Re: st: multiple imputation and propensity score

[Stefano Di Bartolomeo <stefano.dibartolomeo@uniud.it>]

First of all thank you very much for your help.

As you supposed, after calculating the PS I do some 1:1 matching and then I run several models of Cox regression for various outcomes, stratified on matched pairs. I am very curious to learn if all this would be feasible carrying forward all the imputed data sets. For sure, it is beyond my skills. I vaguely surmise that perhaps it could be possible if I used logistic regression instead of Cox regression, but I am not sure and in any case I must use survival analysis. Moreover, how could I draw a graphic of the density distribution of PS in the two groups before and after matching without having a unique PS?
Thank you again.
Stefano

Il giorno 24/ago/2011, alle ore 16.13, Stas Kolenikov ha scritto:

> Can you follow through your analysis in the multiple imputation
> framework? The propensity score will probably go into some regression
> or matching exercise; can you perform these with -mim-? That would be
> the approach most closely consistent with MI framework.
> 
> On Wed, Aug 24, 2011 at 5:28 AM, Stefano Di Bartolomeo
> <stefano.dibartolomeo@uniud.it> wrote:
>> Dear Statalist members
>> 
>> I am doing a study that compares survival after 2 types  of cardiological treatments angioplasty or by-pass. To limit confounding I use a propensity score, calculated as usual by a logistic model. A few covariates had missing values, which I imputed with ICE (5 imputations), separately for each group of treatment. Then I joined again the records in 1 file with 'append' and so have a file with N*6 observations. Then I calculate the propensity score :
>> mim: logistic angioplasty_vs._bypass  + other_covariates
>> Finally I obtain the propensity score with
>> mim: predict pscore
>> As expected, I have 6 sets of propensity scores, one for each set of imputed data (_mj = 1-5) plus the one (_mj = 0)  resulting from the combination of imputed estimators  according to Rubin's rules. Unfortunately, the propensity score of the set _mj = 0 (which is the one I would think correct to use for further analyses) makes no sense, being virtually the same in patients treated with angioplasty or by-pass. The propensity scores of the imputed sets _mj 1-5 instead are ok and distributed as expected in the two treatment groups. I could easily pick up one of this well-working propensity scores for further use, but I know it is not correct. Has anybody ever encountered such a problem? Is it normal that the application of Rubin's rules results in a virtually useless propensity score? If so, how can one properly calculate propensity scores with multiply imputed data-sets?
> 
> -- 
> Stas Kolenikov, also found at http://stas.kolenikov.name
> Small print: I use this email account for mailing lists only.
> 
> *
> *   For searches and help try:
> *   http://www.stata.com/help.cgi?search
> *   http://www.stata.com/support/statalist/faq
> *   http://www.ats.ucla.edu/stat/stata/


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