Affording expensive new drugs

Martin Duerden BMedSci, MRCGP, DRCOG, DipTher, DPH

Often lost in the indignant protests over the NHS not funding the latest ‘wonder drug’ is that such drugs can be eye-wateringly expensive. The latest furore has been over NICE’s provisional determination on Kadcyla (trastuzumab emtansine) – the view from NICE was that this drug is not cost-effective unless the drug company reduces the price, which they have so far declined to do.

Kadcyla
Kadcyla is a combination of the biologic trastuzumab (also known as Herceptin) conjugated with a cytotoxic agent, mertansine. The theory behind this is that trastuzumab links onto cell receptors in certain kinds of breast cancer and delivers the cytotoxic drug to the cell. The drug costs £90 000 for a course of treatment and can potentially extend life by six months compared to another standard second-line treatment in breast cancer.

It is important to recognise when looking at median survival figures that 50 per cent of people in the drug trial did not get this benefit, and 50 per cent got more. Also drug trials present an estimate of clinical efficacy from a selected group of patients that might not wholly translate to clinical practice (clinical effectiveness).

The incremental cost-effectiveness ratio was estimated in the NICE Technology Appraisal for Kadcyla to be £185 600 per quality adjusted life year (QALY) compared to standard treatment.

The Cancer Drugs Fund
All this begs a very reasonable question: should the NHS pay for a drug regardless of how much it costs? When the current government came into power in 2010 it set up a £200 million annual Cancer Drugs Fund (CDF) in England to pay for cancer drugs where either value was as yet uncertain or which NICE has deemed not cost-effective. The simple reason for this fund seemed to be that politicians found it too challenging to deny patients expensive cancer treatments claimed to be ‘life saving’ or innovative, even if the price set by drug companies appeared too high. Possibly, they also wanted to keep drug companies ‘sweet’.

It was anticipated that both the CDF and drug prices set by the drug industry under the elderly Pharmaceutical Price Regulation Scheme (PPRS) would be replaced in 2014 by a negotiated system of drug cost reimbursement called value-based pricing (VBP). After considerable debate about how this might work, and how it might not, VBP never happened and was quietly shelved in 2013. Instead, a new PPRS emerged for 2014 and the CDF lives on.

A new evaluation process
There remain many questions about how rational or fair the CDF system is. Why should drugs for cancer treatment in people dying of cancer (they are rarely life-saving) get priority over people who need noncancer treatment, for example for hepatitis C? Why shouldn’t the NHS negotiate over value for money? Shouldn’t austerity measures apply to drug prices? What else of greater value could the money be better used for? At present the CDF pays for Kadcyla in England, and it has to be said that the role of NICE in determining cost-effectiveness of cancer drugs for the NHS for England and Wales is undermined.

The intent is that NICE will now develop a scheme that enables a wider assessment of drug value than the QALY-based system used by them for Kadcyla. This is called value-based assessment, presumably to indicate its origins in the concepts of value-based pricing. How this will operate is not yet clear, but alongside the QALY it is likely to consider ‘burden of illness’ to give additional weight for loss of healthy life, and ‘wider societal impact’ looking at the impact of disease on function within society rather than the direct effect on health. If this description seems rather vague, how this will translate into the evaluation process remains woolly. NICE are out for consultation on this new appraisal process until 20 June and hopefully after this clarity will emerge.

Declaration of interests
None to declare.

Dr Duerden is a part-time GP in North Wales and clinical senior lecturer, Centre for Health Economics and Medicines Evaluation, Bangor University. He sits on a NICE Technology Appraisal Committee.

Join the discussion below…

NICE CG178 on psychosis – an evidence-based guideline?

Udayanga Perera MB BS and Mark Taylor BSc, MD, FRCPsych, FRANZCP

NICE is renowned for producing impartial and evidence-based clinical guidelines, with a rigorous development process leading to reliable and cost-effective recommendations. NICE recommendations can have far reaching implications on policy at regional and national levels. Despite this, NICE has been associated with controversy, eg the restriction of acetylcholinesterase inhibitors in moderate severity Alzheimer’s.¹

In this editorial we summarise and critique NICE clinical guideline on Psychosis and Schizophrenia in Adults: Treatment and Management (GC178).

Psychosis or schizophrenia?
NICE CG178 replaces the previous 2009 title of ‘schizophrenia’ with ‘psychosis and schizophrenia’. In the introduction NICE defines ‘psychosis’ as a group of psychotic disorders that include schizoaffective disorder, schizophreniform disorder and delusional disorder. However, the term ‘psychosis’ is not found within either Diagnostic and Statistical Manual of Mental Disorders (DSM) or International Classification of Diseases (ICD) diagnostic manuals, and in some services includes bipolar I disorder. This vagueness could be misleading. Furthermore, even a casual glance through the guideline confirms the impression that it concentrates on psychosocial interventions, as though cognitive behavioural therapy (CBT) and family interventions are a panacea across all stages of the illness.

CBT as a panacea?
Comparing the evidence head to head for CBT and family intervention against pharmacotherapy for schizophrenia is beyond the scope of this article. But there is little doubt that the advent of antipsychotics in 1952 had an immense impact on the lives of people with schizophrenia. For example, a population-based study² concluded that long-term treatment with antipsychotic drugs is associated with lower mortality compared to cases where no antipsychotic was used.

There has been increasing interest in the last three decades in CBT as an adjunct to antipsychotic medication in treatment of schizophrenia. NICE draws its recommendations on CBT based on 31 RCTs (N=3052) of CBT versus any type of control. Based on this and another review on CBT for (only) those at risk for psychosis,³ NICE recommends CBT for all people with schizophrenia or psychosis at all stages of the illness, including those at risk for psychosis. A recent larger meta-analysis concluded that CBT has a small therapeutic effect on schizophrenia symptoms and even this small effect is reduced when sources of bias, particularly masking, are controlled for.⁴

NICE has also taken the bold step of recommending CBT and family therapy alone for people with first-episode psychosis who wish it. The guideline acknowledges that psychosocial interventions are more effective in conjunction with antipsychotic medication, but still suggests this intervention alone for one month or less. This is controversial in view of the lack of robust supportive evidence and could potentially worsen outcomes. A related point is that in the guideline NICE seem oblivious to the fact that many patients with acute schizophrenia have impaired insight into their illness and health needs,⁵ and thus may not have capacity to consent to their treatment.

Failure to offer the most evidence-based treatments promptly could be viewed as breaching the duty of care by the practitioner. It is also interesting that NICE seems to have ignored patients who refuse psychotherapy and makes no recommendations on offering medication alone anywhere in the guideline.

The bias towards psychosocial interventions seems mostly based on the premise that antipsychotics are harmful. But it is vital to keep in mind that medication-related adverse effects come to light after extensive research and experience over a long duration. Therefore absence of evidence for adverse effects of psychosocial interventions should not be taken as evidence of absence, as it has been tested less rigorously. Possible adverse effects include stigma associated with prolonged psychotherapy and possible deterioration of mental state due to over-stimulation especially in taxing interventions like CBT. CBT is also costly, time consuming and is not readily available. Its effectiveness depends largely on the skill of the therapist and its fidelity and quality can be difficult to evaluate.

Are all antipsychotics the same?
Out of 110 recommendations in the guideline only 27 (24 per cent) are reserved for medication and most of these are coupled with recommendations on offering CBT and/or family intervention to all patients. It is interesting to note that a contemporary and equally evidence-based guideline, Scottish Intercollegiate Guideline Network (SIGN) 131, has 60 per cent of its recommendations devoted to pharmacological interventions alone.⁶

NICE CG178 makes commendable recommendations on monitoring for medication-related adverse effects, highlighting the differences between antipsychotics. NICE also implies the conventional view that all antipsychotics are equal in efficacy. By way of comparison SIGN 131 advocates considering amisulpride, olanzapine or risperidone (Risperdal) as the preferred medications in the acute phase. This is in keeping with evidence that these medications are more efficacious for acute schizophrenia.⁷

Also, NICE CG178 recommendations do not reflect the differences required in medication dosage in the first episode and the subsequent episodes. On depot or long-acting injectable antipsychotics (LAI) NICE gives a low strength recommendation and says to ‘consider offering’ LAIs. This is not in keeping with evidence favouring LAIs over oral antipsychotics.^8,9

Equally NICE has overlooked the need for a loading dose for the recently introduced LAI paliperidone palmitate (Xeplion), carelessly stating: ‘do not use loading doses of antipsychotic medication’.¹⁰

Conflict of interest?
In our view NICE CG178 promotes some psychosocial interventions, especially CBT, beyond the evidence. NICE CG178 also makes strong nonevidence-based recommendations, for instance that the course of CBT should be at least 16 planned sessions. The guideline’s research recommendations also appear to reflect the interests of the authors. Here they also note an open trial from the Netherlands reporting successful discontinuation of medication in 20 per cent of patients, which merely confirms the well-known fact that about 20 per cent of people who have an acute episode of schizophrenia recover completely.

Conclusion
NICE occupies an important position – it can validate ‘gold-standard’ treatment paradigms that can impact on policy making and research agendas. NICE CG178 should be constrained by the limits of the available evidence and it is unfortunate that the guideline appears at times to reflect the interests of its authors rather than impartial up-to-date evidence.

References
1. Thacker S, et al. The Psychiatric Bulletin 2008;32:32.
2. Tiihonen J, et al. Lancet 2009:374(9690);620–7.
3. Stafford MR, et al. BMJ 2013;346:f185.
4. Jauhar S, et al. Br J Psych 2014;204(1):20–9.
5. WHO. Report of the International Pilot Study of Schizophrenia. Volume 1. 1973.
6. Scottish Intercollegiate Guidelines Network. Management of schizophrenia. SIGN 131. March 2013.
7. Leucht S, et al. Lancet 2009;373(9657):31–41.
8. Leucht S, et al. Lancet 2012;379(9831):2063–71.
9. Tiihonen J, et al. Am J Psychiatry 2011;168:603–9.
10. Taylor D, et al. The Maudsley prescribing guidelines in psychiatry. 11th edition. Wiley, 2009.

Declaration of interests
Dr Perera has none to declare. Professor Taylor was co-chair of SIGN guideline 131 and has accepted fees or hospitality from Janssen and Lundbeck in the last three years.

Dr Perera and Professor Taylor both work in Metro South Brisbane, Australia

Join the discussion below…

Polypharmacy – essential, but also to be avoided

Martin Duerden BMedSci, MRCGP, DRCOG, DipTher, DPH

Polypharmacy is not new – it was first described and discussed in the medical literature at least 150 years ago. Despite this, other than the dictionary definition of prescribing more than one drug at the same time to an individual, it is has been poorly characterised and under-researched.

When I started training in medicine, a long time ago, the medical profession was snooty about polypharmacy: it was considered a sign of poor prescribing and a failure of common sense. In some instances this is still the case, but things have changed and for an increasing number of patients it is considered good practice to use all evidence-based medicines deemed necessary to improve outcome. In the UK, the Quality and Outcomes Framework (QOF) rewards GPs to do this.

King’s Fund report
With my colleagues Tony Avery from Nottingham and Rupert Payne from Cambridge, we set about reviewing the literature on polypharmacy and how best to optimise medicines used in this context for a King’s Fund report (also see pages 26–8 of this issue).1

After a lot of pontificating and discussion with many colleagues we separated polypharmacy into two components – appropriate polypharmacy and problematic polypharmacy.
In the report appropriate polypharmacy is defined as prescribing for an individual for complex conditions or for multiple conditions in circumstances where medicines use has been optimised and where the medicines are prescribed according to best evidence.

Problematic polypharmacy is defined as the prescribing of multiple medications inappropriately, or where the intended benefit of the medication is not realised. These problems increase as drug regimens become more complex. The reality is that people often do not take medicines in the way that prescribers intend and there is considerable evidence that many dispensed medicines remain unused or are wasted.

The numbers of interventions have mushroomed in the last few decades, the population is living longer with chronic conditions and the number of treatable conditions an individual carries has increased. In the real world it is more common to have more than one long-term condition than it is to only have one, but a paradox is that most of the evidence is gleaned from randomised, controlled trials where patients are carefully selected to only have one, and who are unlikely to be frail.

In this context it is not clear that piling more pills into an individual will give the benefits seen in clinical trials and this could be harmful by increasing the risk of adverse drug reactions and interactions, together with impairing quality of life for patients; the pill burden itself may be distressing. The more frail the patient the more likely they are to suffer such harm. How do we resolve this?

Suggested remedies
Many suggestions are obvious but some are worth stating explicitly. For example, it is clear that we need research that is more pragmatic and assesses outcome in the context of multimorbidity and multiple medicines. Alongside this we need guidelines that can describe the best treatment options where diseases commonly overlap, eg diabetes, coronary heart disease, heart failure and COPD. NICE has started to look at this.

When initiating a treatment the rationale for choosing it should be clearly stated. A tip to enhance understanding and adherence is to describe the purpose of the drug on the prescription, eg ‘ramipril 10mg daily to lower the blood pressure and help the heart’. It also makes sense for prescribers to describe in the medical record the reasons to step aside from the evidence, particularly if the patient chooses not to take a treatment.

Multimorbidity and polypharmacy increase clinical workload. Doctors, nurses and pharmacists need to work coherently in teams to tackle medicines optimisation. Our view is that we all need more training in managing complex multimorbidity, polypharmacy and other aspects of medicines management.

Alongside this we need to develop systems that optimise medicines use so that people gain maximum benefit from their medication with the least harm and waste, for example through improved electronic decision support for clinicians and patients, better patient-friendly information systems and judicious use of monitored-dose systems.

If polypharmacy is such a big problem, and we believe it is, where do we start? In the report we suggested some thresholds after reviewing the literature. A pragmatic approach is to focus resources on patients at particularly high risk: for example, those receiving 10 or more regular medicines, or those receiving four to nine regular medicines together with other unfavourable factors (examples of the latter include a contraindicated drug, where there is potential for drug-drug interaction, or where medicine taking has proved a problem in the past).

Knowing when to stop
Perhaps the toughest challenge is deciding when to stop. Many people stay on medicines beyond the point where they are deriving optimal benefit from an intervention. I recently encountered a moribund patient being administered a statin via a feeding tube. When reviewing medications, always consider if treatment can be stopped and recognise that ‘end-of-life’ considerations apply to many chronic diseases as well as cancer-related conditions.

Reference
1. Duerden M. Polypharmacy and medicines optimisation: making it safe and sound. King’s Fund report, 2013, from: http://www.kingsfund.org.uk/publications/polypharmacy-and-medicines-optimisation [accessed 27 January 2014].
See also the associated feature article in the BMJ and subsequent responses. Wise J. Polypharmacy: a necessary evil. BMJ 2013;347: f7033 from: http://www.bmj.com/content/347/bmj.f7033 [accessed 27 January 2014].

Declaration of interests
None to declare.

Dr Duerden is a part-time GP and deputy medical director for Betsi Cadwaladr University Health Board, North Wales, and clinical senior lecturer, Bangor University

Join the discussion below…

Introduction of biosimilars: not to be confused with generics

Jonathan Mason MSc, MPhil, DIC, DUniv, FRPharmS

Generic versions of branded medicines have been with us for a long time – there was a proliferation of generic versions of aspirin, for example, when Bayer’s patent expired in 1917.¹ We will all be familiar with the process through which generics are introduced to the market: the originator brand product has a finite patent life, at the end of which other manufacturers can apply for a licence to manufacture and market a generic version.

Production of generic medicines is reasonably straightforward for synthetic chemical drugs, in which the active substance in the generic medicine is identical to the active substance within the originator brand. While there may be some small differences with respect to excipients used in generic formulations, to all intents and purposes generic medicines are the same as the original brand.

So when is a generic not a generic?
A generic isn’t a generic when it’s a biological drug. Biological drugs (biologics) are, generally, large molecules based on sugars, proteins, nucleic acids or complex combinations of these substances, and are produced through complex biotechnological processes rather than being chemically synthesised.

Since the follow-on manufacturer does not have access to the originator’s molecular clone bank and original cell bank, they cannot guarantee that their version is exactly identical to the original manufacturer’s version (known as the biological reference medicine).

Unlike synthetic chemical drugs, there may be significant differences between the biosimilar and the biological reference medicine in terms of efficacy or adverse effects.² In addition, since biologics are administered parenterally, biosimilars may require different delivery systems from the reference medicine.

In the EU a specially adapted approval procedure has been introduced for certain biologics, termed similar biological med­icinal products – shortened to biosimilars. This procedure is based on a thorough demonstration of ‘comparability’ of the biosimilar to the biological reference medicine. This procedure is more rigorous than for small-molecule drugs proven to be identical to each other, but less testing than for completely new drugs.

Unlike generics, new biosimilars have black triangle status at the time of initial marketing. It is therefore important to report all suspected adverse reactions to biosimilars using the Yellow Card Scheme (refer to the Medicines and Healthcare products Regulatory Agency website). For biosimilars, adverse reaction reports should clearly state the brand name and the batch number of the suspected medicine.

Implications for health systems
The health systems in the UK have a long history of successful introduction of generics. Pharmaceutical companies know that they have a limited window within which to maximise profits from their products before patent expiry, after which generics will almost inevitably be introduced at much lower prices, thereby providing the NHS with windfall savings.

Those of us who work within the NHS as well as those in the pharmaceutical industry all know the game. The rules may have changed slightly over the years, but the basics are the same: NHS advisers will push for maximum use of generics, while industry will push the claims of superiority for their branded products. Generally the first, or in many cases the second, drug of a new class will establish itself as class leader and any subsequent drug in the same class will be viewed as a ‘me-too’ and may struggle to establish anything other than a small niche in the NHS market.

Later entrants to the market will offer lower prices to try to compete with the market share of the first couple of drugs, often with minimal success, although in some cases improved clinical effectiveness may enable later entrants to establish a larger market share.

Use of generics provides benefits to patients and the NHS, not least of which is the fact that the lower price increases overall access to effective medicines. So generic versions of very expensive medicines, such as biologics, would be welcome. Unfortunately, due to the reasons outlined above, true generic versions of biologics will not be produced. We are, therefore, faced with the introduction of a range of biosimilars.

Unlike with true generic medicines, where clinical data for the originator product will hold true for the generic, clinical evidence for each biosimilar will have to be assessed individually. Therefore, it would be inadvisable to switch patients from the agent on which they have been established to a biosimilar in order to save money on acquisition costs. Biosimilars will need to be introduced in a safe and measured way, eg using them in newly diagnosed patients and in patients needing to change therapy.

The good news for patients and the NHS is that manufacturers of biosimilars will seek to maximise benefits of their product and may introduce new administration systems or provide patient support programmes not offered by the originator company, so there may prove to be benefits to both patients and the NHS from the careful introduction of biosimilars.

Conclusion
We need to be cautious and remember that biosimilars are not generics and should be treated as new drugs. Biosimilars may prove to be bio-better or bio-worse: only time will tell for each individual product.

References
1. Jeffreys D. Aspirin: the remarkable story of a wonder drug. Bloomsbury, 2005.
2. Roger SD, et al. J Pharm Sci 2007;10(3):405–10.

Jonathan Mason is clinical adviser (medicines), NHS East London and The City, and pharmacy lead for the Dementia Calls to Action, NHS Institute for Innovation and Improvement

Join the discussion below…

Cardiovascular risk and NSAIDs: prescribing slow to change

Anthony Cox PhD, MRPharmS

NSAIDs expose patients to cardiovascular risk. There is no pharmacological fix other than a decision not to prescribe, or to choose an NSAID with the lowest known risk. The risk is incurred in patients without existing cardiovascular risk factors, even after short-term use.

The recent history of NSAIDs serves as an example of just how hard it can be to influence prescribing attitudes in the face of evidence that clearly indicates harm.

Cardiovascular safety
It was following the voluntary withdrawal of the COX-2 inhibitor rofecoxib in 2004 that the cardiovascular safety of NSAIDs as a whole came under scrutiny. Aggressively marketed as a safety breakthrough in terms of gastrointestinal events as a result of its lack of inhibition of COX-1, rofecoxib was later found in clinical trials to cause cardiovascular events.¹ The NSAIDs were already spread along a continuum of harm with respect to their gastrointestinal toxicity, but similar differences in the cardiovascular risk associated with individual NSAIDs began to emerge as systematic reviews and meta-analyses were performed.²

By 2005 the Medicines and Healthcare products Regulatory Agency was advising on the increased thrombotic risk associated with nonselective NSAIDs, noting that diclofenac’s risk was possibly similar to that of the coxibs. A 2006 European review again highlighted diclofenac’s coxib-like risk.³ Both naproxen and ibuprofen carry a lower cardiovascular risk, with evidence suggesting naproxen’s effect on cardiovascular events is neutral.

Evidence into practice
How much careful consideration of this evidence is apparent in prescribing data? Not enough seems to be the answer. A recent paper usefully examines how this evidence translated into action within 15 countries.⁴ Taking sales and prescription data of NSAIDs in 2011, as a percentage of total sales the high-risk diclofenac accounted for 28 per cent while the relatively low-risk naproxen accounted for less than 10 per cent. Off-evidence prescribing is very common.

While the increased cardiovascular risk to an individual patient may be low, this risk accumulates within the population causing a significant public health issue. Patients may accept the risk of NSAIDs for the symptomatic relief they can provide, but they should be aware of the differing risks they accept with individual drugs.

From 2007 to 2009 the then National Prescribing Centre undertook a large educational intervention including therapeutic bulletins, e-learning materials and therapeutic workshops to influence NSAID prescribing. A time-series analysis of primary-care prescribing data showed a small decrease in total NSAID use, but more importantly a significant fall in diclofenac prescribing and a proportionate increase in the use of naproxen and ibuprofen.⁵ However, there was considerable variation between PCTs, and diclofenac continued to account for 39 per cent of prescribed NSAIDs (naproxen 14 per cent).

Continued resistance to the evidence on NSAIDs was ascribed to the ingrained intuitive decision-making processes prescribers employ. Synthesis of new evidence into practice is not easy, requiring more long-term efforts to instil new evidence into normal practice.

Despite these praiseworthy efforts, there is continued therapeutic inertia on diclofenac. A regulatory solution could be the outcome in Europe. A new NSAID would be unlikely to overcome regulatory hurdles with a cardiovascular risk similar to rofecoxib, especially given the lower-risk alternatives on the market already. Why should an older drug be any different purely because of its history?

Conclusion
Drug withdrawals are not without cost for patients. After the withdrawal of the NSAID benoxaprofen in the 1980s, following cases of hepatotoxicity, the playwright Dennis Potter complained he had been left ‘high and dry’. A proportion of patients taking diclofenac might react similarly if shifted to an alternative.

More recent prescribing data may show diclofenac losing pole position to naproxen, but given its continued high proportion of total NSAID prescribing, relying on influencing prescribers may be too optimistic a policy. Evidence of compliance with regulator warnings is mixed,⁶ and some form of restriction on availability may be the future.

References
1. Bresalier RS, et al. N Engl J Med 2005;352:1092–102.
2. Singh S, et al. Arthritis Res Ther 2006;8:R153.
3. MHRA. MHRA public assessment report: non-steroidal anti-inflammatory drugs and cardiovascular risks in the general population. January 2010. http://bit.ly/144XRYC.
4. McGettigan P, et al. PLoS Med 2013;10:e1001388.
5. Hollingbee D, et al. Prescriber 2011;22(7):45–8.
6. Ricci JR, et al. Am J Manag Care 2009;15(11):e103–8.

Declaration of interests
None to declare.

Anthony Cox is lecturer in clinical pharmacy, University of Birmingham

Do you agree with Anthony Cox? Join the discussion below…

The NICE Innovation Scorecard – a poorly thought through idea

Martin Duerden BMedSci, MRCGP, DRCOG, DipTher, DPH

Innovation is used to describe the introduction of something new or different and normally implies that it is better. A scorecard usually relates to a system for keeping score of a contest and identifying or ranking attainment by the participants.

In August 2012 it was announced that in England each primary-care organisation and hospital provider is to be measured according to its rapid uptake of treatments that have been recommended by NICE. One assumes this chiefly relates to technology appraisals where there is a mandate to fund treatment in the NHS in England and Wales within three months. The scorecard is to be made available on the internet for all to see in autumn (although as everyone knows, autumn in the NHS usually stretches into the following year). Hospitals and commissioning bodies will also be on lists that show which of the latest NICE-approved treatments and drugs are available in local areas – in effect a contest.

Does the NHS stymie innovation?
Why is this naming and shaming necessary? The driver for this appears to stem from the belief that cash-strapped NHS bodies are delaying using NICE-approved treatments or even not funding them at all. Professor Sir Mike Rawlins, the chair of NICE, made it very clear that he had examples of this behaviour in an article in the Health Services Journal and interviews with the media in August. The Department of Health in England says that under the scheme, ‘the NHS will have no excuse not to provide the latest NICE-approved drugs and treatments’ and that it hopes the proposals ‘will bring an end to “postcode prescribing” whereby treatments and drugs are available immediately in some parts of the country, and yet are delayed in others’. Alongside this initiative, NICE say the scheme will see ‘a group’ set up to help local NHS organisations implement NICE guidelines, in an effort to speed up the update of new drugs and treatments.

Healthy drug industry or healthy NHS?
There are some important concerns about Innovation Scorecards. It is a policy driven by politics and such political direction can create conflict in priorities. The Department of Health has a role in supporting the drug industry in the UK and it is clear government policy to drive adoption of innovation in the NHS. Not surprisingly this process has the full support of the Association of British Pharmaceutical Industry who continually raise concerns about the delay in use of newly licensed drugs in the UK compared to other European countries. Thus the drug industry is supported to do well financially at a time of austerity in the NHS.

Innovation in technology vs service development and local priorities
Many NHS bodies are struggling to provide standard NHS services with no growth in funding alongside an NHS system creaking at the seams with built in inflation, particularly as technology becomes more sophisticated and the range of interventions available is rapidly growing. Money spent on one treatment means less available for other purposes; to preferentially fund an expensive intervention for one patient may be to deny other patients another. An example might be an innovative cancer drug used in a third-line setting to extend one life by a few months compared with funding palliative-care services for a number of patients; there may be a NICE mandate to fund the former but elements of palliative-care services have not been subject to NICE technology appraisals.

At a time when Clinical Commissioning Groups in England are trying to set local priorities they are shackled to a system that requires in-year funding, which will use up important reserves. This might be reasonable if NICE guidance was always right and new treatments were truly innovative and better. There are clear examples of such innovation, for example, statins in cardiovascular disease and biological therapies in rheumatoid arthritis. However, many so-called innovations signify small incremental gain over previous or standard treatment at considerable expense measured by the incremental cost-effectiveness ratio using costs/quality-adjusted life years (QALY). The reality is that we pay a large amount for drugs that do not necessarily provide value for money. Hopefully the advent of value-based pricing for new drugs in 2014 may address these concerns.

Maturing technologies alongside immature health economics
A problem with NICE technology appraisal guidance is lack of dynamism in the process as an estimate of cost effectiveness is made at a single point in time in the infancy of a drug. This ‘point estimate’ of cost/QALY is based on what is known about benefits and safety at that stage. In reality the adverse event profile for ‘innovative’ drugs gradually develops over several years of use, usually garnered through pharmacovigilance. Often drugs develop monitoring requirements that had not been known at the time of the health economic analysis, or potential harms emerge that have significant costs attached. This was recently seen both with dronedarone (Multaq) and dabigatran (Pradaxa) where there had been immense pressure to approve these drugs from the industry, clinicians and patients but where now there are significant safety concerns.

Giving the local NHS an option
Another criticism is that NICE often approves a treatment as an option for use where there are other alternatives. This still requires mandatory funding but surely local formulary committees should be able to decide whether this is their preferred option? The Quality Standards set by NICE, largely based on their clinical guidelines are laudable, but Innovation Scorecards may profit the drug industry at the expense of the NHS.

Declaration of interests
None to declare.

Dr Duerden is a GP and deputy medical director for Betsi Cadwaladr University Health Board, North Wales, and honorary senior lecturer at Cardiff and Bangor universities. He sits on a NICE Technology Appraisal Committee.

Vitamin D: deficiency in guidance?

Martin Duerden BMedSci, MRCGP, DRCOG, DipTher, DPH

In February 2012 the four chief medical officers (CMOs) for England, Scotland, Northern Ireland and Wales wrote to GPs, practice nurses, health visitors and community pharmacists recommending that, as a quarter of the UK population are deficient in vitamin D, at-risk groups should be taking vitamin D supplements.

They advised that all pregnant and breastfeeding women and all children between the ages of six months and five years should take a daily supplement. Breastfed babies should be supplemented from one month if the mother has not taken vitamin D supplements throughout pregnancy. Bottle-fed babies may not need this if they consume more than 500ml a day of formula milk as this is already fortified. Furthermore people aged over 65 years ‘who are not exposed to much sun’ should also take a daily supplement. Some of these groups should already receive Healthy Start vitamins if they receive income support or certain other benefits.

Defining deficiency and evidence of harm
Such guidance begs a number of questions. The first of these is how vitamin D deficiency is defined. In our temperate climate it has been claimed that more than 65 per cent of us have either ‘insufficiency’ or deficiency in our vitamin D levels, but it appears that no clear definition exists based on blood levels of vitamin D. This is rather like saying over half the population have hypertension depending on the level of blood pressure chosen to define ‘abnormality’: the norm become abnormal.

Without doubt there is evidence of harm from vitamin D deficiency, in particular rickets and osteomalacia in urban areas where there are high Asian populations. This is not commonplace elsewhere. Surely vitamin D deficiency should be determined by evidence of harm? Is there strong evidence that supplementation in the wholesale fashion recommended results in more benefit than harm? Historically there is evidence of harm in the form of hypercalcaemia from oversupplementation and food fortification from the 1940s and 1950s.

Providing vitamin D supplements
A big challenge for prescribers in the UK is how to go about providing these supplements. This is not described in the CMOs’ guidance. There is a paucity of products that are licensed at the right dose for this purpose. Healthy Start vitamins are in short supply and reserved for the defined groups receiving state benefits. Unless the recommended recipients are prepared to buy (unlicensed) vitamin products from shops and pharmacies, the burden of supply lies through NHS prescriptions. There are at least five problems with this.

The first is that those most likely to benefit are those least likely to buy supplements. The second is that prescribing suitable vitamin D preparations in the volumes required could prove costly at a time when NHS money is in short supply. The third is that, in the absence of a clear NHS policy on appropriate provision, the inclination is to prescribe or supply unlicensed products that have not been quality assured to the same high standards of drugs that are licensed as medicines. There are case reports of harm resulting from excessive vitamin D in some over-the-counter products. The licensing authorities advise care when prescribing in this way and this risk may be more of an issue in pregnancy and in babies, where there is greater vulnerability.

Another problem is that vitamin D drops for children can contain arachis (peanut) or soya oil with associated risks in those with peanut or soya allergy. The final problem is the temptation in older people to prescribe calcium and vitamin D combination products, but recently there has been considerable concern about the increased risks of cardiovascular disease in people taking calcium.

Mass exposure
Probably the point at which this deficiency in guidance seems most absurd is the associated contradictory public health message that sunlight is bad for us and the advice to ‘cover up in the sun’, particularly children. Most fair-skinned people can boost their vitamin D levels by short exposure to sunlight (eg 15 minutes at a time without sunscreen) without running the risk of skin cancer. Surely this is a better public health solution than mass medication?

Declaration of interests
None declared.

Dr Duerden is a GP and deputy medical director for Betsi Cadwaladr University Health Board, North Wales, and honorary senior lecturer at Cardiff and Bangor universities