Of “clever” models and “dumb” spreadsheets”: what is more effective to drive policy change?

What often prevents modelling results from contributing to policy change is not lack of good science, but ineffective communication to the target audience of clinicians and decision-makers.
Dosing of anti-infectives in children is a glaring example of this. While the theory of maturation and allometric scaling are widely assumed as the gold standard within the pharmacokinetic modelling community, a number of international guidelines for dosing in children is still based on weight-bands targeting the same mg/kg dose as in adults. This happens for drugs in neglected diseases, when no directly observed data is available in children, but also for common diseases such as HIV or tuberculosis. This results in millions of children potentially receiving sub-optimal doses.
How can we get across the message of our models and use it to improve policy? Sometimes a simple and easy-to-use solution like an Excel spreadsheet can do the trick better than sleek-looking Visual Predictive Checks and impressively low parameter precision or shrinkage values.