Compromise and Conceit

In a previous post I noted some reasons (that I hope are largely practical rather than ideological) why the Tory reforms of the NHS won’t work, based on my superficial understanding of them. Yesterday I discovered that the Tories have introduced a second – and in my opinion completely mad – reform of the NHS that (it appears) will do the exact opposite of what every sane policy reform in modern healthcare aims to do[1]. This new reform will undermine one of the central planks of any universal health system. In this post I’ll give my opinion of the new reform, explain why I think it won’t work, and describe what I think is a key part of modern healthcare systems, be they private or public.

This post is very long…

The latest reform

In a nutshell, the latest reform is a decision by the government to remove the power of the National Institute for Health and Clinical Excellence (NICE), to refuse approval to market new drugs. NICE used to have the power to ban drugs on the basis of their cost-effectiveness rather than their clinical utility, which makes it a different body to, say, the FDA in the US. However, this role is very important in the control of modern healthcare costs. The Tory decision essentially hobbles the ability of the NHS to control costs and gives the pharmaceutical industry carte blanche to market overpriced and ineffective drugs to desperate people. It also represents the culmination of a 10 or 15 year campaign by the pharmaceutical industry to break down the powers of organizations like the Australian Pharmaceutical Benefits Advisory Council (PBAC) and NICE.

What NICE Did

NICE was charged with assessing new therapies, drugs and interventions for their cost-effectiveness, and recommending whether or not they should be used by the NHS. It is in fact Sarah Palin’s famous death panel, which opponents of Obamacare sometimes claim is responsible for deciding “whether patients live or die.” In fact, it does no such thing – it’s not possible to assess the cost-effectiveness of a new therapy without comparing it to an existing therapy, so NICE can only recommend whether or not a given treatment is worth recommending relative to the existing standard. What actually happens[2] is:

• where there is an existing treatment, whose clinical effectiveness is established, a pharmaceutical company proposes a new, more expensive (usually) therapy
• NICE assesses this therapy, typically by calculating the additional cost of treatment per quality-adjusted life year (QALY) saved[3]
• If the cost per QALY is less than 30,000 pounds (or so) then NICE recommends the treatment be adopted

Note this isn’t 30,000 pounds per life saved, but per QALY. It’s a pretty easy standard for a new treatment to beat. Consider this brief example:

• some drug (statins?) taken annually for the remainder of a person’s life will prevent a repeat heart attack, and the life expectancy of that individual is about 70 on the drug, which costs $100 per year[4]
• On average this drug is taken from the age of 50, so the mean cost is 20×100=$2000
• A new drug (say, kingsfoil) is recommended by Aragorn Inc., that will cost $120 per year but extend life to 80; however, the side effects are such that in a well-designed survey of early-adopters, each year of life saved was valued at 0.7 healthy years. So assuming the same average age of commencement (50), the QALYS saved are 30×0.7=21 years, and the total cost is $120×30=3600. So really we gain an average of 1 QALY at an excess cost of $1600, or 1600/1=$1600 per QALY. NICE would recommend this drug, since the cost per QALY is <$30000
• In fact, for this drug to be not recommended it would have to cost an extra $30000 over 30 years, or $1000 per year, i.e. it would have to cost $1000 more per year than the existing drug, or 10x as much.

It’s worth bearing in mind that my example is for a pretty cheap drug, but there are many treatments that are much more complex than this. Consider the cervical cancer vaccine, which (I think) is about $400 per woman, but which needs to be given to very large numbers of women to prevent one death. The drug is cheap but the QALYs saved are few and far between; but how do you consider the possible elimination of cervical screening in the cost-effectiveness analysis?

In short, NICE’s job is not to tell doctors what treatments they can offer their patients, but to tell them what treatments are publicly funded, and what treatments patients will have to shell out their own cash for. This recommendation is only made when an existing, effective treatment is already in use – to the best of my knowledge NICE has no power to refuse a treatment on cost grounds where no effective treatment already exists[5]. So it is in no sense a “death panel,” unless by “death panel” you mean “a group of people who decide whether you will have to pay yourself for a drug that you will, on average, have to pay $30,000 per year to use, and which is replacing an existing perfectly adequate drug.”

I’ll come back to what that decision (“30,000 per year”) means when we discuss the vulnerability of sick people, and the corpse-crows who prey on them with expensive treatments.

The new policy

The new policy is for NICE to continue to “review new treatments” whatever that means, but to lose the power to refuse registration of treatments on the basis of their cost effectiveness. Instead, the Tories have proposed a system of “value based pricing,” in which the NHS would negotiate directly with drug companies over the price of a new drug. This is what the Australian system does, but the Australian system does it through the decisions of the PBAC, which basically tells the health department what it should pay, and then the health department tells the drug company “this much or fuck off.” This is exactly what happened recently in negotiations with CSL (an Australian company!) over the cervical cancer vaccine – the PBAC said the price that CSL was offering was too high, the health department pointed this out to CSL, and a few months later CSL offered the drug at a lower price.

So it’s possible that the new policy could lead to an equally effective system of price negotiations, except that any such price negotiations require the input of a committee whose advice is treated as final, and based on cost-effectiveness analysis… i.e NICE. So why is the government getting rid of this power of NICE, and why is the  new process going to include considerations such as “how innovative the company has been in producing the drug”[6]?

Some key points about the policy from the Tory website are that it will:

• Introduce a new system of value based pricing, while supporting risk sharing deals with drug companies, meaning that all new clinically effective medicines should be available for clinicians to prescribe.

• Allow NICE to take into account the wider social cost of denying a drug to patients when assessing its value or benefit.

• Shift the decision making power over the supply of new drugs from NICE to front line doctors.

• Shift the burden of proof as to the effectiveness of new drugs from NICE to the manufacturer – the drugs company will have to prove it will work rather than NICE prove it does not.

So they intend to maintain the current procedure of requiring the drug companies to pay for high-cost drugs that don’t work in the patient groups for which the drug is claimed to be effective (“risk-sharing”), but there are two key ideas in these 4 points that are very deceptive:

• take into account the wider social cost of denying a drug to patients: colour me purple if this isn’t a rather extreme example of eco-feminism creeping into the Tory party. Since when was conservative economic and health policy concerned with judging the “wider social cost” of providing or denying medical care? Isn’t this Old Labourism at its core? Who will make this judgement, and how will it be adapted to meet the brutal reality of modern health care cost containment? And how will this accounting be matched to the Tory aims of localism and accountability? This kind of jarring language in a conservative document is cause for suspicion – not because it’s theoretically wrong (it’s a great idea!) but because it doesn’t fit the Tory ideals and it is practically unsupportable madness. The “wider social cost” is only ever assessed in favour of those who have the governing party’s ear – and we’ll see that the people with the governing party’s ear are the pharmaceutical companies, not the rest of the “wider” community.
• Shift the decision-making power… to doctors: This is straight from the Palin playbook, implying that NICE is interfering in individual treatment decisions. This is bullshit. NICE decides what drugs are funded, it’s no more interfering in individual treatment decisions than the government is when it tells a doctor they have to obtain consent for treatment, or that they can’t use homeopathy to treat HIV on the NHS budget[7]. Doctors only ever have clinical decision-making power within the framework that the health system sets, and to pretend that economic decisions are somehow interfering with doctors’ decisions is always propaganda. Doctors’ clinical decisions are always economically constrained – either by what their patients can afford or what the system can afford. If a patient attends a doctor for a free consultation that results in free treatment, no-one can claim that the doctor is able to make decisions unimpeded by economics – we have just shifted the economics away from the individual patient to the system, which usually <i>increases</i> the doctor’s freedom to treat[8]. Given that doctors will always prescribe the best treatment regardless of its marginal cost effectiveness, giving them unfettered power to choose treatment means that the cost of drugs will massively increase, and some other constraint will be put in place to control them.

At the same time the Tories have introduced a 200 million pound cancer drugs budget, which will increase the amount spent on cancer drugs. Removing NICE’s power of recommendation at the same time basically means a bonanza for the drug companies to start selling their most expensive, most ineffective treatments – the ones that they spent lots of money researching for marginal results, and can’t sell under a strong regulatory regime.

The effect of the policy

Based on this information, it seems likely that the “value-based pricing” system that the Tories are introducing is intended to widen the range of high cost, low effectiveness drugs on the market, and to allow for non-economic considerations to enter the decision-making process, including, it would seem, the amount of effort the drug company put into developing the drug, and the “wider social costs” of refusing it. This is going to lead to increased costs of therapies at a time when the NHS is again entering an era of funding constraints and poor future economic growth. Just when the NHS needs every tool at its disposal to contain costs, the very government that has put those funding constraints on it has removed one of its most effective tools.

What this means is that decisions on what drugs to fund will now rest on the individual Primary Care Trusts, who control budgets. This means that the high cost drugs will be most available in the wealthiest PCTs, or that (worse still) poor and badly-run PCTs[9] will make foolish drug purchasing decisions and then be left with funding shortfalls requiring stricter health rationing. This will increase the inequality in health between these areas, where of course the wealthiest PCTs already have the best health outcomes. As Maynard observed in the Guardian, rationing is inevitable in modern healthcare, and the decision is between having it done on an ad hoc basis at the local level, or by experts centrally.

Inequality at the local level is referred to somewhat euphemistically in the UK as the “postcode lottery.” It’s ironic that on their website the Tories are claiming that Labour failed to reduce this postcode lottery, so in response they’re going to do away with NICE; when NICE is one of the most effective tools for preventing inequality from widening. They are instead going to introduce a policy that risks widening health inequality at a regional level.

Good points of the Tory policy

Going on the statements on the Tory website, this new policy will lead to a faster drug approval process, since NICE can be quite bureaucratic, while negotiations between government and drug companies can work quickly. I think this is a plus, but only marginally, since the only drugs NICE affects are those for diseases that have an existing treatment. Making fast decisions without the benefit of an organization like NICE to advise you is also a bad policy plan, since the drug companies will hold all the knowledge and professional expertise; but it appears that the Tory policy will enable NICE to retain its ability to make cost-effectiveness judgements – they just won’t be binding. What this means in practice, I think, is that the price of drugs will be fixed before NICE can complete its assessment, which will then show that they are overpriced and achieve little.

In principle I suppose that taking into account the “wider social costs” of refusing a drug can be good (there is some discussion of its effects on carers in the statement) but I’m suspicious about what this means, and whether these “wider social costs” will largely be presented to government by the pharmaceutical industry-controlled rare diseases lobby, rather than by doctors and social researchers.

Finally, a move to value-based pricing could be an improvement on the NICE system if its done in a coherent, well-organized and structured way, like the Australian system. Unfortunately, the policy as currently presented seems to suggest an ad hoc process with no strong economic analysis behind it. So I suspect that this means that the negotiations will heavily favour the pharmaceutical industry.

Conclusion (at last!)

Ultimately, I think this seems – based on the information currently available – to be a very bad policy, being released at the very worst possible time. With most government departments facing huge cuts and the NHS facing funding cuts in real terms over the next 5 years, it’s madness to introduce a policy which will lead to cost inflation in one of the NHS’s key spending areas. The 200 million pound cancer drugs fund is going to be a bonanza for the pharmaceutical industry that will be of limited benefit to patients or the NHS as a whole, and undoubtedly wouldn’t pass a cost benefits analysis. At a time when immunization rates are plummeting in the UK and many hospitals are suffering serious problems of overcrowding and care quality, that 200 million could undoubtedly be better spent on the front lines, either increasing immunization rates or improving frontline care for all patients, or reducing waiting lists. It looks suspiciously like a sop to the people who funded the Tory election campaign, rather than a serious cost containment policy for what everyone in the UK claims is the most cash-strapped era since world war 2. This is the time to be strengthening NICE and reducing the influence of the pharmaceutical companies on spending decisions, not throwing money at the most marginal drugs.

—

fn1: bar a few minor caveats, that is

fn2: To the best of my knowledge

fn3: QALYs are used because in many cases the treatment may lead to increased lifespan but increased disability, and a QALY adjusts for the perceived value of a year of life with a disability compared to a healthy year – it may be, for example, that people prefer to live 5 disability-free years than 10 disabled years, if the disability is severe

fn4: I’m doing this in dollars because I don’t seem to have a pound sign on my Japanese keyboard

fn5: This would be a cost benefit analysis, but NICE considers cost-effectiveness. My understanding of the economics is that cost benefit analysis is done to compare opportunity costs of two different treatments – so if you had two new treatments with no existing comparator, you could calculate the cost and benefits of each and fund the one that gives a better benefit per pound; but NICE doesn’t do this, and would in fact recommend both treatments on health economic grounds, though I think there are some other aspects of treatment approval on which it can take an FDA-like stance and bar a treatment, though I’m not sure about this

fn6: This is based on the Guardian’s report, so should be taken with a grain of salt

fn7: Rumour has it that this particular rule has been ditched, so who can say what else is going to go pear-shaped?

fn8: Because systems pool risk and/or buy in bulk, enabling costs to be shared and individual treatment costs to be pushed down, making a wider range of treatments available

fn9: and these things often go together!