Cardiff (UHW) were reviewed retrospectively. From these, 43 cases were identified in which USS had been carried out to confirm or eliminate the possibility that a collection was present. The other cases were excluded when a non-infective diagnosis was assumed. The USS report findings were used as a basis for case selection. 37 of these scans were performed by one UDH consultant dental and maxillofacial radiologist, and the remaining 6 scans were performed by UHW radiologists. In the majority of cases, a Toshiba Nemio 20 ultrasound unit was used (Toshiba, Tochigi-ken, Japan). A high resolution (9–12 MHz) linear probe was used to scan the superficial tissues in the head and neck. The main diagnostic features used to make the diagnosis were a round/oval hypoechoic mass, with or without internal debris. The case notes of the 43 patients identified for the study were scrutinised for details of medical and surgical management. Data was entered onto a proforma for each patient ( Fig. 1 ) and the sensitivity and specificity of USS calculated.
43 patients were identified for whom USS had been requested to assess the possibility of the presence of a collection. All cases were regarded as superficial space collections and were diagnosed on USS in the following anatomical regions: submandibular (21), submental (5), submasseteric (6), facial (10) and parotid (1). The age range was 2–62 years with an average age of 37 years. The main cause of infection was dento-alveolar (70%) but other causes included infected plates, sebaceous cysts and a branchial cyst. Of the 43 patients who were scanned for suspected pus collection, surgical drainage was attempted in 36. In the majority of these cases the time interval between USS and drainage was within 24 h. USS and surgical findings for these patients are set out in Table 1 . There were 24 true positive correlations between USS and surgical findings, and nine true negative correlations, therefore, in 92% of patients taken to theatre, findings corresponded. Sensitivity was calculated to be 96% which demonstrates the percentage of patients who actually had a collection who were correctly identified. Specificity was 82% which showed the percentage of patients who did not have a collection who were correctly identified ( Fig. 2 ). A positive predictive value of 92% (the proportion of patients that the USS showed as having a collection who actually had a collection present) and a negative predictive value of 90% (the proportion of patients with negative results on USS who were correctly diagnosed) were subsequently calculated.
|Collection on US report?||Pus on incision?||Outcome||Number of patients|
|No collection||Yes||False negative||1|
|No collection||No||True negative||9|
The findings for the seven patients who were not taken to theatre are set out in Table 2 ; three were not taken to theatre despite having a collection identified using USS. The remaining four patients were not taken to theatre following negative USS findings. In all seven cases, swelling resolved with antimicrobial therapy alone.
|Collection on US report?||Swelling resolved?||Number of patients|