Thread by @tomtom_m

Latest @MRCza and @UCT_news report on excess deaths in South Africa released, covering the period to 11Dec2021. A total of 277 700 excess natural deaths since 3May20. Full report available here: samrc.ac.za/reports/report…

(2/) 1 887 excess natural deaths in the week to 11Dec, up from 1 726 in the preceding week. The p-score this week is 23%, up slightly from 21% the previous week.

(3/) Still no strong excess natural mortality signal from Gauteng or its metros. The p-score in Gauteng in the most recent week is 19% (the fourth lowest of the 9 provinces). Highest in the Eastern Cape (45%) and Mpumalanga (25%).

(4) Some uncertainty re what’s driving the pattern in Eastern Cape.

(5/) p-scores by age have risen slightly for ages 60 and over.
Unnatural deaths have kept uncannily close to the predicted levels for an extended period of time.

(6/) The @nicd_sa released their weekly testing summary today, covering the week to 11Dec2021. The report is available here: nicd.ac.za/diseases-a-z-i….

(7/) Proportions testing positive increased steeply in all provinces, other than Gauteng (which also increased in the week to 11Dec. Nationally, PTP was 35%.

(7b) By age, still testing highest among adolescents and young adults, but has spread to older age groups

(7c) Nationally, the country is shade of Christmas red. More than half of all districts recorded PTP > 30% last week

(8/) BURIED LEDE #1. The proportion testing positive, measured on a daily (as opposed to weekly) basis, has turned in the last few days.

BL#1/2 (This analysis uses only PCR tests (as it is known, see threads passim, that there are delays and problems with loading of antigen results), and split by public-private testing (as known IT problems last week).

BL#1/3 The lower PTP in private is likely due to greater volumes of worried-well testing in private, as well as other selection effects.

(9/) But we can also combine the data from the testing and excess deaths reports, lagging the excess deaths a week behind the testing data.

(10/) BURIED LEDE #2. The pattern in wave 4 is materially different from that in previous waves. A story in 3 graphs

BL#2/2: First. Case and (lagged) excess deaths: in W4, cases have rocketed, but excess deaths? Not so much. Not even in Gauteng.

BL#2/3: Second. Normalised cases and normalised (lagged) excess deaths: in the previous 3 waves, these metrics moved almost perfectly in sync. Not so much. Not even in Gauteng (where cases are > 80% of the peak in W3, yet excess deaths not even 10% of the W3 peak.

BL#2/4: Third. Same plot as the previous, but now presented as a correlation plot between PTP and (lagged) excess deaths. Generally a very stable relationship. Except the red dot for Gauteng, being the most recent week's data.

(11/) OK. So LOTS of people are going to go wild on here shouting about this. Here’s my not-so-hot take. These data are really important. They STRONGLY suggest that the Omicron wave in South Africa will be different.

(12/). But this is NOT definitive proof. We need another week or two data. And this is NOT clinical or virological evidence. And it MAY still be confounded (not least of all by age; we don't think so but still working on that).

(13/) All round the most positive news in weeks. This is cause for optimism. Let's now amass the rest of the data to (hopefully) show that this is all good!
Stay well; stay safe; vaccinate!

(in case clarification required for this and the next few tweets: _IN GAUTENG_; although the figure makes that clear)