Although no politician will commit career suicide by admitting it, rationing of healthcare is inevitable, whatever the healthcare system, recession or no recession. Medicines, particularly biologicals, modern surgery and intensive care are pushing back the boundaries of what can be done to salvage seemingly hopeless situations, to give but three examples. The UK will never be able to pay for all possible medical care for all our citizens, irrespective of cost and benefit. Thus, if medical rationing is inevitable then it should be as fair and transparent as possible, and conducted by experts not politicians. The National Institute for Health and Clinical Excellence (NICE) has the unenviable task of determining who can be prescribed expensive medications or treatments used in large numbers which form a significant percentage of the drugs budget. At the same time, the ‘elephant in the room’ of rationing remains unacknowledged. Furthermore, Health Technology Assessments conducted by NICE project teams are unduly restricted by a rigid remit imposed externally, which precludes reasonable reviewing of indirect data which may be highly relevant to the topic under consideration. Notwithstanding these comments, their recent decision to deny 6–11-year-old children the anti-IgE monoclonal antibody omalizumab as a treatment for asthma, while allowing it for 12 year olds and above, is wrong, and importantly, highlights ways in which the current process is flawed.

Most asthmatic children are easily treated with low doses of inhaled corticosteroids (ICS). Some need an add-on medication, usually a long-acting β-2 agonist or sometimes a leukotriene receptor antagonist. A very few have high levels of morbidity and deaths still occur, despite polypharmacy and high doses of ICS. The evaluation of such children, who have been referred as ‘problematic, severe asthmatics’ should be conducted by specialists.1 When this is done, it is clear that many do not have ‘severe, therapy resistant’ asthma, requiring expensive and invasive medications, but the true problem is that the basics have not been adequately addressed.2,–,4 Nurse led home visits in 71 children with problematic severe asthma revealed that more than 50% had potentially reversible factors, such as poor adherence and ongoing exposure to allergens or tobacco smoke in the home.5 Thus, on a population basis, even in groups of children with apparently severe asthma, only a minority might be considered suitable for omalizumab.

It is wrong, however, to extrapolate from populations to individuals—which is in effect what NICE has done. Although many problematic severe asthmatics can be managed by ensuring administration of basic medications, of those who fail to respond and go on to invasive investigation such as bronchoscopy, half may have sputum eosinophilia despite parenteral corticosteroids.6 It is this group of children who might benefit from omalizumab. The risk of adverse effects from alternatives, such as oral corticosteroids, is considerable.

The NICE ruling on the use of omalizumab in 6–11-year-old children raises a number of issues. One important general issue is that the group who made this ruling consisted of 34 people, of whom only one was a paediatrician (an intensivist, who would not expect to be involved in prescribing omalizumab). One shudders to think what would have happened if this group had passed themselves off as expert witnesses on severe paediatric asthma in the High Court; any barrister would have rapidly demolished them, and the General Medical Council would have disposed of the remains. Would they have been selected as reviewers by a top grant giving body, or a high impact factor journal? They did seek views from one expert respiratory paediatrician and one expert patient (hardly a large sample size; even most medical journals look for at least two and often more specialist reviewers for each manuscript!). Consequently, they made the wrong decision. One solution for families with a child with really severe asthma age 6–11 years who want to trial omalizumab is to move to Scotland, where the diametrically opposite decision has been reached! Another for those who can afford the medication is to pay for it privately.

There are also specific issues related to this NICE review. There was an apparent lack of appreciation of the importance of exacerbations, disrupted schooling and the effects on the wider family when measuring paediatric quality adjusted life years. In fairness, such measurement tools are poorly developed for children. The long-lasting effects of disrupted schooling, affecting future adult life did not appear to be considered. The Committee reached the remarkable conclusion that asthma related admission to hospital is not a relevant end point. Their criticisms of the trial design in the 6–11 year age group were well focussed,7 especially that many had milder asthma and should not have been considered for therapy, but the lack of knowledge of context in the NICE group led them to the wrong conclusion.

However, it is not the main purpose of this editorial to whinge, but to suggest a constructive way forward. Clearly people with generic skills in the evaluation of trial design, healthcare economics and qualitative research are essential in the evaluation of any application to NICE. We suggest that for such specialist areas, a group of experts in the relevant field should be convened. The Royal College of Paediatrics and Child Health Specialist Advisory Committees would be an obvious source to nominate several experts (not just one!). The remit of such a group should be wider than just defining who should be eligible for an expensive medicine. The depth of knowledge and experience of the group should guide those less experienced to fully understand the clinical situation where the drug is being considered.

What would have been an appropriate chain of reasoning had such a group been convened? Paediatricians have rightly campaigned for medicines to be tested in appropriate age-groups, because there may be differences in the nature of the disease, and also different pharmacokinetic and safety issues. The characteristics of severe asthma in school children8,–,11 and adults12,–,14 are the same, and are first detectable at around 3 years of age.15 From 12 years and above, there is ample evidence of efficacy for omalizumab,16,–,19 so safety and pharmacokinetic data are most relevant in children down to age 6 years. The evidence in the 6–11 group comes from a randomised placebo controlled trial in more than 500 children.7 There was efficacy in the primary end point (reduction in exacerbations) and safety. The trial could be criticised on the grounds that children on low doses of ICS were recruited. Most, however, were using at least one additional medication (long-acting β-2 agonist or leukotriene receptor antagonist), and the BADGER study suggests that for most children the plateau of the dose response curve for fluticasone propionate is as low as 200 µg/day,20 so even despite these caveats, the study is important. Taken in conjunction with work in older children, it is difficult to see how the use of omalizumab should not be extended to appropriate children as young as 6 years of age.

The next step would be to define specifically both eligibility criteria and what should be mandatory before omalizumab can be considered. Eligibility might be repeated life threatening exacerbations, or inability to wean the child off oral corticosteroids. Obvious examples of mandatory evaluation would be documentation of adequate inhaler technique, assessment of adherence by accessing prescription records, a home visit, minimising exposure to relevant allergens and documenting with cotinine measurements that tobacco smoke exposure has been minimised. The group should be equally specific about the monitoring during the clinical trial of therapy, and what will be deemed to constitute an acceptable clinical response. The definition of these precise protocols would mean restricting prescription to specialists who can implement them, and, importantly, allow the prospective gathering of data to determine what is predictive of benefit.

The final remit for the group would be to try to define what randomised controlled trials are needed in the future, and to work with the Medicines for Children Research Network and National Institute for Health Research to secure funding. In the case of omalizumab, entry criteria would be children at Step 5 who have been carefully screened to ensure they really do not respond to standard therapy. The current NICE suggestion of a head to head comparison between prednisolone and omalizumab is untenable. The side-effect profiles are not comparable, and corticosteroids could not ethically be withheld if they were needed. Far better would be a multicentre dose reduction study, in children prescribed either very high dose ICS or oral prednisolone. The expert group could collaborate with others to see if there are enough eligible children within the UK to ensure that any proposed studies are feasible.

In summary, we believe that the NICE procedures should be improved. Instead of the adversarial positions that NICE and specialists currently occupy, would it not be better to work together, to ensure that children who need expensive treatment receive it, that scarce resources are not frittered away on those who for whatever reason are not prepared to get the basics right, and that data are gathered prospectively in order to refine future decision making?