3 ms·

I don't really understand any of your objections. > Also...They are measuring Vaccine Effectiveness not Efficacy Effectiveness is what we care about, i.e. rea

I don't really understand any of your objections. > Also...They are measuring Vaccine Effectiveness not Efficacy Effectiveness is what we care about, i.e. real world results. Unless you have some evidence to believe that the distribution chains between the vaccine providers were so imbalanced that one got tainted more than the rest. And if that's the case it would still make the result useful to decide which vaccine you want. > during March 11–August 15 (mentioned twice) Any study is going to be bound within a period of time. What makes you think this period was so special it would not be effective to present a comparison between the vaccines?

pbaka 5y ago
If I understood OC correctly, and from my own understanding of statitics : An arbitrary 30% exclusion rate precludes the effectiveness study from giving any useful information either way : 30% is much larger than a 95% CI. As such, that resulting numbers have no teleological value. A decision can't be made either way. > What makes you think this period was so special it would not be effective to present a comparison between the vaccines? TLDR : it says nothing because too many things varied during that interval which are interdependent time-wise. The starting assumptions are not valid at the end. This is whay the israelis only study short windows of time. ---- Discussion: The study covers only the US during the selected period. The vaccination rates varied , from 25% to 50% (https://usafacts.org/visualizations/covid-vaccine-tracker-states/ https://usafacts.org/visualizations/covid-vaccine-tracker-st...). Delta variant was at less than 20% early june, it is 100% now (https://covid.cdc.gov/covid-data-tracker/#variant-proportions https://covid.cdc.gov/covid-data-tracker/#variant-proportion...). If you are going by the latest numbers from Israel (https://www.medrxiv.org/content/10.1101/2021.08.24.21262423v1 https://www.medrxiv.org/content/10.1101/2021.08.24.21262423v...) which leads US and UK by two months in regards to the Delta variant, effectiveness drops dramatically for delta (~65% for avg. hospitalization endpoint, 42% for infection endpoint in men, 53% for women). Vaccinated people can be infected and retransmit the virus five times more. The Delta has a R0 value (numb. of people a sick person infects) of 6 to 9 regardless of vaccinal status (depending on the study), whereas the other variants had a R0 value with the vaccine which dropped to 1,2-1,6. One is highly exponential in all cases, the other far from quadratic with proper protective measures. Antibody Ct dissapears after six months, also according to the latest Israeli numbers (https://www.medrxiv.org/content/10.1101/2021.08.29.21262798v1 https://www.medrxiv.org/content/10.1101/2021.08.29.21262798v...). Consequently, the study treats the data as an ergodic process, whereas it is clearly non-ergodic : - transmissions can be represented as a graph - previous situation influences future situation depending on varying antibody titers - vaccination occurence itself changes the opportunity for transmission - transmission of different variants depends on encountering differing vaccinal statuses, at different times In conclusion, all the relevant variables have varied too much, and the basic mathematical pressupposition is wrong, whereas the study treats it mathematically as a snapshot uin time. It is completely wrong because of the underlying data: Vaccination rate changes dramatically, in relation to the most prevalent variant at any given time, with an immune response to each variant is extremely different, a different points in time for each vaccine, which itself drops over time. Add to that the above-mentionned exclusion of 30% of the measured population... The israelis on the other hand study shorter 1-2 month windows. It is also far from perfect, but a much closer to a snapshot, so you can disregard the nature of the process which is measured and some notable value remains in regards to making a decision.
belter 5y ago
Indeed they are measuring Effectiveness not Efficacy as they should. There I agree with you. My point is they excluded 32% of their universe due to being immune compromised. If you look at the link, references and quotes I provided the way Effectiveness is defined compared to Efficacy is exactly by including everybody. The second point of the study, that I interpret as showing a little lack of intellectual honesty, is that this exclusion provides for a higher effectiveness rate ( great for media quotes). Also as effectiveness is a measure that varies based on the time when you track it, they should have clarified that, and stated instead the vaccine was this effective from date so and so ( in this case March 11–August 15 ) What about I do a study like this: I let the immunity wane, measure effectiveness 1 year from now, then say Moderna or Pfizer vaccine effectiveness was 25%. In that case you can be sure everybody in the clinical expert group would be very precise in defining when effectiveness is supposed to be measured... :-) As the industry does not an have a standard for that period of time this game gets played around and the subtleties get lost in most media reports. And ultimately, where is the ethics of creating vaccine mandates that include everybody, as its being done in many countries, at the recommendation of experts like the ones responsible for this study? Because some of these mandates put at peril some people, even mentioning risk of legally loosing their jobs, while excluding the immune compromised from vaccine studies? Exclude them from vaccine studies? Then exclude them from the mandates.

pbaka 5y ago

If I understood OC correctly, and from my own understanding of statitics : An arbitrary 30% exclusion rate precludes the effectiveness study from giving any useful information either way : 30% is much larger than a 95% CI. As such, that resulting numbers have no teleological value. A decision can't be made either way. > What makes you think this period was so special it would not be effective to present a comparison between the vaccines? TLDR : it says nothing because too many things varied during that interval which are interdependent time-wise. The starting assumptions are not valid at the end. This is whay the israelis only study short windows of time. ---- Discussion: The study covers only the US during the selected period. The vaccination rates varied , from 25% to 50% (https://usafacts.org/visualizations/covid-vaccine-tracker-states/ https://usafacts.org/visualizations/covid-vaccine-tracker-st...). Delta variant was at less than 20% early june, it is 100% now (https://covid.cdc.gov/covid-data-tracker/#variant-proportions https://covid.cdc.gov/covid-data-tracker/#variant-proportion...). If you are going by the latest numbers from Israel (https://www.medrxiv.org/content/10.1101/2021.08.24.21262423v1 https://www.medrxiv.org/content/10.1101/2021.08.24.21262423v...) which leads US and UK by two months in regards to the Delta variant, effectiveness drops dramatically for delta (~65% for avg. hospitalization endpoint, 42% for infection endpoint in men, 53% for women). Vaccinated people can be infected and retransmit the virus five times more. The Delta has a R0 value (numb. of people a sick person infects) of 6 to 9 regardless of vaccinal status (depending on the study), whereas the other variants had a R0 value with the vaccine which dropped to 1,2-1,6. One is highly exponential in all cases, the other far from quadratic with proper protective measures. Antibody Ct dissapears after six months, also according to the latest Israeli numbers (https://www.medrxiv.org/content/10.1101/2021.08.29.21262798v1 https://www.medrxiv.org/content/10.1101/2021.08.29.21262798v...). Consequently, the study treats the data as an ergodic process, whereas it is clearly non-ergodic : - transmissions can be represented as a graph - previous situation influences future situation depending on varying antibody titers - vaccination occurence itself changes the opportunity for transmission - transmission of different variants depends on encountering differing vaccinal statuses, at different times In conclusion, all the relevant variables have varied too much, and the basic mathematical pressupposition is wrong, whereas the study treats it mathematically as a snapshot uin time. It is completely wrong because of the underlying data: Vaccination rate changes dramatically, in relation to the most prevalent variant at any given time, with an immune response to each variant is extremely different, a different points in time for each vaccine, which itself drops over time. Add to that the above-mentionned exclusion of 30% of the measured population... The israelis on the other hand study shorter 1-2 month windows. It is also far from perfect, but a much closer to a snapshot, so you can disregard the nature of the process which is measured and some notable value remains in regards to making a decision.

belter 5y ago

Indeed they are measuring Effectiveness not Efficacy as they should. There I agree with you. My point is they excluded 32% of their universe due to being immune compromised. If you look at the link, references and quotes I provided the way Effectiveness is defined compared to Efficacy is exactly by including everybody. The second point of the study, that I interpret as showing a little lack of intellectual honesty, is that this exclusion provides for a higher effectiveness rate ( great for media quotes). Also as effectiveness is a measure that varies based on the time when you track it, they should have clarified that, and stated instead the vaccine was this effective from date so and so ( in this case March 11–August 15 ) What about I do a study like this: I let the immunity wane, measure effectiveness 1 year from now, then say Moderna or Pfizer vaccine effectiveness was 25%. In that case you can be sure everybody in the clinical expert group would be very precise in defining when effectiveness is supposed to be measured... :-) As the industry does not an have a standard for that period of time this game gets played around and the subtleties get lost in most media reports. And ultimately, where is the ethics of creating vaccine mandates that include everybody, as its being done in many countries, at the recommendation of experts like the ones responsible for this study? Because some of these mandates put at peril some people, even mentioning risk of legally loosing their jobs, while excluding the immune compromised from vaccine studies? Exclude them from vaccine studies? Then exclude them from the mandates.