Dear friends,
I am conducting a research study about the probability of developing second primary cancers after the diagnosis of a certain index cancer. My primary outcomes are the Standardized incidence ratios (SIRs) and the Absolute Excess Risks (AERs) of developing the second primary cancers.
For some of the resulting second cancers, both SIR and AER are show values that are multiples of 1. For example, a certain second cancer (A) had an SIR of 14 and an AER of 10. Some other second cancer (B) showed an SIR of 8 while having an AER of 0.9. Accordingly, I would say that both cancers showed an association with the index cancer, but I wonder how the differences in AER will affect the magnitude of association, given that the p-value for both second cancers are less than 0.05.
Please advise.
Thanks
The reason why cancer B showed an SIR of 8 and an AER of 0.9 was just because it was a rarer cancer in the general population which you are using for the rate standardisation.
You are just using two different measures of association.
Standard significance tests can be applied to each measure.
It means that the risk of the second cancer(X say) in the index cancer group is less than the risk of X in the general population. Maybe the first cancer somehow inhubits the development of X (speculatively) or there is some other factor associated with first cancer which reduces risk of second cancer (eg treatment given to first cancer)
- Badges
- Science topic
- Similar topics
- Philosophy
- Esthetics
- Interpretation