topic Re: trends over time in SAS Procedures

trends over time

EMB — Tue, 27 Jun 2017 18:12:25 GMT

I am sorry- I will try to explain.

The data analyzed was collected in 4 periods but in each period we collected data about different patients. I would like to assess if there is a significant change in age of first diagnosis in the various periods- for example, we have data on age of diagnosis of asthma for all the new patients in a defined community in 2008 and data on age of diagnosis of asthma for all the new patients in 2012 and in 2016. Mean age at each period was different and the difference is significant. My question is how can I evaluate if there is a pattern in the change (is the mean age increasing linearly over these periods?).

I hope this is more clear.

Thanks again!

Re: trends over time

Reeza — Tue, 27 Jun 2017 18:21:52 GMT

Because the participants are at different times I don't think you need to worry about the time component of this analysis and you can do standard statistical comparison to see if the average age is different between the groups. This would be done via an ANOVA.

The only reason time may need to be a factor is if you had a different criteria for joining at each time point.

Re: trends over time

EMB — Tue, 27 Jun 2017 18:29:29 GMT

Thanks!

Re: trends over time

Reeza — Tue, 27 Jun 2017 18:30:28 GMT

Sorry, that should read, different participants at different times. If it was the same participants at different times you'd have a different issue.