» When To Abandon the Pavlik Harness
When To Abandon the Pavlik Harness

Physical Therapy in Grapevine for Pediatric

For 50 years, children born with developmental dysplasia of the hip (DDH) have been treated successfully with the Pavlik harness. Most studies show that the earlier the treatment, the better the results. The Pavlik harness is not usually recommended for older infants (six months of age or older). The results of this study may suggest differently.

What is developmental dysplasia of the hip (DDH)? In this condition there is a disruption in the normal relationship between the head of the femur (thigh bone) and the acetabulum (hip socket). The socket is shallow and the femoral head doesn't stay in the socket.

DDH can affect one or both hips. It can be mild to severe. In mild cases called unstable hip dysplasia, the hip is in the joint but easily dislocated. More involved cases are partially dislocated or completely dislocated. A partial dislocation is called subluxation.

The goal of treatment is to keep the femoral head in good contact with the acetabulum. A stable hip encourages the development of a normally shaped socket and rounded head of the femur. The proper hip position must be maintained for enough time to stabilize the joint. The hip should be flexed to 95 degrees and abducted (apart) at least 90 degrees. This position keeps the ball (the femoral head) in the best position and allows the ligaments and joint capsule to tighten up.

The best way to achieve this is often with the Pavlik harness. The harness keeps the hip in flexion and abduction. It may be worn until the doctor can no longer move the hip in and out of the socket. Usually this takes about six weeks. As mentioned, the harness is not recommended for children older than six months.

But the results of this study may indicate otherwise. Twenty-four children between the ages of nine months and 23 months (almost two years old) were placed in the Pavlik harness despite the late diagnosis.

Almost half (46 per cent) had a successful result and did not need any additional surgery. Results were obtained in the first six weeks if the harness was going to work. The harness was augmented by a dynamic abduction splint in about half of the group. But the results were not any better with the dynamic abduction splint than without.

In one case, the Pavlik procedure worked for one hip but not for the other. For those who had a failed result with the Pavlik harness, surgery was done to reposition the hip in the socket. The children with the most severely dislocated hips had the worst results.

There are different ways to describe or classify hip dysplasia. The authors of this study used the Graf classification, which uses ultrasound studies to look at the shape of the acetabulum. The more flattened the roof of the socket, and the further off center the head of the femur was positioned in the socket, the worse the Graf grade. Children with Graf type 4 (most severe) dysplasia were the most likely to fail late Pavlik harness treatment.

As a result of this study, the following recommendations were made:

  • The Pavlik harness can be used successfully to treat hip dysplasia in children older than six months.
  • This treatment won't work for children classified as Graf type 4.
  • Use of the harness must not extend past six weeks. If the hip has not gone back in the socket and stayed there, it's not going to.
  • There's no benefit of adding a dynamic abduction splint to the harness.

And the final conclusion: the Pavlik harness can be used successfully to treat developmental dysplsia of the hip with carefully selected older infants and children. A delayed diagnosis is not always an immediate passport to surgery. Surgery can be delayed by six weeks to see if the harness will work in older children. If the harness fails, then surgery can be done.

Reference:  Virginie Pollet, MD, et al. Results of Pavlik Harness Treatment in Children with Dislocated Hips Between the Age of Six and Twenty-four Months. In Journal of Pediatric Orthopaedics. July/August 2010. Vol. 30. No. 5. Pp. 437-442.