There is a SNOMED code that is including ICD codes for patients that are in substance abuse remission. The purpose of the measure is to evaluate patients with a new SUD diagnosis. These patients do not have a new SUD and they often have codes of substance use disorder in remission, such as F15.11 Other stimulant abuse, in remission, which is also listed as a CMS-HCC. This is another example of where an ICD code is not in the value set but is being used under a SNOMED to pull patients in when that ICD code is used. This mapping is being used under IMO and we reached out to them for insight. Below is the their response. 

• The IMO term “other stimulant abuse, in remission” (Lexical ID: 1493348661) is mapped independently to:
  • SNOMED 441527004 |Harmful pattern of use of stimulant|, and
  • ICD-10-CM F15.11.
  • These mappings are made separately, not directly linked across systems.
• The SNOMED code 441527004 is included in an electronic clinical quality measure (eCQM) value set for substance use disorder.
• However, it appears the value set may intend to exclude cases that are explicitly in remission.
• By SNOMED hierarchy logic, though, “harmful pattern of use of X in remission” is a subtype (or child) of “harmful pattern of use of X” — meaning our mapping aligns properly with how SNOMED structures its relationships.
• The issue arises because the value set includes 441527004, which unintentionally pulls in remission concepts, even though the value set’s creator likely didn’t realize this.

Due to this measure pulling in patients to the denominator that are not appropriate it is not an accurate representation of an EC or TINs performance for the measure. Why are these codes being included under the SNOMED? Why are there hidden codes under SNOMEDs? Why aren't these codes just listed under the ICD code Value Sets if they are meant to pull patients into the measure? Can guidance be issued or a change made to exclude these patients from the measure as they do not have a new SUD diagnosis?