New package -rori- on the SSC

[Niels Henrik Bruun]

Thanks to Kit Baum, there is a new package -rori- on the SSC.

rori -- Immediate command for estimation of selection bias through relative odds ratio (ROR) with large sample confidence interval

Response bias is an often neglected consideration when evaluating the results of epidemiologic surveys, usually because of a paucity of information about nonrespondents. The effect of nonparticipation was described by a relative odds ratio (ROR), calculated as the OR (participants)/ OR (source population).
The relative odds ratio (ROR) is computed as the cross product ratio of the participation rates in the four exposure by outcome categories.

-rori- calculates the relative odds ratio (ROR) in a 2x2 table of outcome and exposure. -rori- also calculates the Prevalence ratios of the exposure groups. When exposure isn't binary, it is possible to specify N for the target and response populations.

References

• Austin MA, Criqui MH, Barrett-Connor E, Holdbrook MJ. The effect of response bias on the odds ratio. Am J Epidemiol. 1981;114:137–43
• Nohr EA, Frydenberg M, Henriksen TB, Olsen J. Does low participation in cohort studies induce bias? Epidemiology. 2006;17:413–8
• Nohr EA, Liew Z. How to investigate and adjust for selection bias in cohort studies.

[ericmelse]

Dear Niels,

After ssc installation, I tried to run the examples in the help file.
An error message is returned:

prop not found
r(111);

My Stata version is 17.0, Revision 10 May 2022
Possibly something is missing or some tweak is present in the ado file.

[Rich Goldstein]

Niels Henrik Bruun I notice that the third reference in #1 is incomplete and I would appreciate it if you would supply the rest of the reference

never mind - I see that the full cite is in the help file

[Niels Henrik Bruun]

Rich Goldstein The full third reference is Nohr EA, Liew Z. How to investigate and adjust for selection bias in cohort studies. Acta Obstet Gynecol Scand. 2018 Apr;97(4):407-416. doi: 10.1111/aogs.13319. Epub 2018 Mar 5. PMID: 29415329.

ericmelse I found quite an embarrassing bug. The bug is fixed and an update is sent to Kit Baum.

Thank you both for your comments. It is much appreciated.

[Niels Henrik Bruun]

Thanks to Kit Baum, there is an update fixing the bug mentioned by ericmelse

[Niels Henrik Bruun]

Thanks to Kit Baum, there is an update of -rori-.

The command now handles rows/m/exposure by columns/2/outcome or rows/m/exposure by columns.
In the first case, both ROR and RRF/PR by the outcome are reported in matrices.
In the latter, only RRF/PR are reported.

In the help, it is demonstrated how to use -rori- with datasets.