Preface
The results of Chapter 7 show that annual monitoring is more cost-effective than less frequent monitoring strategies, even though (see Chapter 5) annual monitoring carries the highest rate of statin prescriptions that are ‘unnecessary’: that is, prescription of statins to people whose underlying cholesterol value (or cardiovascular risk) is below a treatment threshold, but whose cholesterol value (and hence apparent cardiovascular risk) measures high on a particular occasion due to within-measurement variability (see Chapter 4).
This cost-effectiveness, despite over-prescription, arises because statins are cheap, so that the main deterrent to their wide use is possible harms, and, because they are well-tolerated, with few side-effects compared to their cardiovascular benefits, in almost all populations the cardiovascular benefits outweigh the few known health risks. Further, as a result of the expiry of atorvastatin patent status in 2012, the costs associated with prescribing statins are lower than the cost of a monitoring programme to allocate (‘ration’) them. Thus, frequent monitoring programmes, with a relatively high false-positive rate and relatively low false-negative rate (see Chapters 5 and 7) are more cost-effective than less frequent monitoring programmes with a lower false-positive rate but higher false-negative rate.
At the early dissemination meeting (see Chapter 7) this interpretation of the cost-effectiveness results was discussed. It was observed that, taken to the extreme, this line of argument would advocate universal statin treatment (in some age-defined population) rather than a monitoring programme. Provisional results from the simulation model of Chapter 5 were presented showing the modelled cardiovascular benefit of universal treatment compared with treatment based on a monitoring programme ().
Modelled cardiovascular event rates under the assumption of universal treatment with 80 mg of atorvastatin, universal treatment with 40 mg of atorvastatin, or treatment allocated according to current UK guidelines, in a primary prevention (more...)
The PPI member of the project steering committee was therefore invited to consider, from a patient perspective, the potential implications of replacing cholesterol monitoring with universal treatment in an age-defined population. The PPI member reviewed drafts of Chapters 2–7, met with the project team for discussion, and prepared a written commentary, which follows below.
To monitor or not to monitor? A patient and public involvement view
Summary As the PPI representative, I was asked to consider the proposition that universal prescription of statins could sensibly be adopted for everyone over a certain age without prior measuring or subsequent monitoring of cholesterol or lipids. My personal conclusion, which I would guess would be the view of many patients and members of the public, is that such a change of practice might turn out, on closer inspection, to be a very finely balanced question and one to be treated with great caution. I suggest a list of factors that should be considered and, if possible, quantified. And – if the balance of advantage still appears to lie with universal prescription – some thoughts on a cautious step-by-step introduction.
The question
- 1.
Now that atorvastatin has come off patent it would cost less to prescribe statins to everyone over a given age than to conduct tests. The known side effects are thought to be slight and the benefits great. We had to consider whether or not this was a safe option with great benefits to health as well as to expenditure. Speaking personally, I find it exciting when a new insight rearranges all the furniture and one knows that the room will never look the same again. But new insights generally arrive with the beauty of simplicity, uncluttered by caveats and conditions. They need to be interrogated.
- 2.
In the present case the insight hangs on a direct comparison of the net costs of two options. In discussion we came to call these ‘measure and monitor’ compared with ‘fire and forget’. The former included the costs of initial measurement and subsequent monitoring, selective prescription and known side effects, and the benefits of reduced cardiovascular incidents. The latter included the costs of universal prescription and known side effects, and the benefits of reduced cardiovascular incidents.
- 3.
As a patient on statins and a lay member of the public, I wanted to know whether or not those costs and benefits painted the full picture. If there were additional costs or drawbacks to ‘fire and forget’ or potential savings and improvements in ‘measure and monitor’ then the gap between the two options might close or the balance of advantage reverse. I suggested a list of potential costs and savings that patients or members of the public might wish to be considered (many of these having been already identified in team discussion).
Potential extra costs with ‘fire and forget’
- 4.
Unknown side effects These might arise either as (a) a result of new medical research or (b) new evidence accruing with the passage of years (to date it has not been possible to study the side effects of taking statins over long periods of time but a universal prescription could place hundreds of thousands of people on statins for 30, 40 or 50 years).
- 5.
Over- or under-prescription Statins prescribed without measuring the patient’s condition would necessarily mean a standard dosage. A low dose would fail to achieve the protection needed by high-risk people, with consequent costs from cardiovascular incidents. A high dose would imply additional costs from side effects, both known and (if any) unknown.
- 6.
Non-adherence I learnt that occasional lapses by a patient in taking statins are generally unimportant. But there are those who discontinue their treatment for long periods or completely, for a variety of personal reasons ranging from anxiety about side effects to over-optimism about their health. Non-adherence brings costs to health. We should expect non-adherence to be greater once statins are prescribed without any prior measurement because many people would doubt that they have any need for them.
- 7.
Labelling Would being told that they need statins lead to more people labelling themselves as having a medical condition to worry about? This might increase costs due to (a) more visits to their GP (see discussion of GP costs below) and (b) an increase in sick leave from work.
Fear of side effects
People vary. Some are keen to take prescriptions, alternative medicines, dietary supplements, etc. Others are reluctant to muck around with their body or are even fearful of side effects. With a universal, unmonitored ‘fire and forget’ policy, this latter group might become significant. One person talks to another. Resistance to taking the prescription would probably spread, non-adherence rates would rise among those who start the regime, and the costs from unprevented cardiovascular incidents would rise.
Media scares
Investigative journalism and marketing managers, in newspapers and TV alike, love a good scare story about health. Media stories are magnified by Twitter and YouTube. People do not trust scientists, pharmaceutical companies or governments. There must be a risk of a scare about the side effects of long-term use of statins by people who do not need them. Perhaps the probability of a scare is no higher than moderate, but the consequences could be highly costly: potentially not only a collapse of faith in universal prescription, but also of growing distrust of statins themselves, leading to more cardiovascular incidents in the population than at present. And perhaps a good scare about statins would smooth the path to a new scare about some other science-based proposal.
Potential new savings for ‘measure and monitor’
- 8.
GPs’ time The main costs of measurement are in staff time, and the GPs’ time is the most costly. Visits to the doctor are triggered by the patients themselves. I question whether or not GPs’ time might be overstated in the present analysis of ‘measure and monitor’: (1) some of the present GP visits should be discounted from the equation (because the patient wants a visit for other reasons); (2) general practices could promote widespread cholesterol monitoring without a prior consultation with the GP – just going straight to blood sampling; and (3) as a minor point, I wonder why the first-year monitoring costs in the model are assumed to be three to four times greater than those for subsequent years – is this dependent on assumptions about GP time?
- 9.
Lower cost measures The study is looking at the effectiveness of different measures in predicting health problems. If (without choosing less effective tools) measures can be adopted that would require fewer GP consultations or fewer visits to the clinic for blood sampling then costs would reduce.
- 10.
Longer intervals between monitoring Given the observation that statins are so cheap and relatively harmless that you might as well do away with measurement altogether, there is at least an argument for extending the periods between taking measurements in a regime of monitoring. Ten-year intervals, rather than five-year intervals, would halve the monitoring costs.
Intermediate positions
- 11.
If, after exploring these possible costs and savings, the balance of advantage still appears to lie with a ‘fire and forget’ regime then it might be prudent to consider partial or incremental implementation.
- 12.
For example, as a first step, universal prescription without measurement might be brought in for those aged > 70 years. The threshold might be reduced progressively to ≤ 60 years. This would seem particularly attractive if the balance of advantage seemed to favour ‘fire and forget’ on those costs that are readily quantifiable, but there were to remain great doubts about the hard-to-quantify factors. These might include, especially, widespread public doubts about the long-term side effects of taking statins and the risk of a media scare story. A cautious first step to those aged > 70 years old would allow time for the public to accept a new approach, time for the public reaction to be assessed, and time for medical research into long-term effects of statins.
Conclusion
- 13.
I hope these thoughts are helpful. I cannot judge how far these factors are amenable to quantification, or even how far the arguments are valid. They are a lay person’s perception, but I think it is reasonable to emphasise that I suspect ‘fire and forget’ would be very exposed to repercussions from public doubts and anxieties, and vulnerable to media scare.
John Stevens
