Our editorial board answer questions on the NICE guidance on zanamivir and inclusion of targets in guidelines
Q NICE has advised the use of zanamivir only when the number of cases of influenza rises above 50 per 100 000 of the population. What intellectual basis is there for dictating an individual's treatment on the basis of how many other people have the disease?
A Influenza is considered to have risen above its baseline rate, and thus to be 'circulating in the community', when consultations for influenza or influenza-like illnesses rise above 50 per 100 000 population per week; above 400 consultations per 100 000 population per week, it is considered to be epidemic.
However, these definitions do not provide an obvious logical basis for the guidance. The answer to the question is not straightforward, but possible reasons include:
- It is difficult to distinguish influenza from influenza-like illness, and an increase in consultation rates increases the chance that a characteristic symptom complex is influenza rather than influenza-like illness. This will improve the effectiveness of the drug (which has no effect on viruses causing influenza-like illness) and hence its cost-effectiveness.
- As the number of cases of influenza rises, the chance of an (unimmunised) at-risk person contracting influenza rises. Since this may have serious consequences, the predicted volume of offset resource use rises, reducing the net cost, and thus increasing cost-effectiveness.
- Zanamivir prevents both uptake and release of virions for influenza A and B virus particles, and can thus be expected to reduce the time period over which someone suffering from influenza will be infectious to others. It may therefore help others to avoid contracting influenza: again, this improves the cost-effectiveness figures.
- Zanamavir's product licence allows its use in those aged 12 and over who present with symptoms typical of influenza when influenza is circulating in the community. It would doubtless have caused difficulty had NICE recommended its use on a different basis to that described in the product licence.
Almost any threshold value for initiating treatment is artificial and has only a relative rather than absolute value. In the circumstances, this aspect of the guidance seems reasonable.
Q Can guidelines be used to define targets for good practice? If so, is there a danger of them becoming implements of policing?
A A guideline can certainly include targets, but probably should not. Targets are a tool for managing one's own performance or the performance of others. Guidelines, on the other hand, should help a clinician decide what is best for a patient or a group of patients.
Clinical performance management focuses on the degree to which patients are offered appropriate treatment. It accepts that this will not happen 100% of the time. It sets targets that represent an achievable improvement over the current situation, and reviews those targets as initial improvements are achieved.
If it is a human and humane performance management system, it seeks to understand why targets may not have been met through no fault of the individuals concerned, and may set targets at different levels to take account of socioeconomic factors, the availability of buildings and equipment etc.
Guidelines have to be realistic, but should not go into these issues: this would make them management documents rather than clinical decision-support tools.
Q Guidelines are essentially formalised clinical pathways that balance risk against resource. Should we be wary of importing guidelines from other countries with resource levels that differ significantly from our own?
A There are many reasons why guidelines from other countries should not be imported into the UK without careful examination of the cultural, legal, financial and organisational environment within which they were created. It is therefore useful to separate:
- Evidence of effectiveness (which should be generalisable to all comparable populations)
- Evidence of costs (which vary substantially within as well as between countries)
- Decisions about healthcare priorities and the threshold of affordability, which depends on social factors such as local epidemiology, cultural views of illness, health service organisation, current spending on health, and willingness of citizens to pay taxes.
If a guideline is written so that these strands are separate, we should be able to import work from other countries on evidence of effectiveness: this is, after all, the basis of the work of the Cochrane Collaboration, international cooperation on health technology assessment, and studies published in international peer-reviewed publications.
Other aspects have to be examined in the light of local circumstances: this is true when applying national UK guidelines to each and every part of the UK, as well as applying international guidelines locally.
We should not reject a guideline produced in another country because it was 'not invented here', but nor should we be uncritical in the way in which the recommendations, as opposed to the evidence, are appropriate to our own practice.
As more patients obtain information from the internet, we will have to explain that these 'environmental' factors make guidelines much more than absolute scientific statements.