They are comparing breakthrough infections among vaccinated people to see which strains backbiter people are infected with and determined odds of infection post vaccination for each strain.
Correct except that if you get vaxxed you *increase* your odds of getting a mutant (vs alpha), and you do not increase those odds if you're recovered-immune. So that's also an inter-group comparison. This is key because obviously it's the mutants which are VOC, and because the vax should NOT be increasing your odds of contracting a (possibly more dangerous) variant. Nobody claimed they came up (or even looked for) a specific probability # for either group.
Yes, they are separately comparing two groups among themselves then comparing those results with each other. (Actually, three groups. They also looked at partially vaccinated.)
If you are vaccinated and you get infected your odds of getting a variant compared to alpha are X.
If you have disease induced immunity and you get reinfected your odds of getting a variant as compared to alpha are Y.
X and Y are NOT overall odds of contracting a covid variant, they are simply looking at people **within each group** who get covid and giving the ratios of strains of infection.
No, that is not what they did. They calculated a model of probability for VOC infection. The predictor were 4 kinds of immune status: naive, convalescent, one dose, fully vaxxed. Based on this status the probability was estimated on how good the immune status makes you guess the infection based on the data. These estimates are also changing over time and age so they included time and age as well. The model is filled with factors that explain all these differences in the probabilities of correctly predicting the variant of infection between immune statuses. I am sorry mate but this study compares between all the groups. You are mixing the reported and non reported results of Odds Ratios comparisons with the actual given out differences. Most importantly here is that the comparisons of vaxxed and unvaxxed are NOT significant. The odd ratios overlap so that the possibility is not ruled out that their
study only accidentally found these differences and the numbers are similar in the real world.
All included groups were people that got infected. There are no cases included where people have not been infected. In the study you find 'OR' written which stands for Odds ratios which is a measurement for the probability of a given estimate in regards to an influencing factor, i.e. OR give us information on how strongly two variables are connected, in this case the connection of the immunestatus and the variant were reported. You also find "logistic regression" which is the statistical interpolation of odds for the probabilities of the included variables, very simplified. You also find the 'CI' = confidence intervals which are statistical estimates on the possible range of the calculated odds ratios in the real word, differing from the calculated relations from the model. You can see some reported ORs in Fig. 2. There you find the ORs for immunestatus and variant relative to naive individuals and the alpha variant. That means: When looking from the relations of the alpha variant, the probability to be infected with Delta was the same (roughly 1) for e.g. convalescent individuals in comparison with naive ones. This differs when you swap convalescent individuals with vaccinated individuals. The OR raises. That means: From the odds of the alpha variant these odds differ in a way that can be best explained by a raise of delta infection probability for vaccinated individuals; the OR is higher than 1, roughly 2, meaning very simplified that for one naive individual you find probably 2 vaccinated individuals that are infected. This increase in probability was significant which means that the probability of this difference to be real is greater than our standard estimates of chance. You are right to stress the point that the naive infections have been the basis for all OR here, but that does not limit the significance of the vaccinated individuals being more probable infected than the unvaccinated ones. It is also important to stress that the real redults are: inside this data set the probability to be right when estimating the virus variant as Delta, Beta or Gamma is higher for vaccinated individuals when compared to naive ones. But meaningful for the actual question vaccine efficacy against infection: No significant difference for naive and convalescent immunestatus vs significant higher probability of infection for vaccinated.
The only odds they calculated were within the groups. Just because a vaccinated individual is more likely to catch a variant over the original doesn't mean that they are more likely to catch covid than someone with natural immunity. They didn't calculate overall odds of getting covid or whether one group was more likely than another to get covid.
Please indicate WHERE IN THE STUDY you think it indicates what you're saying and I'll go look again.
An increased risk compared to other vaccinated people...
"after vaccination" -- means they are comparing the same person before and after vax
No it doesn't and that's not what they did.
what are they comparing the after vax status to? LOL
They are comparing breakthrough infections among vaccinated people to see which strains backbiter people are infected with and determined odds of infection post vaccination for each strain.
Correct except that if you get vaxxed you *increase* your odds of getting a mutant (vs alpha), and you do not increase those odds if you're recovered-immune. So that's also an inter-group comparison. This is key because obviously it's the mutants which are VOC, and because the vax should NOT be increasing your odds of contracting a (possibly more dangerous) variant. Nobody claimed they came up (or even looked for) a specific probability # for either group.
"if you get vaxxed you *increase* your odds of getting a mutant (vs alpha), and you do not increase those odds if you're recovered-immune"
If a vaccinated person gets covid it's more likely to be a variant.
If a recovered-immune person gets covid it's as likely to be a variant as the original.
There's no comparison in this study of which of the two groups is more likely to catch covid overall.
Yes, they are separately comparing two groups among themselves then comparing those results with each other. (Actually, three groups. They also looked at partially vaccinated.)
If you are vaccinated and you get infected your odds of getting a variant compared to alpha are X.
If you have disease induced immunity and you get reinfected your odds of getting a variant as compared to alpha are Y.
X and Y are NOT overall odds of contracting a covid variant, they are simply looking at people **within each group** who get covid and giving the ratios of strains of infection.
No, that is not what they did. They calculated a model of probability for VOC infection. The predictor were 4 kinds of immune status: naive, convalescent, one dose, fully vaxxed. Based on this status the probability was estimated on how good the immune status makes you guess the infection based on the data. These estimates are also changing over time and age so they included time and age as well. The model is filled with factors that explain all these differences in the probabilities of correctly predicting the variant of infection between immune statuses. I am sorry mate but this study compares between all the groups. You are mixing the reported and non reported results of Odds Ratios comparisons with the actual given out differences. Most importantly here is that the comparisons of vaxxed and unvaxxed are NOT significant. The odd ratios overlap so that the possibility is not ruled out that their
study only accidentally found these differences and the numbers are similar in the real world.
Please indicate where in the study they calculated overall odds of any of the groups catching covid.
All included groups were people that got infected. There are no cases included where people have not been infected. In the study you find 'OR' written which stands for Odds ratios which is a measurement for the probability of a given estimate in regards to an influencing factor, i.e. OR give us information on how strongly two variables are connected, in this case the connection of the immunestatus and the variant were reported. You also find "logistic regression" which is the statistical interpolation of odds for the probabilities of the included variables, very simplified. You also find the 'CI' = confidence intervals which are statistical estimates on the possible range of the calculated odds ratios in the real word, differing from the calculated relations from the model. You can see some reported ORs in Fig. 2. There you find the ORs for immunestatus and variant relative to naive individuals and the alpha variant. That means: When looking from the relations of the alpha variant, the probability to be infected with Delta was the same (roughly 1) for e.g. convalescent individuals in comparison with naive ones. This differs when you swap convalescent individuals with vaccinated individuals. The OR raises. That means: From the odds of the alpha variant these odds differ in a way that can be best explained by a raise of delta infection probability for vaccinated individuals; the OR is higher than 1, roughly 2, meaning very simplified that for one naive individual you find probably 2 vaccinated individuals that are infected. This increase in probability was significant which means that the probability of this difference to be real is greater than our standard estimates of chance. You are right to stress the point that the naive infections have been the basis for all OR here, but that does not limit the significance of the vaccinated individuals being more probable infected than the unvaccinated ones. It is also important to stress that the real redults are: inside this data set the probability to be right when estimating the virus variant as Delta, Beta or Gamma is higher for vaccinated individuals when compared to naive ones. But meaningful for the actual question vaccine efficacy against infection: No significant difference for naive and convalescent immunestatus vs significant higher probability of infection for vaccinated.
The only odds they calculated were within the groups. Just because a vaccinated individual is more likely to catch a variant over the original doesn't mean that they are more likely to catch covid than someone with natural immunity. They didn't calculate overall odds of getting covid or whether one group was more likely than another to get covid.
Please indicate WHERE IN THE STUDY you think it indicates what you're saying and I'll go look again.