April 1977 & April 2020
Exploring peculiar hikes in monthly excess non-COVID mortality in England & Wales (rev 1.1)
By now it should be abundantly clear to subscribers that something highly peculiar was going on during April 2020. We may try and pin the blame on false negative test results but this doesn’t explain why this problem only reared its head in April, and then among older females. A false negative is a false negative and test performance (sensitivity/specificity) is not going to be age and gender specific unless we have designed a magic test. Neither does this strained argument consider that we are talking physician-certified death certificates upon which more than a mere test result hangs.
On top of this there’s the peculiar finding of a similar inexplicable hike in excess age and gender standardised mortality April 1977, and this prompted me to do a bit more ARIMA modelling to see how the two anomalous Aprils came out in the wash.
What I’ve baked up this morning is automated ARIMA modelling of monthly excess non-COVID mortality by quinary age group and gender for the period Jan 1975 – Dec…