Controlling Dx Type for ICD10 codes

Data Integrity Checks
Summary:	Some dxs can't be Dx Types Primary Admit Diagnosis, Comorbid Diagnosis, Admit Diagnosis or Acquired Diagnosis
Related:	ICD10 Diagnoses, Primary Admit Diagnosis, Comorbid Diagnosis, Admit Diagnosis, Acquired Diagnosis, Dx Type
Firmness:	hard check
Timing:	complete
App:	CCMDB.accdb
Coding:
Uses L Problem table:	not relevant for this app
Status:	needs review
Implementation Date:
Backlogged:	true

• Cargo

• SMW

• Categories: 

• form:

Proposed Solution

Encode which codes can not be of a particular type; this will likely be fewer checks, and it will also likely be an easier and more deterministic answer without ifs-thens-buts; it will be a slightly less tight check but with fewer false positives.

• Add three columns to the s_ICD10 table:
  • comorbid_not_allowed (Comorbid Diagnosis)
  • admit_not_allowed (Admit Diagnosis
  • acquired_not_allowed (Acquired Diagnosis)
  • primary_not_allowed (Primary Admit Diagnosis)

_after Built query "z_ComoAdmAcqu_Primary_List", exported to excel, sent to Barret; Allan said this is low priority, so putting a review data of 3 months from now added this back to the task list, been almost 6 months

added: 2022-02-17 action: 2022-08-10

• Cargo

• Categories

• put cross checks into CCMDB.accdb to prevent bad entries

why do this change outside the wiki?

The change will require editing every single record in the ICD10 Diagnosis List, so doing it on the wiki would be time-prohibitive. If we do it externally the new fields would be imported into the wiki, which would then again be the master repository for this.

• We will use a tool like AutoWikiBrowser or similar to get the changes integrated into pages.

Implementation

It was decided that this is worth doing, and that the above is a reasonable way to do it. Allan is working on generating the list.

Wiki cleanups when done

_after Make sure that these examples that were mentioned over time are adressed: Medical noncompliance has some special rules around this Cardiac arrest Severe sepsis, VAP, recurrent Pneumonia should never be a Comorbid Diagnosis Past history, removal of breast (mastectomy) should never be an Acquired Diagnosis / Complication or Admit Diagnosis

added: 2022-02-17 action:

• Cargo

• Categories

If we do this, should we do more?

• Discussed but rejected possibility to restrict impossible dx / sex field combinations.
• this would be the right time to also fix/implement the following:
  • is_pathogen: see Category:Infection requiring pathogen
  • pathogen requirement (see Category:Infection requiring pathogen, Category:Infection requiring pathogen, Category:Potential infection)
  • min nr of combined codes field