CausalCompV1Readme.txt for CausalComp.V1.xls

The Excel-spreadsheet allows easy calculation of several summary estimates for an exposure of interest - outcome association in the presence of a dichotomous covariate which can be a confounder or not and an effect-measure modifier (multiplicative scale) or not.

The spreadsheet "CausalComp.V1.xls" is offered as open code under the conditions of the GNU-GPL license (Free Software Foundation. The GNU General Public License, accessed June 8, 2005 (www.gnu.org/licenses/licenses.html)) and can be downloaded at www.brighmanandwomens.org/pharmacoepid free of charge.
The spreadsheet is self-explanatory but we strongly recommend reading and citing the manuscript "Insights into Different Results from Different Causal Contrasts in the Presence of Effect-Measure Modification" by Til Stürmer, Kenneth J Rothman, and Robert J Glynn, published in Pharmacoepidemiology and Drug Safety, 2006; 15: 698–709.

When you open the spreadsheet, you will be asked whether you want to enable the macros contained in the spreadsheet. You should enable the macros so that the spreadsheet can work properly.

Instructions of use are in the blue box. Enter your parameters in the YELLOW cells. The spreadsheet will then calculate the stratum-specific cell counts (ORANGE cells) and relative risks as well as the Mantel-Haenszel, Maximum Likelihood (iterative process), Standardized Mortality/Morbidity Ratio, Propensity Score matched, and Inverse of Probability of Treatment Weighted summary estimates (GREEN cells). 

We have tested the spreadsheet extensively. If you should notice any errors or have suggestions for improvement, I would appreciate if you could contact me (til.sturmer@post.harvard.edu).

Boston, March 15. 2006 - Til Stürmer