Comparison of clinical estimations and radiological measurements in the context of paediatric cervical masses: How accurate are we?

Objectives: Cervical masses represent a common presentation to the paediatric ENT outpatient clinic. Despite their high prevalence, few represent an underlying malignant pathology. Currently, there is no universally accepted algorithm for management of such cases in the United Kingdom. Local guidance often centres around clinical interpretation of size when determining the need for investigation. With such emphasis placed on size, the authors sought to investigate the accuracy of physician palpation and in turn explore the validity of common referral pathways.

Methods: A retrospective methodology was used to address the research objective. All paediatric ultrasound scans conducted during a 24-month period were reviewed to compare the clinical and radiological assessments of diameter. A 2 tailed paired student's t-test was selected to evaluate the differences between the two measurements and ultimately assess clinician accuracy.

Results: 753 paediatric cervical ultrasound scans were conducted during the 2-year period. Initial analysis identified three discrete ultrasonographically diagnostic groups: lymph nodes (n = 532), non-lymph node masses (n = 162) and normal anatomy (n = 59). Further analysis of clinical estimation and radiological measurement demonstrated a statistically significant difference (p < 0.02) for the lymph-node group. A statistically significant difference was not demonstrated within the non-lymph node group (p = 0.66). When clinician assessment resulted in a value outside of what was considered accurate, lymph node sizes were predominantly overestimated (62/91, 68%). Conversely, non-lymph node masses where the clinical measurement was not within 5 mm of the ultrasound measurement were more commonly underestimated in size (11/45, 55%) but this was not statistically significant.

Conclusions: Our study demonstrates a statistically significant tendency for clinicians to overestimate paediatric cervical lymphadenopathy. Calliper measurement may further improve accuracy of assessment.