@ProfMattFox 1/
The odds ratio from a case-control study is an unbiased estimator of the
a. odds ratio in the underlying cohort when we sample controls among non-cases
b. rate ratio in the underlying cohort when we use with incidence density sampling
No rare outcome assumption required.
@ProfMattFox 2/
Because the odds ratio is approximately equal to the risk ratio when the outcome is rare, the odds ratio from a case-control study approximates the risk ratio in the underlying cohort when we sample controls among non-cases and the outcome is rare.
But...
@ProfMattFox 3/
... for an unbiased estimator of the risk ratio (regardless of the outcome being rare), we need a case-base design, not a classical case-control design.
Of course, all of the above only applies to time-fixed treatments or exposures.
As for the causal interpretation...
@ProfMattFox 4/
If there is confounding, selection bias, or measurement bias in the underlying cohort, the same applies to the case-control.
For example, if we need to adjust for selection bias due to loss to follow-up in the cohort, we also need to adjust for it in the case-control study.
@ProfMattFox 5/
For more realistic applications of case-control studies with time-varying treatments, see the paper led by @barbradickerman on #TargetTrial emulation with case-control designs:
1/ We recently confirmed the effectiveness of the Pfizer-BioNTech vaccine outside of randomized trials @NEJM. nejm.org/doi/full/10.10…
Studies like ours are being used to promote a vaccine passport to travel in the US, UK, and European Union.
A few clarifications are in order.
2/
Before we start, a disclaimer:
Vaccine passports involve complex ethical, economic, and political considerations.
Here I talk exclusively about scientific issues. The goal is that those making decisions have a better understanding of what we do and do not know as of today.
3/ Based on our study, we can say confidently that the vaccine is highly effective in preventing you from getting sick with #COVID19.
Based on our study, we can't say confidently that the vaccine is highly effective in preventing you from getting infected and infecting others.
2/ Porque las cifras oficiales cuentan como UCI cualquier cama donde se puede instalar un respirador:
quirófanos, salas reanimación postquirúrgica, unidad coronaria, UCI pediátrica...
Y no cuentan que el 70-75% de UCIs de verdad suelen están ocupadas en periodos no pandémicos.
3/ Así que el número de personas ingresadas en UCI por #COVID19 en Madrid es mayor que el número de camas reales de UCI en cada hospital.
Por ejemplo:
"La Paz" tiene 30 camas de UCI y 31 ingresados por COVID-19
"12 de Octubre" tiene 24 camas de UCI y 34
ingresados pr COVID-19
New York and Madrid had similar epidemics until they spectacularly diverged.
In March, both cities were caught by surprise and shut down because of #COVID19.
In September, the situation is under control in NY and alarming in Madrid.
Why?
2/ Let’s start with the similarities: two big, dense cities with a large network of public transit and lots of visitors.
An explosive outbreak of #SARSCOV2 overwhelmed their contact tracing system and their hospitals. A lockdown was required to reduce the public health disaster.
3/ By June, both places had succeeded in bringing down the number of new cases. That's precisely what lockdowns do.
In July, new cases started to increase in Madrid until reaching one of the highest incidences in Europe.
New York has not seen any increase in new cases yet.