Thanks very much Tamara. This is very helpful.

On April 1st, the file was downloaded around 10 am Mountain Standard Time (approximately 1pm EDT). On July 15th, I downloaded them around 4pm EDT. As far as I remember, both times I got them from the CQM library on cms.gov. (http://www.cms.gov/Regulations-and-Guidance/Legislation/EHRIncentivePrograms/eCQM_Library.html)

Tamara,

CMS and ONC are working on a plant to prevent this situation from occurring again. Could you provide insight on what time and from which site the files were downloaded on April 1st and July 15th?

Thank you in advance for your help,
Sharon

FYI - I compared all of the measures from the April 1st file to the one downloaded today.

SCIPInf1 showed file changes, but they were underlying the xml, not changes to the actual logic or text

SCIPInf2 shows changes to the logic, mostly it seems that the denominator exceptions are in different order for every population. I've attached the April 1st copy of this measure as well.

Thanks... also, the version we downloaded on April 1st had Excel files with all of the valuesets and all values that were on the valuesets in each measure. These are no longer present in the one I downloaded today. These were VERY useful to us, I'd like to recommend that you continue to provide those.

We are currently looking into a cause for this discrepancy and will post an additional reply once determined.

These are the files that we downloaded on April 1, 2013

we downloaded a copy of the specifications file for the EH measures on 4/1/13 when the april specifications were released. That file had the line mentioned above.

I checked the one out on the CQM library website today and confirmed it no longer contains the mention of valueset 2.16.840.1.113883.3.666.5.1105.

This concerns me... when was this change made? Why were we not informed of this change? What other changes have been made that we should be aware of?

ADDITIONAL INFORMATION: After filing this comment I did a comparison of the file from April 1st with the one I downloaded today. There were changes to:
the References
the human readable Initial Patient Population text
the human readable Denominator text
the human readable Denominator Exclusion text
addition of the condition (length of stay <= 120 day(s)) in the initial patient population on both population 1 and 2
The name of the valueset in the April 1st doc was slightly different than the one from today so the logic lines pertaining to chest xray and ct scans were slightly different in text

I'll attach the file from April 1st for this measure to this issue report, so you can see it for comparison to what you have now.

Thank you for addressing this concern. A review of the April posting in eSpec Navigator indicates that this new value set is incorporated in the measure and is contained within the data criteria.

A review of VSAC indicates the updated value set for version 3 of the measure (Hospital Meaures-Radiologic Finding-Normal SM-CT: 2.16.840.1.113883.3.666.5.2365) includes the correct codes. The previous value set (Hospital Measures-Radiological findings normal SNOMED-CT Value Set: 2.16.840.1.113883.3.666.5.1105) includes the correct codes and is present for reference to version 1/2.

Could you please indicate where you are viewing the following attribute as described below : Attribute: "Result: Hospital Measures-Radiological findings normal" using "Hospital Measures-Radiological findings normal SNOMED-CT Value Set (2.16.840.1.113883.3.666.5.1105)"

Actually, my whole point with this is that the LOGIC was changed, but if you check in the data criteria section of the april specs, this line is present:

•Attribute: "Result: Hospital Measures-Radiological findings normal" using "Hospital Measures-Radiological findings normal SNOMED-CT Value Set (2.16.840.1.113883.3.666.5.1105)"

It did NOT contain the new valueset OID here, it contained the old one.

Upon further review of this measure we have determined that some data elements can be difficult to capture. After discussion with the measure steward, it was decided to remove the radiologic findings: abnormal denominator criteria, and instead include a denominator exclusion with radiologic finding: normal chest x-ray/CT scan. It is expected this will be an easier data construct to find. By moving the logic and revising the radiologic finding, the original value set with abnormal findings (2.16.840.1.113883.3.666.5.1105) is no longer used by the current version of the measure and a new value set was added to the logic (“Hospital Measures-Radiologic Finding-Normal SM-CT” SNOMED-CT Value Set (2.16.840.1.113883.3.666.5.2365). The April specifications you note use this new approach. Additional information can be found under CQM-226.

You can still find the original value set in VSAC because all versions must remain available.