Up to 90 percent of hip dysplasia cases identified during the first six weeks of life resolve spontaneously.1 The effectiveness of universal screening for developmental dysplasia of the hip (DDH) in preventing long-term disability is unknown. One British study reported in 1998 that the rate of infants requiring operative hip procedures was the same before and after the implementation of universal DDH screening (0.66 to 0.85 per 1,000 children before implementation; 0.78 per 1,000 children after implementation).2
It is unlikely that a uniform recommendation supporting or discouraging universal screening for DDH will be made in the near future. Some experts suggest that technology is the answer because clinical examinations lead to many false-positive results, and imaging studies may be more objective. Radiography is not feasible because the hip is poorly ossified at birth. Therefore, ultrasonography has been suggested for use as a screening or diagnostic tool for DDH.3,4 In this issue of American Family Physician, Drs. Storer and Skaggs5 recommend the use of hip ultrasonography as a diagnostic tool to assess clinical improvement from noninvasive interventions such as the Pavlik harness.6
There are two types of newborn hip ultrasound assessment: dynamic and static. Dynamic ultrasonography evaluates the stability of the hip by measuring the movement of the acetabulum; static ultrasonography assesses anatomic hip abnormalities in fixed positions. Although some experts consider imaging to be more objective than clinical examinations, studies have shown that hip ultrasonography is operator-dependent, and that interpretation is subjective.7 For example, the Graf classification is used in most of the literature to differentiate among the spectrum of hip abnormalities, including dysplasia, stability, subluxation, and dislocation.8 When experienced and well-trained ultrasonographers examine a static assessment of a normal hip, they generally produce the same results (interoperator and intraoperator reliability is 98 percent).6 However, when abnormal static assessments are examined, agreement between a first and second interpretation of the same test by the same person is only moderate (kappa, a measure of agreement taking chance into account, is 0.41, with 1.0 being perfect agreement). When a second person interprets the same ultrasound results, agreement between the two assessments is only fair (kappa = 0.28).8,9 Grading of dynamic hip stability shows only moderate agreement between examiners (kappa = 0.42).10 Community-based hip ultrasonography is unlikely to reach even these levels of reliability or agreement because the operator likely will have less experience and training than operators who participate in clinical trials.
In addition to the problems in performing and interpreting hip ultrasonography, universal screening produces high rates of false-positive results, which lead to unnecessary and potentially harmful follow-up radiography and interventions (e.g., the Pavlik harness, which has a low but definite rate of avascular necrosis of the hip).11
Ultrasound screening could be limited to infants with risk factors for DDH (i.e., female, breech position at delivery, or a family history of DDH). However, only 10 to 27 percent of infants diagnosed with DDH in population-based studies have identified risk factors. Using female sex as a risk factor would require screening about 51 percent of all infants.12–15 Among infants with risk factors, only 1 to 10 percent have DDH, severely limiting the positive predictive value of screening based on the presence of risk factors.7,9,10
Hip ultrasonography has been studied as follow-up and diagnostic testing in specialty settings to which infants who failed Ortolani and Barlow examinations were referred.16 In specialty settings, hip ultrasonography decreased the number of children treated for DDH and decreased the number of hip-related hospitalizations and subspecialist office visits. However, it had no impact on the number of infants who eventually underwent hip surgery for DDH. This suggests that diagnostic testing with hip ultrasonography may be useful in subspecialty settings.
Hip ultrasonography does not appear to be a useful primary care addition to universal clinical examination screening for DDH. Because it does not improve the already high false-positive rate, routine hip ultrasonography seems to do little more than increase health care costs. However, it may be valuable as a diagnostic tool for subspecialists by limiting unnecessary interventions.