[CoachKandSportsguy]

Just a heads up to help prepare people for a bunch of states releasing new COVID stats this week. CoachK and her team has prepared the records for her MA hospitals, as well as being on the call with the Mass department of public health (MDPH). Also remember that the states' DPH are requesting the information to satisfy some political issue, whether pressure from public opinion or backup for a past decision or a future decision being contemplated to be issued.

First, remember that the state public health departments request this information from hospitals, and that hospitals have to often times figure out how to report the data from the data collected in the electronic health records (EHR) in the medical records system. The data requested is not usually as simple as a search on a data field with a Yes No flag, but a series of if this then that or else if that then this or else default type logic.

The reports are based upon each state's definition of the stat, and are sometimes the definition is left up to each hospital to report on their best efforts, or they agree on something can be used as broadly / generally across the different hospital systems which has a less accurately but the state can feel good about the data quality to publish something.

Second, because each state has their own DPH, states do not coordinate and standardize amongst themselves. that is important when comparing state to state, or even nationally.

So what you will start to see is a bit of a mish mash of percentages which should NOT be to be taken as perfect, not definitive proof of what is trying to be measured. There may be large swings as coachk has seen at their different hospitals, yet, what will be presented is a state total number, biased to each states population distribution. ie NY state has a bipolar population - large percentage of population in a very small urban area, and a very small population in a very large rural area.

the first releases have been questioned by reputable doctors as not what would be expected, which also means that the data definitions might be questionable, misunderstood, or assumed to be perfect. . . they are not, but also something in the absence of nothing may be better, or may not, it always depends. . .

So our suggestion is to observe and find the data interesting, but not to make any judgments about absolute accuracy of the stats, because they are not. . . they are a first attempt at trying to answer the question. . .

over time, the definition will become better, but never the first time. . .

[Altavia]

Thanks for the explanation and heads up.