Evaluating generally applicable advice and deciding if relevant
We came across a set of principles that someone follows to consider whether generally applicable medical advice is sound and applicable to that person. Given that no individual can do detailed analysis (be it due to lack of time to knowledge). Principles and heuristics are a practical answer.. Seem very reasonable, as humans are clearly not all equal.
It seems some of these principles, very relevant to the conduct of clinical trial, could also be more generally relevant. For example, when evaluating if best practice for management of IT services in large multinationals is applicable to a challenger… So we would be interested in your opinions!
1. Money aware
It goes without saying… if there is an unbalanced incentive to a decision maker or strong influencer, consider if it could be a key factor in the choice that was made, vs the target costs/benefits.
2. Check for bias (including unconscious)
Human nature often leads us to ignore some risks and wish for unreasonable benefits. The same can be the case when considering subsets of risks and subjects.
3. Consider the whole system
More than the process of defining and setting up a trial, you will be impacting an existing ecosystem of people, organisations and suppliers.
4. Test everything and keep testing
All assumptions are just that until they are proven with appropriately realistic tests. Scenarios and input and output data are key.
5. Insist on a full impact analysis that demonstrates the evidence on which it relies
Expose and explore all the data available before considering a new initiative. Often it reduces the effort by better targeting the improvement of measurement of its effect.
6. Ensure informed consent is part of it
This ensures that the trial is explainable to the subjects, both the expected benefits as well as the risks. Part of the Nuremberg code.
7. Follow the Nuremberg code
Trials should be to the benefit of society and risk limited by the humanitarian value of the experiment.
Clearly following the money is key, checking bias of all the decision makers and adding a different perspective, consider not only how quickly that shiny system will be up and running but who will keep it so including looking after security and much more can be said.
Testing is never enough, and how much can it be automated to lower the cost of change?
Also conducting an impact analysis, looking at the underlying data is key to surfacing the current gaps and why that particular solution is the recommended one to address them.
Finally, a service provider that ensures awareness of the risks by the IT customer and user community before making changes is always more likely to succeed.