Drugbaron Blog

May 24, 2021 no comments

Why Small Beats Big: the Hidden Cost of Too Many Voices

The coronavirus pandemic has taught us a lot of lessons.  But some, such as exactly why small biotech companies are increasingly making all the running in early-stage drug discovery and development, are slightly less obvious than others.

Just this week, we have seen the UK Scientific Advisory Group for Emergencies (SAGE) urge the Government to slow the lifting of restrictions out of concern that the so-called Indian variant of covid might be spreading more quickly and could cause a further wave of infections and possibly therefore hospitalizations and deaths.

What does that have to do with the power of small teams in drug discovery?

The common thread is incentives.  Consider the position of the SAGE group: they are charged with forecasting what might happen to coronarvirus cases in the future, armed with mathematical models whose power largely depends on the assumptions upon which they are based.   Armed with such models, there are certainly scenarios that could see significant challenges for healthcare providers in the future particularly if new variants emerge that current vaccines protect against only weakly.  And all these scenarios are mitigated by caution now.  Keep everyone at home for longer, and the risk of future surge is, undoubtedly, reduced.

Hence the advice offered by the SAGE group to the Government, who are charged with making the decisions as to which restrictions stay in place and which are lifted.

Now consider the same scenario from the perspective of the Government ministers.  Keeping restrictions in place has substantial costs in many different ways: economic costs, impact on mental health, slower and poorer treatment of non-covid illnesses not to mention the strong ethical imperative to avoid restricting liberty except in extremis.  In short, as it has been throughout, the decision what to do has been a fine balancing act between the costs and benefits of restrictions.

Next, look at the impact of the advice from SAGE.  Because of their remit, to look at the best way of reducing the impact of the virus, they have set out the best way, in their view, to achieve that narrow goal.  In many ways, the advice is not all that surprising – at every stage throughout the pandemic more restrictions would have reduced viral spread.  And if there were no other considerations, it would have been the right choice to make.  Arguably, if everyone in the country stayed in complete isolation for a month, the virus would disappear altogether with every transmission chain hitting a brick wall simultaneously.

However, the advice has a more insidious effect.  The bad outcomes they forecast are far from certain.  Indeed it currently seems more likely than not that the vaccine will be effective against the variants and while cases may surge among the young, unvaccinated groups, hospitalisations and deaths will remain low.  But imagine the worst-case scenario does play out and deaths rise substantially, perhaps because of weaker-than-expected protection by the vaccines.  Now everyone will point to the SAGE advice and criticise the Government for ignoring what, with the benefit of hindsight, would prove to have been sage advice (if you pardon the pun).

The situation is made worse by the supposed “expert” status of the SAGE panel, and the Government’s own unwise statements about “listening to the science”.  Science is supposedly the ultimate rational arbiter, so being seen – with the 20/20 vision of hindsight – to have ignored “the science” is an exposed position to adopt.

Without the public advice from SAGE, the Government would likely balance the risks and proceed with lifting all the restrictions, rightly believing that the costs substantially outweight the benefits.  But that course of action becomes harder with the advice in hand.

In short, advice is not free.  Yes, it may improve your knowledge of the situation, but it also creates a risk – particularly if you, as the decision-maker, choose to ignore it.  Even if a bad outcome with a tiny likelihood now occurs, which could have been avoided had the advice been followed, the decision-maker is likely to be censured.

Play this scenario out in a biotech setting, and the problem of large teams becomes immediately apparent.  The more voices you have around the table, the more likely it is when something goes wrong that one or other of the many pieces of advice the decision-maker received would have mitigated it.

Faced with such risks, decision-makers usually decide to follow most, if not all, of the advice they are given.  The experts on toxicology, ADME, formulation, manufacturing, pharmacology, statistics and the rest all demand more studies, more information to help avoid calamity.  And if the money is there to pay for it, then the experiments get done “out of an abundance of caution”.

Larger teams, with more advisors, simply make work.  Other people on the team may be pretty certain that the risk that’s being mitigated is miniscule, but they don’t want to be caught out in the harsh glare of hindsight.

Unfortunately, drug discovery and development shares another key characteristic with the coronavirus pandemic: both are low-validity environments.  That is, there is a lot of uncertainty about how things will develop in the future.  The case for having less advice gets stronger and stronger, the less predictable the circumstances become.

The problem is the spectre of unknown unknowns – things that can go wrong that simply cannot be predicted from the current status, no matter how well-informed you are.  And in early-stage pharmaceutical R&D, the majority of failures occur for reasons not only that the team didn’t know about, but importantly could not have known about.

Almost a decade ago, DrugBaron used Monte Carlo simulations to explicitly demonstrate that, in such low-validity environments, gaining more information usually costs more than that information is worth in terms of improved decision quality.    But exactly the same models demonstrate that the benefit of seeking more advice is quickly beaten by the (largely hidden) costs of doing so.

In short, you should only ask for advice when you genuinely have a gap in your knowledge which you consider critical for making the decision in front of you.  You certainly should not ask for advice simply to provide cover for your decision-making (so that, if things go wrong, you can point to how much advice you had sought).  The UK Government is finding out now just how painful it is to have “expert” advisors providing a view on just one aspect of a complex multi-variable optimisation.

Large and well-funded teams are most at risk from this penalty – where the hidden cost of the advice exceeds the utility of the new information gained.  After all, the more people there are in the team, the more voices there will be that need to be heeded, or (more bravely) ignored.  And with enough resource, it is nearly always possible to mitigate every risk if you don’t care how inefficient or slow you are moving.

By contrast small teams largely side-step this problem.  There is hardly anyone around to offer an opinion that the decision-maker might need to risk ignoring.  Moreover, there are few resources to be deployed so even if an unlikely bad scenario transpires, the decision-maker has the protection of resource-limitation.  They had to use the limited resources to mitigate the most likely risks, and therefore were “unlucky” to be victim to such an unlikely happening.

In an environment as low-validity as early-stage pharma R&D, the difference is stark: the small team moves quicker, expends massively less resources, but has only a marginally smaller chance of succumbing to an avoidable mistake.  Calibrating exactly how small to be, how little advice to seek, is an open debate – it may not be as small as the zero-person biotech companies DrugBaron has often talked about.  But nor is it as large as a typical pharmaceutical company with access to so much expert advice that decision-makers risk being paralysed by the need to cover so many bases.

As we move beyond the coronavirus pandemic, we take with us important lessons (here is another…) about how to operate successfully in low-validity environments.   The huge cost of discovering and developing new medicines is frequently in the headlines, as the efficiency of pharmaceutical R&D stubbornly rises.  One of the few weapons we have to reverse that decline in productivity is to keep teams small, and above all think twice before asking for more advice – even from DrugBaron!

5 2 votes
Article Rating

Yearly Archive

Would love your thoughts, please comment.x