Risk sharing and drug pricing
327. The OFT report recommended that the problems
of price setting with limited data and uncertainties about treatments
at launch could be mitigated using risk-sharing schemes. The financial
risks associated with uncertainty due to lack of evidence would
be split between the NHS and the manufacturer. As Professor Peter
Smith told us:
There are three [possible] approaches to take. One
is to wait and do more research; one is to take a provisional
decision that one may reverse in the future; and the other is
in some way to get the companies to share the risks of accepting
the drug early.[319]
328. Key to effective risk-sharing is the adequate
measurement of the effects of treatment following its introduction
to the NHS. Witnesses told us that "monitoring studies"
should be conducted to determine whether the NHS or the manufacturer
should pay for the treatment. According to Professor Jon Nicholl,
an initial price may be set, based on best estimates of efficacy:
At a later date when the monitoring study is concluded,
the cost-effectiveness of the treatment can be recalculated, and
the appropriate NHS price of the drug reviewed. If the treatment
is less cost-effective than previously estimated there may be
reimbursements paid for earlier treatments, as well as a change
in the price paid for future treatments.[320]
329. Professor Claxton agreed that a lower price
could be set initially, while such studies are carried out, to
ensure that the NHS does not lose out financially:
If the type of evidence required cannot be provided
with provisional approval, then price ought to be reduced so that
the cost-effectiveness of the drug is not uncertain.[321]
330. However, there are problems with this approach.
They include the difficulty of imposing clinical trial conditions
in the real world of the NHS. According to Professor Nicholl,
the universal availability of drugs used under risk-sharing schemes
means that there are no comparator groups and no controls. This
means that the uncertainties observed initially are not resolved:
One cannot collect evidence which says that the outcomes
for patients who are being treated by the treatments being made
available under a risk-sharing scheme are better than for patients
not being treated by those drugs. In turn the consequence is that
we cannot resolve the uncertainties in the cost and effectiveness
as we try to do in the first place. [322]
331. Such studies are expensive, and who should pay
must be determined. Professor Nicholl and others stated that they
should be publicly funded, possibly with a charge to industry
to cover the cost of acquiring the information or consideration
of the cost when determining the price of a product.[323]
332. NICE and the Department of Health have recently
accepted a risk-sharing scheme for the multiple myeloma drug Velcade
(bortezomib)[324] and
the OFT report indicated the positive aspects of risk-sharing.
However, significant problems have been observed with such schemes
in the past. The scheme to evaluate the benefits and risks of
beta interferon and glatiramer for multiple sclerosis does not
appear to have been successful (see box below for details). Witnesses
warned that "we should be wary of rushing into this as a
generalised mechanism".[325]
333. Some treatments may be more suitable for risk-sharing
than others. Bortezomib, for example, lends itself to such a scheme
because there is a protein marker that indicates whether a patient
has responded to the drug or not.[326]
Dr Barker told us that such markers were likely to be more common
in the future:
increasingly we are developing markers of response
and so it will be more practical for medicines to be given with
a marker of response as a diagnostic so that everybody has the
confidence, the patient has the confidence, the clinician has
the confidence, the NHS has got the confidence that this is being
given to a patient for whom it will work.[327]
334. Professor Nicholl also indicated that risk sharing
schemes would be suitable to determine the relative prices of
two treatments where patients and clinicians have not yet decided
which treatment is preferable.[328]