The pregnancy related maternal death rate apparently doubled in two years, and nobody smelled a rat? That kind of change is unreal, unless there was a genocide, a civil war or other major cataclysm. It has to be bad data, or as the writer points out, a change in data collection protocol.

So what should be done? Take a lesson from how the cosmic distance ladder is established in Astronomy, read all about it

You can’t use the same measurement principle at all distance scales. So when changing from one measurement system to another you have to ensure a large region of overlap so you can calibrate the new one against the established one.

Analogously, if you are trying to measure something (maternal mortality rates) with known time-dependent behaviour under a stable (even if wrong) data collection protocol, a new, possibly more accurate, protocol (even one that has been calibrated to boots on the ground reality checks for a subset) has to be phased in over a time scale comparable to the one over which real change is suspected to take place, so you can model and separate the real change in numbers from the one due to the change in protocol.

What can we do about the past? If adoption of the new protocol was not instantaneous and if we know the proportion of new vs. old protocol usage for a sufficient number of points in time, the real time-dependent effects and protocol-dependent effects can be separated.

I stop to miau to cats.

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store