Todavía riendome de los que pensaron que las #restricciones iban a ser por 15 días jajajaja... No puedo dejar de reirme. #argentina
Ahh y ya pasamos a Suecia en muertos por millón y ellos ya pasaron la segunda ola...
Ahora si la población supiera que la forma de minimizar mortalidad es hacer #restricciones diferenciadas por 90 días y abrir todo para menores de 50/60 que tienen super baja mortalidad ya hubieran incendiado todo. Te engañaron como a una colegiala...
Y cuando digo TODO, es TODO... boliches, todo!!! Chau máscaras, chau todo... no queres que los vulnerables se sientan seguros en la calle. Y si vivis con un vulnerable, vos tampoco tendrías que sentirte seguro, así minimizas riesgo.
Cuanto más estiras el tiempo, más posibilidades le das a una variante del virus para que le gane la pulseada a las otras variantes, generando lo que está pasando hoy. Mirá que bonito como las nuevas variantes (más rápidas) toman prevalencia y estan haciendo desastre.
Inocentes palomitas despues de 1 año piensan que se puede controlar un virus respiratorio con Ro=4.3 (ahora). Y que testear te da algo más allá de gastar plata que podrías haber usado para reforzar el sistema hospitalario. Como dice @madorni: "Saluden a la Argentina que se va"
Pero bueno, si alguien con 2 dedos de frente hubiera hecho los cálculos necesario (que no eran rocket science) para darse cuenta de todo esto, no hubieramos tenido que escribir un technical report de 40 hojas con @LDjaparidze explicando todo esto y más. medrxiv.org/content/10.110…
Pero lo mejor despues escuchas a @edufeiok y @JonatanViale quizás 2 de los mejores periodístas de Argentina que invitan gente que habla de los números y ni siquiera se tomaron el laburo de contrastarlos con las estadísticas públicas.
1/n Epidemiological models are underdetermined. What @pjakma is doing is showing exactly that. You have to fight underdetermination with tooth and nails. I have wrote about this before, but as the video shows with that much variability it is pointless. Are there alternatives?
2/n Yes. There are 2 types of models, exploratory and predictive. Usually your model with fit one category, it is very unlikely your will fit in both. Fitted models help you in the predictive realm, but the key point is to realize that under-determination is a thing.
3/n That forces you to have to diminish the amount of moving parts, and try to remove as many high-dimensional stochastic inputs as possible through averaging. That's the approach we took on our work.
1/n A pedido de un amigo que me la pidió en español (no tengo muchos seguidores que puedan hacer uso) va mi análisis sobre la edad de las hospitalizaciones en AMBA. El original acá:
2/n Un amigo me preguntó muy preocupado hace unos días que hay de verdad sobre todos los médicos que salen a decir que cada vez hospitalizan gente más joven en Argentina/AMBA.
El diagnóstico: "Anecdotitis"
3/n Cuando alguien con entrenamiento científico es confrontado con una 'verdad', lo primero que hace es tratar de saber más al nivel de detalle. Ya se lo que estas pensando. Te digo que los científicos son una especie rara y en peligro de extinción (y no aparecen en televisión).
1/n A friend asked me a few days ago what was true about the saying of medics about younger people being hospitalized more often recently in Argentina.
The results are in and the diagnosis is: "Anecdotitis".
2/n Like any scientists that is confronted with some 'truth', you try to figure out the nitty gritty details of it. I know what you are thinking, but scientists are a pretty rare species nowadays. I would say an endangered species.
3/n The first thing anyone with at least 2 fingers of forehead is to go to the case file and look at the distribution of death. Well, not unexpected AT ALL.
1/n If you are going to call someone an antivaxxer just make sure that the person in question may know *a bit* more than you about the topic in question. And can actually show proof he is NOT. Think once, think twice, THINK before embarrassing yourself!!!
2/n But you know, I may have converted to being an antivaxxer recently (that would be a good escape). I just think that what antivaxxer parent would vaccinated their kid 2 weeks ago on the 5th of April of 2021
3/n Why do I do this? Why do I willingly show the vaccination history? Because I am tired of people that has not done the research to judge in binary terms something that it is not. If you haven't done the research it is on you, not me.
1/n I find it interesting the poor understanding of risk by medical professionals. Just for the record, I am pro-vaccines, but I am anti-ignorance more. If you have risk from 2 independent events (covid positive and vaccination) the risk gets added.
2/n From probability we know that: P(A or B) = P(A) + P(B) + P(A and B)
For context, what is the probability of dying from COVID-19 or dying from Vaccine Adverse reaction.
3/n Purist would say then that we need to take the conditional probability, so the formula would be:
P(dying | COVID-19 or Vaccine) = P( dying | COVID-19 ) + P( dying | Vaccine ) + P( dying | COVID-19 and Vaccine )
Can someone tell me if I am getting this right? If I haven't been proved positive of SARS-Cov-2 by either PCR or Antibodies authors can still count me as #LongCOVID anyways and Journal editors disagree with that? Mhhh, that sounds like Journal doing some science for change.
Apparently I was getting it right. So 39% of the sample was actually positive for SARS-Cov-2, while probably the whole sample was positive for "Probable Acute Viral Infection". NOW, the last phrase in statistics have a known name: "Crapping on your sample".
But let's follow on the 'false-negatives' if 61% of your sample are false-negative, can anyone explain to me: Why are we testing?