Originally posted by Bill the Cat
View Post
Originally posted by Cow Poke
View Post
But where's the fun in stopping there?
Science reporting sucks, in general, but this is an unusually clear example not just of how badly it sucks, which is little more than a bromide today, but how it sucks, specifically, giving us a chance to examine the failures in science reporting looking for underlying causes. Far too often, science reporters get the facts wrong because they can't comprehend the science. Excuse me while I cringe. But in this case, the suckage is systemic, making for a more irenic and detached critique.
Compare the Hill reporter's synopsis of the study ...
.
Researchers determined that the waning immunity had to do with the amount of time since an individual was given the second shot rather than due to the highly infectious delta strain.
with the study's findings as reportedly immediately following the synopsis ...
.
Reduction in vaccine effectiveness against SARS-CoV-2 infections over time is probably primarily due to waning immunity with time rather than the delta variant escaping vaccine protection.
Emphasis on the "probably primarily" redacted from the reported synopsis.
By the numbers from the study itself, there is a decline in effectiveness directly associated with the delta variant, but it's judged as less important than the decline due to the passage of time. To wit a decline against delta from 93 to 53 percent after four months is contrasted with a decline against non-delta from 97 to 67 after four to five months. Where the decline against delta measures 40 points after only four months, the decline against non-delta is only 30 after four to five months.
.
Among sequenced infections, vaccine effectiveness against infections of the delta variant was high during the first month after full vaccination (93% [95% CI 85–97]) but declined to 53% [39–65] after 4 months. Effectiveness against other (non-delta) variants the first month after full vaccination was also high at 97% (95% CI 95–99), but waned to 67% (45–80) at 4–5 months.
Scientists speak carefully and deliberatively, in tentative terms so as not to go beyond their data, with close attention not merely to their results, but to their confidence in the results, reflecting their training. Journalists, conversely, are schooled away from modifiers, which are presented in the classroom in emotionally evocative terms like "weasel-wording," biasing against modifiers even when clearly indicated as above. Nobody wants to be called a weasel, and especially not by your boss.
That's why I conclude the suckage is systemic, in this case. In science, modifiers aren't included for sport. In science, all modifiers matter! (Oh no I didn't!) Science researchers are just as interested in being able to claim definitive results as anyone else. (Cue the preening mathematical researchers pointing at exact results in the hundreds of thousands of decimal places. It must be mortifying, being a mere scientist.)
A couple of minor things for numbers wonks, because this is me. Look at the confidence intervals [in boxed parentheses] and the measures (leading figures in the enclosing open parentheses) and note the measures aren't necessarily midpoints, e.g., the range for 67 isn't plus or minus a constant, but rather 67+13 on the high side v. 67-22 on the low. That's not an arithmetic error, it's how the modeling accounts for the distribution of data.
Then look at the confidence ranges, from the tightest at four points, (99-95), for full vaccination against non-delta after a month, to the loosest, at 35 points, (80-45), for non-delta after four to five months, significantly larger than the 26 point range for delta, (65-39), after four months. That's a clear statement that the delta variant was not dominant at the beginning of the study period, but became dominant for the majority of the study period, creating larger sample sizes for delta infections, and hence, tighter confidence intervals.
Obviously, given my druthers, independent of my membership in House Pfizer, I'd rather be exposed to non-delta. But as a member of House Pfizer, I'll take 67 percent protection over 53 percent every time.
I know Bill is keen on natural immunity, so I'll add that this study doesn't address it, but that I'm keeping a weather eye out for any studies that look at or ideally compare the decline in the post-infected population, but I know already most of that population has also been vaccinated now so his cohort is going to be especially difficult to track down.
Comment