Graf (1984) asks whether sonography can reliably classify infant hip dysplasia without radiation. Drawing on cadaveric dissection of 20 hip structures and 3,500 infant hip sonograms, the paper proposes a four-type classification based on two measurable angles on a standardized coronal section. This is the foundational paper establishing the static sonographic technique and the Graf classification still used today.
The newborn acetabulum is almost entirely cartilaginous. Plain radiograph shows only ossified structures, misses most of the morphology, and carries radiation risk in a screening population. Graf gave clinicians two measurable angles on a standardized coronal ultrasound section that grade hip morphology from day 2 of life, radiation-free.
When you see a newborn with DDH risk factors (breech, female sex, family history), the Graf type tells you exactly what to do: Type IIa under 3 months: observe, it is physiologic Type IIb over 3 months: reimage at 8-week intervals Type IIIa: start Pavlik harness Type IIIb: escalate to rigid abduction or casting Type IV: urgent reduction
Knowing that up to 90% of hips under 3 months are Type II is the key to avoiding overtreatment. Without this age-stratified distinction, every mildly immature newborn acetabulum becomes a brace candidate.
Germany, Austria, and Switzerland built universal screening programmes on this classification. Subsequent work (von Kries et al., Lancet 2003) confirmed reduced operative rates under screening — all tracing back to the angle thresholds defined here. Critics note real inter-observer variability in angle measurement and the risk of over-sensitivity; the Type IIa/IIb age cutoff at 3 months is precisely what Graf built in to guard against that.
Graf (1984) asks whether sonography can reliably classify infant hip dysplasia without radiation. Drawing on cadaveric dissection of 20 hip structures and 3,500 infant hip sonograms, the paper proposes a four-type classification based on two measurable angles on a standardized coronal section. This is the foundational paper establishing the static sonographic technique and the Graf classification still used today.
The newborn acetabulum is almost entirely cartilaginous. Plain radiograph shows only ossified structures, misses most of the morphology, and carries radiation risk in a screening population. Graf gave clinicians two measurable angles on a standardized coronal ultrasound section that grade hip morphology from day 2 of life, radiation-free.
When you see a newborn with DDH risk factors (breech, female sex, family history), the Graf type tells you exactly what to do: Type IIa under 3 months: observe, it is physiologic Type IIb over 3 months: reimage at 8-week intervals Type IIIa: start Pavlik harness Type IIIb: escalate to rigid abduction or casting Type IV: urgent reduction
Knowing that up to 90% of hips under 3 months are Type II is the key to avoiding overtreatment. Without this age-stratified distinction, every mildly immature newborn acetabulum becomes a brace candidate.
Germany, Austria, and Switzerland built universal screening programmes on this classification. Subsequent work (von Kries et al., Lancet 2003) confirmed reduced operative rates under screening — all tracing back to the angle thresholds defined here. Critics note real inter-observer variability in angle measurement and the risk of over-sensitivity; the Type IIa/IIb age cutoff at 3 months is precisely what Graf built in to guard against that.