06-15-2021
Ely
Fluorite | Level 6
Member since
01-31-2018
- 6 Posts
- 1 Likes Given
- 0 Solutions
- 1 Likes Received
-
Latest posts by Ely
Subject Views Posted 648 09-14-2020 12:33 AM 788 08-12-2020 10:06 AM 842 08-12-2020 07:03 AM 815 12-03-2019 08:49 AM 851 11-27-2019 03:18 AM 940 11-26-2019 07:43 AM -
Activity Feed for Ely
- Posted Re: Model specification in Proc mixed for time interval dependent variable on Statistical Procedures. 09-14-2020 12:33 AM
- Liked Re: Model specification in Proc mixed for time interval dependent variable for SteveDenham. 09-14-2020 12:30 AM
- Got a Like for Re: Model specification in Proc mixed for time interval dependent variable. 08-12-2020 11:29 AM
- Posted Re: Model specification in Proc mixed for time interval dependent variable on Statistical Procedures. 08-12-2020 10:06 AM
- Posted Model specification in Proc mixed for time interval dependent variable on Statistical Procedures. 08-12-2020 07:03 AM
- Posted Re: Duplicate rows between a time period in transversal data on SAS Programming. 12-03-2019 08:49 AM
- Posted Re: Duplicate rows between a time period in transversal data on SAS Programming. 11-27-2019 03:18 AM
- Posted Duplicate rows between a time period in transversal data on SAS Programming. 11-26-2019 07:43 AM
-
Posts I Liked
Subject Likes Author Latest Post 1 -
My Liked Posts
Subject Likes Posted 1 08-12-2020 10:06 AM
09-14-2020
12:33 AM
Hello Steve, Many thanks for your help. The gamma function fit the data beter. Please how can I interpret the random effects? I have tried to interpret the covariance of the random intercept (when significant) but I am not sure if it is the best way. Regards
... View more
08-12-2020
10:06 AM
1 Like
Hello @SteveDenham, Thank you for your quick reply. A patient can be followed-up over time between different hospitals with the same unique ID. Regarding this statement: " Next, what about interactions between your independent variables? Are you willing to assume that males on tttA have the same effect on interval as females on tttB? If you can make that assumption, then the model statement looks OK. And last, how many unique patients are involved. To have pat as a continuous effect, you will almost certainly need to sort your data by hospital and patient." I cannot assume that males on tttA have the same effect on interval as females on tttB. I will test sex*treat interaction There are 455 distinct patients (350 on tttA and 105 on tttB) Yes my data are sorted by hospital and pat See below results with and without sex*treat interaction How to interpret the intercept ? Regarding this statement: "Next, I suspect that interval is not a variable with Gaussian error. Intervals like this give rise to Poisson distributed counts, and are generally gamma distributed. If that is the case, you will likely need to shift to PROC GLIMMIX. " Should I run the model using PROC GLMMIX with a gamma link and select the model with lower AIC? Regards
... View more
08-12-2020
07:03 AM
Hello All, I am trying to estimate for a given indication : 1. the effect of sex, age, treatment and dose on the interval between injections (of the treatment injected in the hospital) over time. 2. the effect of sex, age, treatment on the dose (qualitative, dosing interval) over time. I also want to include the hospital experience in the analysis. There are 2 options for this: nested the patient in the hospital or include the number of patients injected at hospital level and a dummy indicating whether the patient change the hospital or not. Patients don't receive the same number of injections (actually the number of injections varies from 1 to 19) and cannot switch between treatment over time. data Have;
*infile datalines ;
input pat hospital Sex $ Age_at_initiation treat $ change_hosp dose_units dose_interval $ Inj_nr date mmddyy10.;
format date mmddyy10.;
cards;
3 9 M 80 tttB 1 1600 Off-label 1 02/19/2015
3 13 M 80 tttB 1 1200 High 2 08/31/2015
3 13 M 80 tttB 1 1200 High 3 01/10/2016
1 1 M 69 tttA 1 100 Low 1 12/02/2012
1 1 M 69 tttA 1 100 Low 2 05/05/2013
1 1 M 69 tttA 1 200 Low 3 07/07/2013
1 2 M 69 tttA 1 200 Moderate 4 11/15/2013
1 2 M 69 tttA 1 300 Moderate 5 04/24/2014
2 6 F 25 tttB 0 300 Low 1 03/27/2016
2 6 F 25 tttB 0 500 Moderate 2 10/04/2016
;;;; For the first model, the dependent variable "interval" is the difference between consecutive date of injections. I have tried the following code: proc mixed data=have covtest ;
class treat sex hospital;
model interval = age sex treat dose_ interval inj_nr / CHISQ S covb ;
random int / subject=pat(hospital) type=UN G ;
run; My questions are: a. Is this model specification answer to my research question? b. How to validate and interpret the results? Thank you in advance for your time and help. Regards
... View more
12-03-2019
08:49 AM
Many thanks for your help
... View more
11-27-2019
03:18 AM
Hello, Thank you for your help. In July and August, I want to apply the previous product (ProdB). The rules are: - to apply the previous product when there is gap. The patient is at any time under treatment - to apply the dominant product when there is an overlap of 2 products in the same month: If prod A is taken 12 days and Prod B 19 days, then B is dominant Regarding @andreas_lds request, I am still trying what he suggested to me.
... View more
11-26-2019
07:43 AM
Hi everyone, I am having a trouble when trying to derive monthly evolution of patients per line of treatment. I would like, for a period between 2016/01/01 and 2019/07/31, to split each patient information per month in order to be able to only keep the month on target. Dataset Pat Poduct Line Line_first_date Line_end_date 1 ProdA 1 16/12/2010 23/02/2016 1 ProdB 2 10/03/2016 30/06/2017 1 ProdC 3 18/09/2017 20/01/2019 1 ProdD 4 05/03/2019 11/12/2019 Want Pat Month Month_start_date Month_end_date Product Line 1 1 01/01/2016 31/01/2016 ProdA 1 1 2 01/02/2016 29/02/2016 ProdA 1 1 3 01/03/2016 31/03/2016 ProdB 2 1 4 01/04/2016 30/04/2016 ProdB 2 1 5 01/05/2016 31/05/2016 ProdB 2 1 . . . . . 1 . . . . . 1 . . . . . 1 41 01/05/2019 31/05/2019 ProdD 4 1 42 01/06/2019 30/06/2019 ProdD 4 1 43 01/07/2019 31/07/2019 ProdD 4 Do anyone have an idea of how to process? Thank for your help.
... View more