[Measurement on printed paper strips or on the ultrasound device with computer assistance - which technique is more...

Two different measurement techniques of ultrasonograms of the infant hip were evaluated with respect to reproducibility of the Graf classification and variation of α- and β-angles.

In a cross-sectional, blinded study, the hips of 207 consecutive newborn babies (101 male; 106 female) were sonographically screened at an average of 2.64 days after birth. Each hip was measured twice by three investigators with different levels of experience - a paediatric orthopaedic surgeon, a senior surgeon and a trained medical student. A mobile ultrasound system (SONOLINE G60S®, Siemens, Erlangen, Germany), equipped with a 7.5 MHz linear transducer, was used. Both hip joints were measured twice by all three investigators. The measurement was performed 6-8 weeks later in a blinded manner. The sonograms were initially printed out on high-quality paper strips and measured by pencil, ruler and goniometer. Finally, each investigator evaluated the same sonograms computer-assisted, using the trackball and dashboard of the ultrasound system.

Concerning intraobserver reliability, we observed a significant reduction of variation both for α- and β-angles in favour of the classic measurement on printed strips (p < 0.05). The interobserver calculation also detected a trend for higher angle variation when the angles were measured electronically. The reproducibility of Graf classification was not influenced by the kind of measurement technique. The outcome was not affected by investigator's level of experience (p > 0.05).

This study demonstrated considerable advantages for the classic measurement of paediatric hip ultrasonograms with pencil and goniometer on printed paper strips compared to computer-aided measurement concerning variation of α- and β-angles.