Clinical screening for developmental dysplasia of the hip in Northern Ireland

Background

Developmental dysplasia of the hip describes a spectrum of disease ranging from minor acetabular dysplasia to irreducible dislocation and may lead to premature arthritis in later life. This term has replaced the more traditional “congenital dislocation of the hip,” which assumes dislocation present at birth and does not include hips that are poorly developed or subluxed. Early diagnosis is thought to improve outcome1 2 3 because non-surgical treatments can be used to optimise development of the hip joint. Despite this, however, some patients diagnosed early and treated conservatively may still require surgical intervention to obtain a stable reduction. The risk of requiring surgery increases as age at diagnosis rises, with nearly all patients who are older than 18 months at diagnosis needing open reduction.4 This has prompted the widespread adoption of neonatal screening programmes aimed at reducing the incidence of late presentation.

Clinical screening was introduced in the United Kingdom in 1969, and in 1986 the Standing Medical Advisory Committee recommended that screening should occur within 24 hours of birth, on discharge from the hospital of birth, and at 6 weeks of age.5 Despite these efforts, late presentation has not been eliminated, and in some places it has risen. 6 7 Although concern has been raised about the screening technique,8 9 10 the evidence shows that neonatal screening by experienced staff is effective. 11 12 Nevertheless, recent reports casting doubt on the efficacy of screening for developmental dysplasia of the hip 13 14 led the Medical Research Council to call for a formal evaluation of the UK neonatal screening programme.15

Ideally screening should use a simple reliable test that has been shown to have high sensitivity and specificity, is inexpensive, and provides results that suggest a clear course of action. As these criteria are not fulfilled in the case of developmental dysplasia of the hip, the term screening has been criticised,13 and surveillance has been suggested as a more appropriate alternative.

In Northern Ireland, screening primarily consists of clinical examination using the Ortolani and Barlow tests 1 2 and assessing the range of hip abduction and leg length discrepancy. It is carried out in the hospital of birth by hospital staff and again by the general practitioner at 8 weeks of age. The health visitor assesses the infant at 10 days, 7 months, and 18 months, gait being assessed on the final occasion. These assessments are carried out in accordance with the guidelines in the Northern Ireland child health system surveillance programme. Uniquely to Northern Ireland, a record of these checks is kept in the personal child health record or “red book,” which is issued to every newborn in the region.

The problem

A study of the outcome of screening of children born in the province during 1983-7 reported a high incidence of developmental dysplasia of the hip (1.75 per 1000) and a late incidence (diagnosis after 6 months) of 1.14 per 1000.16 As this was a relatively poor outcome it was felt that the service needed to be improved.

Key measures for improvement

The overall incidence of developmental dysplasia of the hip in Northern Ireland is unknown because no data are available on the numbers of patients who have successful conservative treatment. We therefore used the incidence of children requiring operative intervention as a measure of the efficacy of screening. We assumed that the overall population incidence has remained reasonably static in the region over the past 10-15 years to enable direct comparisons between this and the previous study.

Gathering information

We did a comparative retrospective study to determine the late incidence of developmental dysplasia of the hip in children born during January 1991 to December 1997. Late incidence was defined as diagnosis after 6 months of age and requiring operative intervention. We identified all children born in Northern Ireland during the study period who had had operative intervention for hip dysplasia by December 2000 from the Belfast Orthopaedic Information System (IT Department, Musgrave Park Hospital, Belfast, Northern Ireland). We also reviewed theatre logbooks and consultant and nursing records to ensure complete sample capture.

Operative intervention was defined as closed reduction, open surgery, or arthrography followed by immobilisation. We excluded children who had arthrography without immobilisation as well as children with cerebral palsy, spina bifida, arthrogryposis, and other neuromuscular disorders. We believe that sampling for this study is complete because Northern Ireland has a stable population and a unified healthcare service and there are only two regional centres with a small number of staff dealing with the condition.

To ensure that the results were comparable with the previous study we defined late incidence as diagnosis after 6 months of age. However, we also obtained results for those presenting after 3 months to enable comparison with other studies. Age at diagnosis was defined as the time at which the orthopaedic consultant made a written record of the diagnosis.

Strategies for change

Since the previous study, several changes have been implemented in Northern Ireland. These include:

• Improved staff training—targeting all members of staff responsible for child health surveillance including health visitors and general practitioners

• Greater use of ultrasonography—mainly for cases in which there is clinical doubt about the status of the hip joint or in children considered to be at a high risk of developing hip disorders

• Centralised nurse led clinic—introduced in 1995 to cope with the increase in orthopaedic referrals. Specialised nurses examine patients referred from general practitioners, paediatricians, and other orthopaedic consultants and use radiography and ultrasonography as indicated. Children who are thought to have an abnormality are urgently referred to a consultant paediatric orthopaedic surgeon. Those at high risk are kept under review until radiographic findings are satisfactory

• Improved access to consultant paediatric orthopaedic surgeons through the nurse led clinic and by regular consultation between the disciplines concerned with treating this condition.

Effects of change

A total of 196 hips in 179 children born within the study period required operative intervention for developmental dysplasia. As there were 173 746 live births during this period, the operative rate was 1.03 per 1000. Of the affected children, 156 (87%) were girls and 23 (13%) were boys, giving a ratio of 6.8:1. The age at diagnosis ranged from 2 days to 83.5 months (table 1), the mean age being 9.0 months. Seventeen children had bilateral developmental dysplasia of the hip, 14 of whom were not diagnosed until after 3 months of age and eight of these after 6 months of age. If we used 3 months as the threshold for late diagnosis, 140 children (154 hips) were classified as late (0.81 per 1000 of the population). The figure decreased to 102 children (110 hips) with a 6 month threshold, giving a late incidence of 0.59 per 1000 population. This incidence is lower than the 1.14 per 1000 reported for 1983-7.16 table 2 gives the prevalence of risk factors for developmental dysplasia of the hip among affected children and table 3 shows the reasons for referral.

Of the 196 hips, 130 (66%) were definitively treated by closed reduction with or without adductor tenotomy (table 4). Forty eight hips initially managed with closed reduction subsequently required open surgical intervention. The need for open surgery was significantly more prevalent in hips of patients diagnosed after 6 months of age than in those diagnosed earlier (33 (30%) v 15 (17%), ²=5.02, P<0.025). Eight hips (19%) in patients identified aged 0-3 months required open reduction after failed closed reduction compared with seven (16%) in patients identified aged 3-6 months, an insignificant difference.

Seventy one hips in children under 3 months of age presented to the orthopaedic service. Twenty nine were initially deemed not to require intervention but were found to be dysplastic or dislocated 3.4-12.9 months later. Most of the children were kept under review, and consequently 17 of the 29 abnormal hips were diagnosed before 6 months of age; the remaining 12 were not treated until after the child was 6 months old.

Nine of the 29 hips (in seven patients) had been examined by ultrasonography at the initial presentation. Three of these early scans were reported as showing abnormal results, but the hips were not treated at the time. The reasons for this are being investigated.

Lessons learnt

The fall in late presentation of developmental dysplasia of the hip in our study compared with a study in 1983-7 is probably due to the changes in screening practice. Screening will never eliminate the need for surgical treatment, as shown by the fact that 42 children in our study required operative intervention despite being diagnosed within the first 3 months of life. Nevertheless, competent screening practices should reduce the late presenting operative rate.

Our study highlighted several areas where further improvements could occur. We found that a surprising number of children with risk factors such as breech presentation and family history were overlooked and not referred early to the orthopaedic service. In addition, 29 affected hips (16% of the total affected) were referred to orthopaedic services before the children were 3 months old but were not diagnosed immediately. Reasons for this include errors in technique of clinical examination, lack of reliability of the Barlow and Ortolani tests,8 the relatively infrequent use of ultrasonography during the early study period, and the possibility that not all late presentations are preceded by neonatal instability.17 Most of the children whose hips were initially thought normal presented in the first four years of the study (1991-4) before the nurse led clinic was established, and almost half were later shown to have bilateral dislocated or dysplastic hips. There were also a few false negative ultrasonography results.

Next steps

Centres that routinely use ultrasonography to screen neonates report lower incidences of late presentation than ours, and in a study by Marks et al there were no late cases.18 However, such an approach raises concerns about overdiagnosis and overtreatment. It is therefore unlikely that it will be possible to eliminate late diagnosis. Nevertheless, our results suggest there is room for improvement. Theoretically, if all the patients seen within the first 3 months of life had been diagnosed at the first assessment, all patients with risk factors of breech presentation and positive family history had been screened and diagnosed early, and there was no system and human error, a further 31 children would have been diagnosed and treated early. This would have reduced the incidence of late presentation to 0.41 per 1000, which will remain a target to aim for in the future.

A subsequent audit of children born after the introduction of the centralised nurse led clinic is intended to assess fully its impact on clinical screening. We hope through this audit and by future improvements in screening to meet our target and further reduce the late diagnosis of developmental dysplasia of the hip.