I am not sure I follow this. This isn't my field and I have no idea what
a provider is. (Other people may have the same issue.) But reissuing
-duplicates- with the -provider- variable too might help. That is, it
sounds as if you seek duplicates on both -mrn- and -provider-.
Nick
[email protected]
Michelle Unterbrink
I'm calculating readmission rates for a hospital with many providers. I
have been able to successfully calculate overall readmission rates; but
I also need to calculate readmissions from a particular group of
physicians. In other words, I need to know interval between admissions
for patients that are discharged by a group of providers, then admitted
later to any provider.
My data is set up with rows corresponding to visits, and patients have
unique identifiers (mrn).
Interval between admissions:
bysort mrn (adm_date) : gen readmit_interval = (hours(adm_date -
dsc_date[_n-1])/24)
Number of times a patient is admitted:
duplicates tag mrn, generate (readmissions)
The preceding commands generate statistics for all admissions,
regardless of provider.
I generated a new variable to identify the providers of interest:
For 'provider' variable, 1 = a provider of interest, 0 = other providers
gen provider = (md_name==1|md_name==5...)
Using the provider variable, it's simple to calculate readmissions for
patients discharged from and readmitted to the providers of interest,
but this doesn't capture readmissions to the other providers:
bysort mrn (adm_date) : gen readmit_interval_hosp = (hours(adm_date -
dsc_date[_n-1])/24) if provider==1
How can I modify my readmission formulas to capture readmissions only
when a patient is discharged from a provider of interest? Any help would
be much appreciated!
*
* 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/
