Extended newborn screening: Which diseases?
R. J. Pollitt, Neonatal Screening laboratory, The Children's Hospital, Sheffield S10 2TH
In 1995 the NHS Health Technology Assessment Programme launched its first year by commissioning a systematic review on the role of neonatal hearing screening in the detection of congenital hearing impairment and two reviews on newborn screening for inborn errors of metabolism. These were followed later by reviews on screening for haemoglobinopathies and for cystic fibrosis (all available on http://www.ncchta.org/ ).
Unfortunately, there was no clear path to implementing the findings of these reviews, generating a great deal of frustration amongst those who had invested time and effort in the initial reviews as well as other clinicians and academics who were eager to innovate and improve their services to patients.. Implementation of hearing screening and newborn screening for sickle cell disease is now well advanced but with cystic fibrosis and extended screening for inborn errors of metabolism progress has been slow. A major stumbling block has been the difficulty of assessing the complex and diverse issues involved.
There have been many attempts to provide a universal framework within which prospective screening programmes can be evaluated. The ten principles promulgated by Wilson and Jungner in 1968 provided the foundations upon which to build. In the UK responsibility for advising on screening policy is now firmly centred in the National Screening Committee (NSC) which has developed the ten principles into 23 criteria. The NSC Handbook provides a valuable basis for assessing prospective screens and under 35 general headings describes 87 specific items of information that must be assessed by systematic review. However, such a universal framework can create problems when applied to newborn screening. For rare metabolic diseases, the specified level of evidence would be unachievable within a realistic cost and time, even if the ethical problem of equipoise within the required randomised controlled trial could be overcome. Additionally, the new criteria, like the Wilson and Jungner originals, are predicated on screening adults. They ignore the family dimension implicit in screening newborns. For example, despite its inability to modify outcome, there is strong support from many parents for newborn screening for Duchenne muscular dystrophy because of the genetic information that it can give and its role in easing the family journey. If the concept of informed choice is to be taken seriously there seems no reason (other than the very real one of resource constraints) why such a screen should not be offered.
Very slow progress has been made in exploring the possibilities of tandem mass spectrometry (MS-MS) screening in the UK, though elsewhere in the world it is becoming commonplace. This technique can replace older methods of detecting phenylketonuria and, using the same Guthrie blood spot, simultaneously screen for a variety of other disorders of intermediary metabolism. The technical complexity of the screening method and subsequent confirmatory tests, the multiplicity of disorders that can be covered and their differing natural history, and the lack of quantitative data on incidence and test performance have all caused problems during the debate. There is difficulty in reconciling the various professional paradigms and some of the NSC criteria appear inappropriate when considered in the context of everyday clinical experience.
A two year pilot study of MS-MS screening, initially limited to medium-chain acyl-CoA dehydrogenase deficiency (CMAD), started in England in April 2004. This will provide data on test performance, and should give greater confidence in the practicability of the technique, though within limits having firmer numerical data is unlikely to affect the decision whether to continue the screen. The main factors in considering the range of diseases covered are likely to be the positive predictive value of an abnormal screening result and the ease with which a presumptive diagnosis can be confirmed.