|Home | About | Journals | Submit | Contact Us | Français|
Insufficient evidence exists to support universal screening
In this week's BMJ, Wald and colleagues propose a universal screening strategy for familial hypercholesterolaemia.1 They suggest that serum cholesterol should be measured in children aged 1-9 years during routine visits to primary care, and that those with abnormal total cholesterol (greater than 95th centile) should have genetic tests or clinical investigations to confirm the diagnosis. A population cascade screening programme could then identify the parents of children who screen positive for the disorder.
This proposal is based on their meta-analysis of screening for familial hypercholesterolaemia. This study showed that measuring serum cholesterol in children age 1-9 can detect 88%, 94%, and 96% of cases, with false positive rates of 0.1%, 0.5%, and 1%, respectively.1 Their proposal is based on the ability of the test to detect the disorder with a reasonably high detection rate and a relatively low false positive rate. The authors present no new evidence for the long term health benefits or potential harms of identifying and treating children with familial hypercholesterolaemia.
Much attention has been paid to screening for lipid disorders at a young age because half of children with high concentrations of total cholesterol and low density lipoprotein will continue to have raised lipids in adolescence and early adulthood, and early identification and treatment in certain populations of adults can prevent coronary heart disease.2 A screening programme could identify three groups of children with abnormal cholesterol concentrations—children with monogenic dyslipidaemias, such as familial hypercholesterolaemia; those with undiagnosed secondary causes of dyslipidaemia (diabetes, hypothyroidism, etc); and those with multifactorial dyslipidaemias (polygenetic or related to risk factors, such as obesity). The group most likely to benefit from screening, earlier identification, and treatment would be children with familial hypercholesterolaemias. In these children, treatment with statins and bile acid resins improves lipid profiles and intermediate outcomes.2 In children with abnormal lipids but without familial hypercholesterolaemia (multifactorial dyslipidaemias), evidence shows that medical or behavioural interventions do not improve lipid levels.2
The US National Cholesterol Education Panel and American Academy of Pediatrics recommend screening in children with a positive family history of hypercholesterolaemia or those with risk factors.3 This has been problematic because of a high false negative rate in detecting high serum cholesterol, ranging from 17% to 90%, as a result of variable definitions of positive family history and differing thresholds of abnormal cholesterol. Taking a family history is also associated with a high false positive rate, with 25-55% of children and adolescents qualifying for serum cholesterol screening on the basis of family history alone.2
The screening strategy proposed by Wald and colleagues seeks to identify only those children with familial hypercholesterolaemia by requiring specific criteria for a clinical diagnosis: total or low density lipoprotein cholesterol above a given value, raised serum cholesterol in a first degree relative, and a family history of tendon xanthomata. As reported by Wald and colleagues, detection rates are relatively high, but even with a relatively low false positive rate, a universal screening programme without genetic confirmation will identify a large number of children who do not have the disorder. However, a programme that incorporates genetic confirmation of the diagnosis is likely to be expensive.
A third strategy, which is more appropriately called case finding rather than true screening, is cascade screening, where the family members of all patients with known familial hypercholesterolaemia undergo clinical diagnosis or genetic testing. This strategy is endorsed by the UK National Screening Committee and supported by the 2000 health technology assessment report, which concludes that such a case finding strategy in relatives of patients with familial hypercholesterolaemia followed by a clinical or genetic diagnosis would be most cost effective.4
Unfortunately, we have no direct evidence on the adverse effects of any of the above screening strategies. No data are available on the safety of long term treatment with drugs started in childhood or adolescence. Although lipid concentrations in children with familial hypercholesterolaemia can be improved with treatment,2 we have no evidence of a long term benefit on health. If a benefit exists, the difference between this benefit and that associated with detecting and treating the disorder in adults would need to be examined, and there is currently no evidence to do this.
Finally, while the false positive rate may be low, those children who are found to have raised cholesterol but who do not have familial hypercholesterolaemia or are false positives may be treated unnecessarily. Treatment in children with non-familial hypercholesterolaemia has not been shown to improve health outcomes in children or adults,2 and again the long term safety of lipid lowering agents in young children has not been determined.
On the basis of current evidence, the most cost effective approach to identifying people with familial hypercholesterolaemia is case finding in the family of those known to have the disorder. There is insufficient evidence to support universal screening with either serum cholesterol followed by clinical or genetic confirmation, or family history taking followed by serum testing.5
Competing interests: None declared.
Provenance and peer review: Commissioned; not externally peer reviewed.