st: scoring system using marginally significant predictors

[Marcos Vinicius <bioestatistico@inbox.com>]

Hello ,
I would like  to share a strategy and have your opinion.
Backward elimination was used to reach the final multivariate  logistic regression model using p<0.10 as entry and stay level  From the final model  We then created a weighted scoring system by rounding all regression coefficients up to the nearest integer (ie, the smallest integer greater than the estimate). This method was based on the beta coefficients 
Once the final model was defined, we created integer weights for each variable. We calculated these weights by multiplying the model coefficients by 10.As like the article below we considered the statistically significant predictors and also the marginally significant predictors (0.05<p<0.10 ) to determine the scoring system. 

Is it ok or the scoring system should be strictly derivated using  only statistically significant predictors?
Thanks,
Ann Thorac Surg. 2008 Jun;85(6):1938-45. doi: 10.1016/j.athoracsur.2008.03.014.
Preoperative prediction of the occurrence and severity of complications after esophagectomy for cancer with use of a nomogram
Thanks,
Vinicius

____________________________________________________________
FREE 3D MARINE AQUARIUM SCREENSAVER - Watch dolphins, sharks & orcas on your desktop!
Check it out at http://www.inbox.com/marineaquarium



*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/