How to test for trend of incidence rate over time?

[Ben Wyse]

Below is what I used but I am uncertain about. Can anyone tell me whether I am right? What is the type of trend test here (it said score test but I need more info such as whether it tries to fit to a linear trend)? What is the degree of freedom of the chi2? Is there other command to conduct a trend test?

Many thanks!

set up data

stset follow_up, failure(disease) id(study_id)

To calculate incidence rate by age group:

stptime, at (20(20)60)

To test for trend of incidence rate over time

stsplit age_group_by10, at(10 20 30 40 50)
stmh age_group_by10

Below is the result of stmh. I take that the trend is small (RR=1.028 increase per year) but highly significant (chi2=58.16, p<0.0001).

. stmh age_group_by10

         failure _d:  disease
   analysis time _t:  follow_up
                 id:  study_id

Score test for trend of rates with age_group_by10
  with an approximate estimate of the
  rate ratio for a one unit increase in age_group_by10

RR estimate, and lower and upper 95% confidence limits

        ----------------------------------------------------------
             RR       chi2        P>chi2      [95% Conf. Interval]
        ----------------------------------------------------------
          1.028      58.16        0.0000         1.021      1.036
        ----------------------------------------------------------

[Phil Bromiley]

You didn't get a quick answer. You will increase your chances of useful answer by following the FAQ on asking questions – provide Stata code in code delimiters, readable Stata output, and sample data using dataex.

I'm afraid that I do not do this kind of analysis so I really can't help you with the substance of your problem.