Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: st: Unique Clients with Multiple IDs that need to be integrated
From
"Justina Fischer" <[email protected]>
To
[email protected]
Subject
Re: st: Unique Clients with Multiple IDs that need to be integrated
Date
Fri, 12 Oct 2012 00:45:39 +0200
Hi Emily,
this is a very long account.....please apologize my frankness, but what is your question? Your data look like perfect panel data to me now (with a unique id and time variable = day_year)
HTH
Justina
-------- Original-Nachricht --------
> Datum: Thu, 11 Oct 2012 15:19:52 -0700
> Von: "Emily Putnam-Hornstein" <[email protected]>
> An: [email protected]
> Betreff: st: Unique Clients with Multiple IDs that need to be integrated
> I am working with a large substance abuse treatment dataset with roughly
> 900,000 observations for approximately 200,000 unique clients.
>
> In working with these data (these records were being linked to a second
> dataset), it was discovered that a number of clients had been
> inadvertently
> assigned a new id when returning for a follow-up treatment service. In
> other
> words, the agency thought it was linking/working with dataset of unique
> clients, only to discover that many records were clearly for the same
> individual.
>
> To complete the data linkages the decision was made to maintain one id/one
> record for each client, but to create up to 10 new variables with the
> alternative ids listed so that after the linkages were completed, data for
> these other ids could be pulled to construct a longitudinal record of
> service encounters for a given individual. They were attached to a
> client’s
> record as depicted in the data below. So, for example, the first client
> (cid1=1) was not found to have any duplicate ids, so fields altid1=cid1
> and
> altid2-altid10 are all empty (I have only shown alt1-altid3 for
> simplicity).
> But for the second client (cid1=2) it was discovered that this client has
> information under two other ids “65958” and “62781”
>
> To be able to construct a longitudinal record of service encounters for
> each
> client I need to recode the records currently appearing as cid1=65958 and
> cid1=62781 to cid1=2 (and I need to do this for up to 10 ids for some
> clients). Additionally, the correct client id is listed as an alternative
> id
> for those records I want to recode. So for cid1=65958, altid1=2 and
> altid2=62781 (I have also shown this below, under “Later Data”).
>
> Finally, I mentioned the data linkages earlier because it is important
> that
> I maintain the first (as sorted) cid1 as that is the id that links to
> other
> data source – although it is not necessarily the id associated with a
> client’s first admission.
>
> I will not burden the listserv with the completely inelegant code I have
> been working on for the last two days (consisting of attempts to reshape,
> merge subsets of records, etc. and has gotten me nowhere). Nor will I
> attempt to justify decisions that were made in creating this dataset that
> was given to me…but would greatly appreciate any and all guidance! Thank
> you
> and apologies in advance if I have explained this poorly. Emily
>
>
> +---------------------------------------------------------------------------
> -----------------+
> | CLIENT modality_type adm1
> dis1 cid1
> altid1 altid2 altid3 |
>
> |---------------------------------------------------------------------------
> -----------------|
> 1. | 1 Admission 12 Dec
> 06 . 1
> 1 . . |
> 2. | 1 Discharge 12 Dec 06
> 31mar2007 1
> 1 . . |
>
> |---------------------------------------------------------------------------
> -----------------|
> 3. | 2 Admission 08 Mar
> 07 . 2
> 2 65958 62781 |
> 4. | 2 Discharge 08 Mar 07
> 17mar2007 2
> 2 65958 62781 |
> 5. | 2 Admission 14 Jun 07
> . 2
> 2 65958 62781 |
> 6. | 2 Discharge 14 Jun 07
> 30jun2007 2
> 2 65958 62781 |
>
> |---------------------------------------------------------------------------
> -----------------|
> 7. | 3 Admission 14 Jul
> 06 . 3
> 123171 3 . |
> 8. | 3 Discharge 14 Jul 06
> 03aug2006 3
> 123171 3 . |
> 9. | 3 Admission 02 Oct 06
> . 3
> 123171 3 . |
> 10. | 3 Discharge 02 Oct 06
> 22oct2006 3
> 123171 3 . |
> 11. | 3 Admission 30 Apr
> 07 . 3
> 123171 3 . |
> 12. | 3 Discharge 30 Apr 07
> 20may2007 3
> 123171 3 . |
>
> |---------------------------------------------------------------------------
> -----------------|
> 13. | 4 Admission 07 Mar
> 07 . 11
> 11 15262202 . |
> 14. | 4 Discharge 07 Mar 07
> 27mar2007 11
> 11 15262202 . |
>
> |---------------------------------------------------------------------------
> -----------------|
> 15. | 5 Admission 20 Dec
> 06 . 13
> 66234 13 172805 |
> 16. | 5 Discharge 20 Dec 06
> 07jan2007 13
> 66234 13 172805 |
> 17. | 5 Admission 01 Mar
> 07 . 13
> 66234 13 172805 |
> 18. | 5 Discharge 01 Mar 07
> 12mar2007 13
> 66234 13 172805 |
>
> |---------------------------------------------------------------------------
> -----------------|
> 19. | 6 Admission 13 Feb
> 07 . 31
> 14997478 31 190690 |
> 20. | 6 Discharge 13 Feb 07
> 06mar2007 31
> 14997478 31 190690 |
> 21. | 6 Admission 22 Jun
> 10 . 31
> 14997478 31 190690 |
> 22. | 6 Discharge 22 Jun 10
> 23sep2010 31
> 14997478 31 190690 |
>
> (LATER DATA)
>
>
> +---------------------------------------------------------------------------
> ----------+
> | CLIENT mo~y_type adm1
> dis1 cid1 altid1
> altid2 altid3 |
>
> |---------------------------------------------------------------------------
> ----------|
> 67821. | 14922 Admission 19 Dec 06 .
> 65958 2
> 65958 62781 |
> 67822. | 14922 Discharge 19 Dec 06 30jan2007
> 65958 2
> 65958 62781 |
> 67823. | 14922 Admission 08 Feb 07 .
> 65958 2
> 65958 62781 |
> 67824. | 14922 Discharge 08 Feb 07 12mar2007
> 65958 2
> 65958 62781 |
>
> +---------------------------------------------------------------------------
> ----------+
>
>
>
>
> *
> * 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/
*
* 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/