In May this year, I wrote an article for Nautilus Magazine in which I described the effect of Goodhart’s Law on the usage of performance metrics in our health care system. In short, one interpretation of Goodhart’s Law states: ‘If a measure becomes a target, it ceases to be an effective measure’. What this effectively means is if you choose anything which you can measure as the ruler with which you assess people then this measure will be gamed by the participants of this little game. They will try to optimize this measure at all costs, as this is now effectively the only aspect upon which they will be evaluated. Frequently this then leads to unintended side effects.
In the realm of health care this can be witnessed when reasonable sounding metrics, such as length of time until hospital readmission or emergency room waiting time, are used to evaluate and reward or punish hospitals. These hospitals will then only optimize this metric (by for instance preventing needed patient readmissions or assigning the majority of the doctors to the emergency room at the detriment of other hospital areas), which usually leads to unintended side effects and neglects other relevant factors. Resultingly, health suffers. Even if the original metric was initially well chosen, it will be subverted and the approach can backfire badly.
To combat this or at least reduce its impact, reliance on as many (suitably weighted) metrics as possible would partially avoid these acts of negligence. Ideally, these metrics (or at least their weights) should be kept secret so that hospitals are not even tempted to game the system. If this piqued your interest, I definitely recommend you to check out the article in the link above!
Engineering Life 