HIF proposes $600 million annual budget to evaluate health impacts

One feature of the HIF that is raising eyebrows is their proposal to
spend $600 million *per year* on assessing health impacts. This is
apparently more than an order of magnitude greater than the entire
budget of NICE. The reference to the $600 million from the HIF book

Aidan Hollis and Thomas Pogge, The Health Impact Fund, Making New
Medicines Accessible for All, A Report of Incentives for Global Health,

Page 30-32


Health impact assessment would be expensive, given the need to assess a
variety of medicines globally. There would, of course, be some economies
of scale from assessing many medicines at the same time, and
efficiencies from assessing the same medicine year after year. However,
a reasonable perspective is that if the HIF had an annual budget of $6
billion, it could spend about $600 million on administration and
assessment, with the bulk being devoted to assessment. This would make
it by far the largest health assessment agency in the world. For
comparison’s sake, NICE (the UK’s National Institute for Clinical
Excellence) has a budget of approximately $50 million. NICE publishes
around 25 technology appraisals, 12 clinical guidelines and 60 pieces of
interventional procedures guidance each year (NICE 2004). The HIF would
have, assuming a stock of about 20 medicines registered at any time, a
requirement to evaluate the impact of those medicines around the world,
which would be a much more difficult process than that undertaken by
NICE. However, there could be considerable external benefits from such
an assessment process, including primarily that it would enable better
prescribing as the relative therapeutic benefits of different products
were better understood.

The HIF would be by far the largest health assessment agency in the

A budget of $600 million, spent on roughly 20 medicines at any given
time, yields an average budget per year per drug of $30 million. How
would this be spent? Part would be allocated to evaluating clinical
evidence. Current estimates of the cost of trials can be found in Holve
and Pittman (2008), who estimate that head-to-head studies range in
price from approximately $2.5 million for relatively small studies to
$20 million for large studies. Such studies, of course, would not be
conducted every year; some such studies could be performed by the
registrants, though the HIF could also commission its own independent
studies where needed. Observational studies range in cost from $1.5
million to $4 million. The HIF would require observational studies in
different settings, though not every year, so this could be quite
costly. However, it is likely that observational studies would be less
expensive in developing countries. Systematic reviews of evidence tend
to cost up to around $0.3 million. The HIF would also require a
substantial auditing function to ensure that the products were being
distributed and used in ways consistent with the findings of the
observational studies. Finally, there would be a significant overhead
component related to obtaining the functions of the technical branch and
other operational branches, which could be shared across products.

Errors using inadequate data are much less than those using no data at