Do numbers lie?

I once led a hospital quality improvement project team to execute a data collection plan that involved recording the time it takes for patients to leave the hospital after being medically cleared for discharge. The 2 week long exercise seemed to go well and when I looked at the numbers I found a pattern. The pattern wasn't related to the problem we were trying to solve but that each floor seemed to have its own distinctness with regard to the timepoints. I was sure something went wrong. So I brought the nurse heads on the different floors all together and asked them what it is they were actually measuring. It turned out that everybody had a different definition of the timepoint they were required to record. While for some it meant the time the patients had cleared all their bills and obtained their take home medicines, for others it meant the point when patients were handed over the discharge summaries. And then there were others with slightly different definitions of their own. While you make be thinking it was a poorly planned exercise, the fact is that definitions were clear to those on the project team but not to those on the frontline. These kind of communication flaws are not news to managers. However what is important to understand is that the flawed data gathering could have led to the wrong inferences and thereby wrong solutions to the problem at hand. And this is not just happening in quality improvement but in clinical and managerial audits as well.
So beware...always look for the story behind the numbers!