Realism and Learning in Pharmaceuticals
Bilyard recommends defining what a drug-development study must establish and what can be lost if it fails. Testing critical requirements early can reveal an unviable product before further investment, while clear goals make later root cause analysis possible.
THE IDEA
Realism and Learning in Pharmaceuticals
Dr Kevin Bilyard, former CEO and Board Director of Immodulon Therapeutics, describes how goals and tolerance for failure change during drug development:
‘The vast majority of board members and investors in biopharmaceutical companies have a deep understanding of how risk can vary between different stages of research and development.
Early in the discovery stage, a setback or failure will be a disappointment. Still, you won't necessarily conduct a full‐blown investigation into why the candidate drug didn't meet your requirements unless there is cause. This is because you understand that in pharmaceutical research it's impossible to be sure that all research studies will succeed. After all, you are carrying out work at the frontiers of science where no one knows what the outcome is going to be. Most stakeholders accept losses, and setbacks are an intrinsic part of the business and will normally be concerned only if there is an excessive failure rate.
However, at the development phase, where a candidate drug has passed through the research phase, a different set of considerations are in play. It can be tempting to approach drug development in a very cautious, incremental way. But the problem is that you can find yourself having spent millions of dollars on one study after another. Then you carry out a major study that shows that you do not have a viable product, perhaps because of an insufficiently strong statistical signal on a key characteristic for the drug.
As a result, there has been a growing practice of using multi‐disciplinary teams of medics, scientists and clinical trial specialists to design and oversee drug development. These teams can design clinical trials to meet the key scientific, regulatory and statistical evidence requirements “head on” as well as understanding practical matters such as how you might recruit volunteers for the trial.
Nowadays, most clinical trial teams aim to approach the development phase with a clear, comprehensive and practical sense of what is a necessary and appropriate goal and what might be lost if the study fails, to which the answer is, sometimes, everything.
This approach can't guarantee success, but it can help to ensure trials are neither over‐ambitious nor under‐ambitious. Losses will arise, but they will typically be lower and sooner than if the development programme was poorly conceived. The phrase you hear now is “fail fast”.’
The same principle applies to other projects: define the goal and acceptable losses before starting. Without these, root cause analysis (RCA) lacks a clear standard for judging what went wrong.6 After all, if you experience a setback and want to understand what went wrong, you need to be clear what ‘wrong’ means. As Yogi Berra said: ‘You've got to be very careful if you don't know where you are going because you might not get there’.7
When to use it
Developing a new drug and designing clinical trials to test critical assumptions early.
Implementing a new IT system by prototyping high-risk components first.
Trying a new medical procedure with a clear plan for evaluating success and failure criteria.
Starting a new business venture using lean startup principles to test hypotheses quickly.
FROM THEORY TO EVERYDAY LIFE
In practice
Pharmaceutical drug development
Bilyard distinguishes expected setbacks during discovery from expensive failures later in development. He recommends multidisciplinary trial design that tests scientific, regulatory, statistical, and recruitment requirements early enough to expose an unviable candidate before further spending.
IT system development
When implementing a new IT system, it's important to set clear goals and acceptable failure criteria upfront. Instead of building incrementally without testing core assumptions, teams can prototype key features early to identify fatal flaws quickly, minimizing wasted resources.
A MOMENT TO REFLECT
Put it to work.
What is your current tolerance for failure in your projects, and how do you define 'acceptable failure'?
Can you identify a situation where you were overly cautious and missed the opportunity to 'fail fast'?
How might setting clearer goals help you conduct more effective root cause analysis after a setback?
CONTINUE THE CONVERSATION
Source texts
- Beyond the Five Whys Root Cause Analysis and Systems Thinking (James C. Paterson)