RE: st: RE: creating an index variable: number of presentations to hospital by participant

[Alison McCarthy <statslearner@live.com>]

Dear Nick,

This is exactly what I was looking for thanks! I had the bysort patient_id year arrivaldate, but was missing the brackets.
Thanks for sending the reading material too.
Really appreciate your advice
Alison 

> From: n.j.cox@durham.ac.uk
> To: statalist@hsphsun2.harvard.edu
> Date: Thu, 28 Oct 2010 10:54:06 +0100
> Subject: RE: st: RE: creating an index variable: number of presentations to hospital by participant
> 
> That is, 
> 
> bysort patient_id year (arrivaldate): gen order_in_year = _n 
> 
> A tutorial on -by:- is included within 
> 
> SJ-2-1  pr0004  . . . . . . . . . . Speaking Stata:  How to move step by: step
>         . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  N. J. Cox
>         Q1/02   SJ 2(1):86--102                                  (no commands)
>         explains the use of the by varlist : construct to tackle
>         a variety of problems with group structure, ranging from
>         simple calculations for each of several groups to more
>         advanced manipulations that use the built-in _n and _N
> 
> which is accessible to all via the Stata Journal website. I got the above text (and other hits) within Stata by 
> 
> . search by, sj 
> 
> -- and if you do that yourself you will get a clickable link to the corresponding .pdf. 
> 
> Nick 
> n.j.cox@durham.ac.uk 
> 
> Mitch Abdon
> 
> You may add the year variable in the -by- prefix. You need to include
> the year variable in the -sort- as well. See "help by".
> 
> Alison McCarthy
> 
> > regarding the formula below. Is there a way to also split this up by year of presentation?
> > This means you have the number of ordered presentations for each participant calculated separately for each year.
> 
> Kieran McCaul
> 
> >> You'll need some sort of patient ID that identifies individuals.
> >> Assuming you have that, then:
> >>
> >> sort patient_id arrivaldate, stable
> >> by patient_id: gen order=_n
> >>
> >>
> >> look at -help sort- for an explanation of the option -stable-.
> >>
> >> Without this option, sort will randomly order observations with the same
> >> patient_id and arrivaldate.
> >>
> >> Of course, if you have the arrival time as well, you could sort on:
> >>
> >> sort patient_id arrivaldate arrivaltime, stable
> 
> Alison McCarthy
> 
> >> I am working with a large data file containing presentations to a
> >> hospital emergency department (ED). A number of individuals make more
> >> than one presentation, which I have seen via the command 'duplicate'.
> >> The organisation of my data is below:
> >>
> >> ARRIVALDATE repeaters part_ID age DISCHARGEDATE
> >> 30-Mar-10 0 70 50 30-Mar-10
> >> 30-Mar-10 0 70 50 30-Mar-10
> >> 31-Mar-10 0 70 50 31-Mar-10
> >> 01-Apr-10 0 70 50 01-Apr-10
> >>
> >> As can be seen, this person presented twice to the ED on the one day;
> >> the second person presenting twice in a two-day period.
> >>
> >> I wish to create a new variable which tells me the number and order (in
> >> terms of date) of presentations for each participant, and would greatly
> >> appreciate some guidance.
> 
> *
> *   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/