Hussein Mohamedbhai, Stephen Ali
Introduction & aims
Tracheostomy has a well established record of providing relief to imminent airway compromise. However, the most effective and safest tracheostomy is that done electively. It is the duty of the clinician to perform an elective tracheostomy for head and neck cancer judiciously. Increasingly, predictive scoring systems are being utilised as an objective model. The aim of this study is to design a simple, sensitive and specific scoring system which can be used patient’s with head and neck cancer undergoing surgical management.
Retrospective case note review was conducted of all patients at a tertiary head and neck cancer unit who were diagnosed with a head and neck cancer. Categorical regression analysis was completed and P values which indicate statistical significance were obtained.
149 patients were included in this study. The most statistically significant factors were then included and scored as according to their positive predictive value. This produced a score based upon T (T staging), R (Reconstruction), A (Anatomy of tumour), C (Co-morbidities), H (History of previous head and neck cancer), Y (lateralitY- bilateral neck dissection). Applying a threshold score of 5 demonstrate a sensitivity of 89%, a positive predictive value of 90%, a specificity of 89% and negative predictive value of 86%. Applying this scoring system to the patient cohort demonstrated a mean score of: 2.6 (intubated), 7.1 (primary tracheostomy), 6.4 (late tracheostomy).
This test can appropriately predict those patients not needing tracheostomy and can act as a reliable screening tool in pre-operative airway planning. Our tool showed concordance between those patients undergoing primary tracheostomy and late tracheostomy. It could provide a way to aid management, and reduce the frequency of post-operative tracheostomies. Thus, there is the potential that patients could be managed safer, with a reduced morbidity and mortality.