I10 vs. I10C

[hbuite...@gmail.com]

We noticed in the test specifications that states “NIST-SS-010 - DG1-3.3 SHALL be valued "I9CDX" or "I10C" or "SCT"”. However, the implementation Guide on Page 71 indicates that "I10" should be used. The Errata does not address a change in coding system, and furthermore, the latest V2.8.2 Chapter 2 Code Tables does not include an "I10C" value. However, given the I9xxx variances available, a special designation within I10 would not be unexpected, although never communicated and should not be put forth through the test specifications. Can you clarify the source of this new coding system? We also request that if indeed a targeted code needed to be established that this is communicated through an Errata by the implementation guide owner, and that it be formally recognized by ONC.

[sheryl...@nist.gov]

Hi Hans,

 

NIST is conferring with the syndromic surveillance subject matter experts who worked with NIST in determining that I10C should be used instead of I10.  An update will be provided as soon as available.

 

Thank you for your question,

 

Sheryl

[sheryl...@nist.gov]

Hi Hans,

 

The requirement in the SS Tool (“NIST-SS-010 - DG1-3.3 SHALL be valued "I9CDX" or "I10C" or "SCT") is supported by the ONC Edition 2015 Final Rule of October 2015. This conclusion is based on review of the relevant information in (1) the PHIN Guide Rel 2.0 of April 2015, (2) the Errata to the PHIN Guide Rel. 2.0 of August 2015, (3) PHVS_AdministrativeDiagnosis_CDC_ICD-10CM Value Set, and (4) HL70396.

 

The PHIN Guide R2.0 on Page 71 does indicate that "I10" should be used for PV2-3 (“PHVS_CauseOfDeath_ICD-10_CDC   The list provides ICD-10 codes and associated cause-of-death titles for the most detailed listing of causes of death. This list is maintained by CDC NCHS.").  The associated Errata (which is named in the Rule), however, corrects this statement and says on page 17 for PV2-3 under Revision/Resolution, "New Value Set: PHVS_AdministrativeDiagnosis_CDC_ICD-10CM", addressing a change in the coding system; and on page 21 it lists this value set for DG1-3.

 

The PHVS_AdministrativeDiagnosis_CDC_ICD-10CM (http://phinvads.cdc.gov/vads/ViewValueSet.action?oid=2.16.840.1.114222.4.11.7356 ) named in the PHIN Guide R2.0 and Errata indicates “I10C” as the entry in table HL70396.

 

Regarding V2.8.2 Chapter 2 Code Tables – this version is not being used; HL70396 is a dynamic HL7 table (divorced from any specific HL7 version) and includes “I10C” (see https://www.hl7.org/Special/committees/vocab/table_0396/index.cfm ). Additionally, Table 0396 included in PHINVADS for this IG also contains I10C, although a specific binding to it for PV2-3.3 is not clear. In either, case I10C is specified.

 

NIST hopes the above information clarifies the source of this new coding system (I10C).

 

Thank you for your question,

 

Sheryl

[hbuite...@gmail.com]

Sheryl:

We are concerned that "I10" fails and request that it passes as it is conformant with the published and recognized guide. Neither the implementation guide nor the errata indicates that "I10C" is required. In fact, it only states to use "I10" and the Errata does not make any reference to changing "I10" to "I10C. The errata only indicates a fix to the broken link for the value set, and the value set itself does not indicate that table is known as "I10C" by HL7.

While Table 0396 includes it, that does not automatically change an implementation guide unless that guide makes the changes AND is recognized. Neither happened. Consequently we must ask that NIST adjusts its tools to at least not fail "I10", while it could opt to accept "I10C" as well. Not doing so would set a precedent that implementation specifications can change by virtue of test data definition rather than formally published and recognized guides.

[sheryl...@nist.gov]

Hi Hans,

 

Regarding your questions:

 

1. We are concerned that "I10" fails and request that it passes as it is conformant with the published and recognized guide.  Neither the implementation guide nor the errata indicates that "I10C" is required.  In fact, it only states to use "I10" and the Errata does not make any reference to changing "I10" to "I10C. The errata only indicates a fix to the broken link for the value set, and the value set itself does not indicate that table is known as "I10C" by HL7.

1. If you follow the link of the value set (PHVS_ AdministrativeDiagnosis_ICD-10CM) that is referenced in the Errata document (which is named in the ONC 2015 Final Rule) you get to this information.

The value set indicates the code system that needs to be sent in the message instance. Although not explicitly mentioned in the Errata, it is clearly derivable. Additionally, most Guides would not explicitly list what needs to be sent as the code system, though the PHIN Guide does.

1. On page 75 in the PHIN Guide R2.0 (named in the ONC 2015 Final Rule) it says for DG1-3.1:

This component contains a code value (concept ID) from one of the following code systems:

PHVS_AdministrativeDiagnosis_CDC_ICD-9CM

Or

PHVS_AdministrativeDiagnosis_ICD-10CM

Or

PHVS_Disease_CDC

Conformance Statement SS-011: The implementation SHALL support all 3 value sets

1. “I10C” indicates the “PHVS_ AdministrativeDiagnosis_ICD-10CM” coding system, which is used for morbidity coding.

2. On page 21 in the Errata is further confirmation that the “PHVS_ AdministrativeDiagnosis_ICD-10CM” coding system (i.e., “I10C”) is to be used for DG1-3.

 

1. While Table 0396 includes [I10C], that does not automatically change an implementation guide unless that guide makes the changes AND is recognized.  Neither happened. 

1. Table 0396 is an HL7 Dynamic table and I10C was added explicitly to support the Errata. Additionally, the PHINVAD version of Table 0396 also includes I10C. We agree that the requirements for table 0396 aren’t all that clear in the IG, but the intent of including I10C for the purposes mentioned are clear.

 

1. Consequently we must ask that NIST adjusts its tools to at least not fail "I10", while it could opt to accept "I10C" as well.  Not doing so would set a precedent that implementation specifications can change by virtue of test data definition rather than formally published and recognized guides.

1. The information in the PHIN Guide 2.0 Errata overrides the information in the PHIN Guide 2.0. NIST conferred with the CDC Syndromic Surveillance subject matter experts. They confirmed that “I10C” is correct for DG1-3.3, that the example messages in the NIST tool are correct, and that the NIST tool is correctly validating the syndromic messages when it checks for ICD-10CM codes (for DG1-3 and for PV2-3 as well).

    

NIST’s # 1 rule is to test to the requirements and only to the requirements. NIST never intentionally interjects requirements. When issues such as this one arise, NIST seeks SME help to resolve them. Errata are created to document the resolution so we have something to refer to when the Test Cases are created for the NIST tools. When a requirement is not clear, NIST must make decisions based on what the intent of the IG author was. The PHIN Guide/Errata authors told NIST their intent regarding I10C, and NIST worked with them to document it correctly as stated above.

 

Thank you again for your questions,

 

Sheryl

[hbuite...@gmail.com]

Sheryl:

Thank you for the update. I would like to quibble with the indication that it is clearly derivable from the Errata. When navigating from the Errata to the value set, the initial page does not list "I10C" anywhere. When clicking on any of the Detail hyperlinks, it is still not listed. Not unless you hit the Download Value Set option and you actually download the value set spreadsheet do you realize that that spreadsheet has more columns and column I states indeed that "I10C" is to be used.

The spreadsheet includes more than what is displayed on the PHIN VADS page. The PHIN VADS web page is apparently only a summary, not the full set of data (i.e., columns). This is not indicated at all on the page, so downloading the value set is not necessarily an obvious step as it appears everything relevant is on the we page itself. Only downloading the spreadsheet will disclose that information.

This style of implementation guide documentation is not very helpful and clear, and would appreciate NIST also passing on to the implementation guide owners to document their guides more closely to how other V2 guides that have recently been published and included in certification editions are documented. That would avoid this type of confusion.

Thank you!

[sheryl...@nist.gov]

Hi Hans,

Thank you for taking time to provide your feedback on this subject. Please rest assured that NIST has shared this feedback with the syndromic surveillance implementation guide owners, and they are taking it very seriously.

Best regards,

Sheryl

NIST