Sorry, I sent the CPT codes to the participating sites' contacts individually.

To share with the whole group, the CPT codes for amputation of lower limb is 
27888, 27889, 27880-27886.
Yes, I have reviewed the sites' query reports and had numerous email exchanges 
with the sites' contacts. Some learning points from the experience:

1.      The queries were built on KUMC terminology instead of master GPC 
terminology, resulted in query variations across sites. As an example, not all 
sites have clinical service information, thus size of the base population 
(primary care) is not comparable across sites. Also, some sites used diagnoses 
ICD9 V-codes or status codes to encode lower limb amputation, which is not 
correct because it specifies history of amputation that are frequently found in 
problem list (actual date and location of the service cannot be reliably 
determined). Not all sites have CPT codes loaded in i2b2 (MU and IU). MU is 
expected to load it into i2b2 by mid-September.

2.      UTSW had a specific problem with inclusion of their affiliated Parkland 
hospital where they only receive CPT codes instead of inpatient procedure 
codes, yielding larger denominator (Q2 = eligible patients) but smaller 
numerator (Q4 = patients had lower limb amputation procedure). Inclusion of CPT 
codes increased amputation numbers significantly for UTSW. It also increased 
the number a little for other sites. This may be due to physicians performing 
amputation procedures at facilities not owned by the site.

3.      The queries were complex with multiple panel queries and time specific 
querying.

With the updated queries, NYC CDRN's updated number seemed consistent with 
other sites (KUMC, MU, MCW). IU's updated number is half of what was originally 
reported (rate is reduced significantly), but is still much higher than other 
sites. UTSW's amputation numbers still seem to be off. By limiting the 
denominator to outpatient UTSW clinics, their base population denominator 
dropped significantly, from >40,000 to ~9500, while all other centers have at 
least 11,000 patients. Given the size of UTSW, it is hard to believe the 
number. Further investigation is still needed, but we may not have the time for 
this proposal submission.

Thanks,

Mei

From: Dan Connolly
Sent: Monday, July 25, 2016 11:24 AM
To: Mei Liu; <[email protected]>
Subject: RE: Feasibility Query: National Diabetes Amputation Prevention Study 
(NDAMPS)

Mei, you were going to send the CPT codes. Did I miss them? Perhaps you sent 
them somewhere other than gpc-dev? Please share with the whole group.

Did you get the printed queries from MU, MCW, UTSW, and IU?

Have you resolved the issue of very high and very low rates?

--
Dan
________________________________
From: 
[email protected]<mailto:[email protected]> 
[[email protected]] on behalf of Mei Liu [[email protected]]
Sent: Monday, July 18, 2016 10:26 AM
To: <[email protected]<mailto:[email protected]>>
Subject: Feasibility Query: National Diabetes Amputation Prevention Study 
(NDAMPS)
Thank you all for submitting the patient count for the NDAMPS feasibility query.

Based on the submitted numbers, we observed very large variations in the 
amputation rates between participating sites (KUMC, MU, MCW, UTSW, IU, and NYC 
CDRN), with UTSW and NYC being very low, IU being very high, and KUMC, MCW, and 
MU being in the middle. Before we can perform power calculation with 
confidence, we need to make sure that the queries were ran consistently and 
whether the sites are counting amputations differently.

This is urgent because we cannot submit the proposal (due July 31) without a 
reliable power calculation. I hope to discuss this over the dev call tomorrow.

Thanks,

Mei
-------------------------------------------------
Mei Liu, PhD
Assistant Professor
Department of Internal Medicine
Division of Medical Informatics
University of Kansas Medical Center
Office: 913-945-6446
Fax: 913-588-4880

_______________________________________________
Gpc-dev mailing list
[email protected]
http://listserv.kumc.edu/mailman/listinfo/gpc-dev

Reply via email to