Personalized Animal Studies

I was engaged by a major pharmaceutical client to develop a global data repository of pre-clinical results, that is, animal and pre-human test results. The idea was to avoid duplication by having access to the raw and processed data from all prior studies. In an interview with one of the animal researchers I ran into a wall.

I can’t just include the quantitative results from my experiments in the shared database. People will misinterpret the raw data without the contextual insights I gathered during the course of the experiment. Only I knew which rats were naturally calm or naturally aggressive, something that the raw data just doesn’t show. Only I knew when a normally dominate and active rat was just having a bad day, even without the treatment. I can share the results of my study, but I refuse to share the data because others will just use it out of context. Pharmaceutical Industry Client (2005). Personal Communication

This could be interpreted as an instance of the counterproductive behavior Knowledge is Power+, but more likely it was an instance of having been burned by others who built their own Narrative Bias+. The contextualized data didn’t support their conclusions so instead they went back to the raw data and drew their own conclusions.

These were cloned rats and yet the researcher was able to attach personalities to each of the 30 rats in the study. The researcher was able to draw first-hand conclusions on the affect of the treatment on each individual within the rat population. This was not a numbers game.

Yet in human clinical trials, wild-type humans, we do our best to hide the individuality of participants in the trials. We generalize, sanitize, homogenize and pasteurize the data and call it science. What does this say about the reliability of our clinical trial results? Of course we eventually must generalize our treatment regimen to a large enough population of patients so we can make money. This is commercialization+ aka FDA Approval. But before this moment, we are still in Research and Development, and we should not fool ourselves.

A much better approach, and eventually the industry will get there, are mini clini’s. These are small 6-10 person clinical trials where we do pay attention to the reactions of each individual receiving the treatments. We do leverage the intellect and professionalism of our clinical researchers to interpret the results. We do not generalize or quantify the results in an exercise of cheap heuristics+. Instead we step back from a comprehensive review of each patient file and make a judgment call on whether or not the treatment is worthy of further investigation.

How this happens, and how we eliminate investigator bias from this exercise, is a key challenge that has been tackled in the practice of World Class R&D.