I don't mean to say I love the CDC and think they're perfect and arent being influenced or biased by the current government, but I do want to share some insight on data analysis to help people understand why they're doing this (and it actually IS for a very good reason), because there are some not-common-knowledge statistical principles in play here that make this make MUCH more sense. TL;DR at the end.
(Firstly, I want to add in a small piece of context that you can't actually do percent change calculations if starting from 0 cause like. Well "it's like 5 times the original value" so 5 × 0 = 0 so no change? It just doesn't work.)
Updating these baseline values gets us an incredibly more accurate measurement, because they are in no way measuring the amount of covid in a community. They are measuring the *rate of change over time*:
Note that they here list "overall levels of the virus in wastewater" as a SEPARATE value, which is used IN CONCERT with the rate of change data, and not being REPLACED by it.
Using the baseline of zero for rate of change tracking gives the information: how much more covid virus is in here as compared to no covid? Which can give you one piece of information: how much total covid virus is in here right now? That's important but is easily covered by their separate data category, How Much Covid Is In Here Right Now, which they already have because that's what they measured in the water. Finding this value using analysis would just be using this data point to find itself. It's redundant and doesn't give us any more information than we already had.
Using the baseline of the measurements from 2024 gives us this information instead: how much are the levels of covid virus changing in this location? This gives us a TON of important information. Some examples of this could include early warning of a major outbreak in an area; identifying which communities are are most in need of extra support; tracking gross transmission rates; tracking rates of breakthrough infections (could indicate a new variant that current vaccines aren't effective for); and so many other things.
Even more importantly, doing this STANDARDIZES and EQUALIZES the data and keeps small numbers from slipping through the cracks and being eclipsed by bigger numbers. The CDC had some info on that same page about this:
I want to expand on another aspect of percent change calculations to make it a bit clearer.
As an imaginary example, say that wastewater samples taken from a city in December of 2024 had enough virus to indicate about 1000 infected people were in that city. Then say they took two more samples which indicated 1050 and then 1200 infections.
Using 0 as a baseline, the percent rate of change from 1 to 1000 is a 99,900% rate of change; 1 to 1050 is a 104,900% rate of change from baseline, 1 to 1200 is 119,900% change. It's all urgent, but it's pretty close. Nothing is making it SIGNIFICANTLY better or worse. The clearest thing we can learn from this is just: there's lots more covid than there was when there wasn't covid. And that's a good point but unfortunately isn't new information. We already know there's more covid than when there wasn't covid.
Using the first value of 1000 instead: 1000 vs 1050 is a 5% rate of change from baseline; 1000 vs 1200 is a 20% change. That is a really significant difference! What was going on when the second sample was taken??
Ultimately, at this point, 0 cases of covid can be considered an outlying value. It's so different from any number we've seen since 2020 that it makes it extremely difficult to get the kind of specific, detailed information needed to form any meaningful, actionable insight. We KNOW there's an enormous gulf between when there were 0 cases and now. We don't need to measure that, for the same reason data tables tracking covid over time don't start in 1973. There is a massive quantity of potential useless data between now and 1973, and it only makes our data more watered-down and less useful to include that.
I made an example with graphs. Dataset:
Using 100,000 as a baseline we get a lot of good information about how the numbers are changing:
Using the "objective number zero" as a baseline:
Hmm. Not really getting much out of that one if I'm honest with you.
Tl;dr This is good data practice. They are not redefining what is considered "low" as compared to not having any virus because they are not trying to determine how much virus is present. They are redefining "low" to mean "the lowest extreme recent data value" in order to narrow their focus, because they are tracking rate of CHANGE, not rate of INCIDENCE. The updated baseline makes it possible to gain specific, detailed information about how covid transmission is changing over time, which can give them - and us - actionable insight. Using zero as a baseline for this is utterly useless.