Interpreting meaning in influenza vaccination study results

[DBecker1988]

Here is the study: Evidence of bias in estimates of influenza vaccine effectiveness in seniors

The authors hypothesis is that those who get the influenza vaccination are more likely to be heath conscious individuals which might account for lower incidence of infection. I.E. healthy elders are motivated to be healthy and make healthy choices such as eating right, exercising, and getting their yearly vaccination. From this notion the author hypothesis that the main reason they are less likely to get sick is not from the vaccination, but rather from the other measures these people are taking to be healthy (exercise and diet)

Thats all fine and dandy in theory, but I am having some issues interpreting how the results of the authors statistical analysis correlates with the conclusions. Perhaps someone who is more statistically inclined than myself can help by explaining the meaning of the results:


Results: The relative risk of death for vaccinated persons compared with unvaccinated persons was 0.39 [95% confidence interval (95% CI), 0.33–0.47] before influenza season, 0.56 (0.52–0.61) during influenza season, and 0.74 (0.67–0.80) after influenza season. The relative risk of pneumonia hospitalization was 0.72 (0.59–0.89) before, 0.82 (0.75–0.89) during, and 0.95 (0.85–1.07) after influenza season. Adjustment for diagnosis code variables resulted in estimates that were further from the null, in all time periods.

• What is the meaning of the numbers 0.39, 0.56, 0.72 etc.?
• How did these numbers come to be?
• Lastly, what does it mean that they consistently show the tendency to approach 1 with the progression of time? (see also table 5 on page 342)

One other confusing issue:

• Why is it that "healthy individuals" represent a demographic with more comorbidities? Isn't the reverse true? Perhaps I'm misinterpreting the data.

[CharonY]

Odd I could have sworn that I have posted here already. In any case, I still have not found the time to read the paper, but the numbers refer to relative risks. In this case it is the ratio of deaths in vaccinated vs non-vaccinated individuals at the given time points. Thus 1 would indicate that the risk is identical in bots populations, whereas 0.39 indicates that a vaccinated person would have 2.5 lower risk than non-vaccinated people.

 

 

Lastly, what does it mean that they consistently show the tendency to approach 1 with the progression of time? (see also table 5 on page 342)

 

I would have to read it to be certain but it could mean that over time, the relative risk of both population become identical. Judging from OP I would say that the researchers indicate that the differences in relative risk of death is not due to immunity to flu as the strongest effects do not seem to coincide with the actual occurrence of flu. Again, solely based on the OP, not based on actually reading the paper.

 

To me, that would not be terribly surprising as the general risk from dying of flu (even if infected) is generally low (for industrialized areas, at least). So I would not be surprised that flu vaccination protects from getting the flu, rather than from dying (again, the presented data refers to death and not to getting sick).