Abstract
The aim was to determine reliability in treatment threshold based on USS angular measurements between observers involved in the DDH hip screening programme at the NOC and assess the effect of image orientation on the accuracy of these measurements.
3 independent observers measured alpha and beta angles on bilateral hips in 10 consecutive patients seen in the DDH hip screening clinic. All scans were performed by a single radiographer and observers used the same set of USS images for a given patient. Each observer measured alpha and beta angles a total of 4 times: conventional ultrasound image projection (with the ilium horizontal) (round 1), Graf's anatomical projection (round 2), and both techniques repeated 1 month later (round 3 and 4 respectively) to assess intra-observer reliability. To determine its effect on treatment threshold taking into account alpha and beta angles and patient's age, the consistency between observers' management recommendations was evaluated for each round. Possible outcomes were: 1) patient discharged, 2) no treatment needed yet, but follow-up required, 3) start treatment.
Intra-observer reliability for conventional projection was moderate (Kappa 0.58), and improved for anatomical projection (Kappa 0.65). Inter-observer reliability, as a surrogate measure of consistency in management recommendations between observers, ranged from fair to moderate across the 4 rounds (Kappa 0.30 – 0.50). However, contrary to previous recommendations, reliability was better with conventional projection (Kappa 0.41 (95% CI 0.11–0.72)) compared to anatomical projection (Kappa 0.36 (95% CI −0.01–0.73)). The overall agreement in management recommendations, pooling all results across 4 rounds, was 51.3% (Kappa 0.39 (95%CI 0.15–0.63)).
This audit supports the argument that anatomical image projection improves intra-observer consistency. However, as with all USS measurements, angular measurements were highly user dependent and treatment threshold based on USS may not be as consistent as anticipated.
