My Cleveland Indians have now won 13 games in a row. That along with their overall record and stellar play over the entire season tells me they are a good team. However, at any given point the Indians or any particular player could play really poorly. The similar challenge with ratings is they can give us a false snapshot at any particular point in time. For example, if we measure infection rates at just a particular period of time, we are prone to come to the wrong conclusions. As an example, this article by Consumer Reports names 19 teaching hospitals with higher than the national average infection rates. It is based on a snapshot at one particular time.
The Hospital Compare data tends to give us all a short-term view. In just the most recent data from September 2016, nearly 15% of hospitals had an increase or a decrease of 30% or more in the infection rate of clostridium difficile (aka C. diff). These variations really pop when looking at the data over many quarters with the Quality Ratings Interactive. For example, it is apparent that Cary Medical Center has a two quarter infection issue whereas Castleview had a high rate just in the June quarter.
While we love winners and losers, what really matters is a long-term sustained view of how facilities perform over time.