– Written by Anonymous – a drug developer with a quantitative background
Explore different viewpoints with our PHUSE Blog series ‘An Alternative Point of View’. Hear from anonymous industry experts as they discuss controversial topics. Remember, these are opinions, not facts. While we value diverse perspectives and open dialogue, note that views expressed may not always be based on established facts. Join us for an engaging journey of exploration and discussion.
Disclaimer: This text is not based on experimental evidence or any type of systematic review. Rather, it reflects personal observations and perceptions.
W.A. Wallis once defined statistics as a body of methods for making wise decisions in the face of uncertainty. This definition emphasises the need for statisticians to engage with the implications of their work and actively contribute to decision-making processes. That is, it is about leveraging statistical methods to inform choices rather than limiting ourselves to the technicalities of data analysis.
The real life of statisticians in drug development
The role of statisticians in drug development is pivotal. This is, for example, reflected by the ICH E9 guidance on statistical principles for clinical trials. As a downside, this created a comforting bubble for statisticians. With the focus on generation of ‘firm evidence of effectiveness’, it also paved the way to almost completely remove the uncertainty from the decision-making. Instead, we are using well-designed clinical trials and a p-value as a decision criterion in a strictly mathematical way. If any kind of judgement is required, it is often left to the clinical colleagues once the evidence has been presented. This is even more pronounced in the analysis of safety data. For decades, the approach, in essence, was to throw summary tables over the fence, with the notion that the study outcome is now a medical question. The worst case is a situation where the reliance on firm statistical thresholds leads to a kind of paralysis – an inability to act when the evidence is not clear-cut.
This practice can inadvertently leave physicians navigating complex decisions with limited guidance.
Need to bridge the gap
To bridge this gap, statisticians must move beyond their comfort zones. They should be encouraged to engage in the judgemental aspects of decision-making, drawing from interdisciplinary knowledge. This is not merely an expansion of their role, it is a necessity in a field that often operates under uncertainty. The landscape of decision-making often requires looking beyond p-values and engaging with a broader spectrum of evidence – especially when it comes to assessing safety. It is unfortunate that this is not commonly part of the standard curriculum in the education of statisticians. However, we should not leave physicians alone when navigating complex decisions with limited guidance. While the integrity of statistical inference is crucial, it is not enough when decisions need to be made on weaker evidence from diverse and often unreliable sources.
One potential solution comes from the field of epidemiology: the Bradford Hill Criteria. (1) This set of criteria helps evaluate causal relationships and provides a framework for assessing evidence in a more holistic manner. Yet, despite their utility (2) and the availability of updated interpretations, (3) they are frequently overlooked by statisticians focusing solely on inferential statistics.
Moving forward
I feel there is an opportunity, perhaps even a need, for statisticians to embrace a more integrated role. By considering a broader range of evidence and engaging in the qualitative aspects of decision-making, they can provide invaluable support to clinicians. This collaboration can lead to more informed, nuanced and ultimately safer decisions that account for the complexities of real-world healthcare.
In conclusion, moving beyond p-values is not just a statistical imperative, it’s a professional responsibility. The intersection of statistics and medicine must be navigated with an understanding that decision-making involves not just methods but also judgement, context, and a commitment to patient safety and risk management. Let’s encourage a culture where statisticians feel empowered to step out of their comfort zones and contribute even more meaningfully to the discussions that shape healthcare.
References
- Bradford Hill A. (1965). The environment and disease: association or causation? Proceedings of the Royal Society of Medicine, 58(5): 295–300.
- Buchanan J & Li M. (2021). Safety Signaling and Causal Evaluation. In: Quantitative Drug Safety and Benefit-Risk Evaluation. Practical and Cross-Disciplinary Approaches. Ed. Wang W, Munsaka M, Buchanan J, Li J. Chapman & Hall, CRC. The American Statistician, 77(1). https://doi.org/10.1201/9780429488801.
- Fedak KM, Bernal A, Capshaw ZA & Gross S. (2015). Applying the Bradford Hill criteria in the 21st century: how data integration has changed causal inference in molecular epidemiology. Emerging Themes in Epidemiology, 12:14.