The natural history of Perthes' disease: Risk factors in 212 patients followed for 5 years

BACKGROUND: The prognosis in Perthes' disease varies considerably according to certain risk factors, but there is no concensus regarding the relative importance of these factors. We assessed the natural history of the disease and defined prognostic factors of value in deciding the proper treatment. PATIENTS AND METHODS: During the 5-year period 1996–2000, a nationwide study on Perthes' disease was performed in Norway. 425 patients were registered. The present study involved the 212 children (mean age 5.1 years, 77% boys) who were affected unilaterally and who had been treated with physiotherapy only (which is considered not to change the natural history). They were followed by taking radiographs at the time of diagnosis and after 1, 3, and 5 years. At the 5-year follow-up, the outcome was evaluated according to a modification of the Stulberg classification: good (spherical femoral head), fair (ovoid femoral head), and poor (flat femoral head). RESULTS: The 5-year radiographic results were strongly dependent on 4 risk factors: age 6 years or more at diagnosis, total femoral head necrosis, height of the lateral pillar of the epiphysis less than 50% of normal height, and femoral head cover less than 80%. As the number of risk factors increased from 0 to 4, the proportion of patients with good radiographic 5-year outcome decreased from 79% to 0% and the proportion with poor outcome increased from 3% to 91%. INTERPRETATION: Most children under 6 years of age do not need any special treatment. In older children, no special treatment is indicated if the whole femoral head is not necrotic and the femoral head cover is > 80%. In the most severe forms of the disease (i.e. more than 2 risk factors), surgical containment treatment seems advisable.