I have 4 populations of different races that I am following for a period of time. I am looking for the incidence of a particular condition (lets call it condition A) after an event x. (lets say joining a particular business)... I find that one particular population (lets call it population k) has significantly lower incidence of the condition A compared to other populations . Upon further analysis I realize that event x occurred in most members of the population k a lot later than the other three populations.. (ie most members of the business joined the business a lot later than other 3 groups).
I want to know whether the lower incidence of condition A is due to the late occurrence of event x or do they actually have a lower incidence.
How do I approach this problem? I am thinking of getting kaplan meier curves of condition A in all 4 subgroups..what do I do there after?
Kaplan-Meier analysis would work; you would have to construct the analysis so that the time axis is the time after exposure to event A, and not the time since first joining the study (unless you can construct your graph so that the time of exposure to 'A' is indicated on the graph at the same point for everyone). The log-rank test (and its variants) will determine which K-M survival curves are different from one another.
You should also consider person-years of exposure to event 'A' before the disease of interest occurs. But try to avoid the mistake of not allowing enough time for the disease of interest to manifest itself after the exposure---for example 30 people X 6months each = 15 person-years, but six months may not be enough time for the disease to develop biologically so using person years there would be misleading. What is the median time since exposure until the disease occurs? And the minimum time? You can compute incidence rates per year of exposure, and incident rate ratios between the four groups. You can stratify disease rates by time since exposure, for example
..... want to know whether the lower incidence of condition A is due to the late occurrence of event x or do they actually have a lower incidence.... this can be better addressed using Logistic Regression using 'time since exposure' as independent variable.
You should calculate observation period for each person (or person-year for each population) and then you should calculate relative risk of the effect A for each population using Logistic (or Poisson) Regression.
You can use available statistical program packages for example EPICURE (Preston D.L., Lubin J.H., Pierce D.A. et al. (1993). EPICURE User’s Guide. Seattle WA : Hirosoft Corporation) or other.
You can do a so called delayed-entry model, which accounts for the time just before the time at risk.
Look for Crowther et al. (2016) Joint modelling of longitudinal and survival data: incorporating delayed entry and an assessment of model misspecification. Stat Med. 2016 Mar 30;35(7):1193-209.
Analysis can be done with STATA. This works either for survival analysis and incidence estimate studies.
- Badges
- Science topic
- Similar topics
- Medicine
- Epidemiology