Huge pet peeve of mine when articles discuss a study but don't link it. Link

When you look at what else we do in the modern world to avoid/mitigate a similar level of mortality, these seem like no brainers, especially the masking. A 0.19 per 100,000 reduction seems small, but the mortality rate was ~50 per 100,000 in Canada. This is basically a 4% reduction in deaths by masking. For a minor cost measure with no long term economic outcomes, this should have been an easy one for people to get on board with.

School reduction in spread (~10%) shouldn't surprise anyone who knows school aged kids. They're germ balls, and multiple kids per household means there's effectively no "bubbles", so of course stopping that spread would have a huge impact, but then kids are stuck at home, so it has a huge cost too.

SIPOs (Shelter in Place Order) had a 5% reduction in hospitalizations, which is honestly smaller than I was expecting.

See below for the mortality summarization.

Specific NPIs. Of the 6 NPIs studied in per capita mortality, bar/restaurant closures experienced the largest effects 4 weeks after implementation, corresponding to 1.08 fewer deaths per 100,000. Although we did not find any more evidence of fewer deaths per capita, limited gatherings (–6.41%), business (–5.32%) and school closures (–3.98%) were associated with decreased mortality growth rates after 2 weeks, whereas SIPOs (–1.66%), masks (–1.9%), and school closures (–8.29%) corresponded to reductions after 3 weeks. SIPOs were the only NPI studied at the ≥4-week lag for growth rates and were associated with a 1.95% drop in the mortality rates. Although the authors observed mitigative effects of travel restrictions on case growth rates, this was not the case for mortality.

I imagine its a tricky thing to study in any detail, and its interesting to look back now to see the effectiveness of the measures put in place.