Kathleen Ellis

1
Diagnoses and Procedures Mapped to SNOMED CT

• Kathleen Ellis
• Veterinary Medical DataBases

2
Vocabularies

• Some systems allow mapping concepts to pull-down
  lists controlled vocabularies
• Diagnoses
• Fracture of Femur
• Laminitis
• Procedures
• to charge screens (must be detailed to be able
  to be mapped)
• Spay
• Chemotherapy Doxorubicin
• Ultrasound

3
Benefits

• Time saver
• Health information staff may spend time on
  other endeavors
• Fewer staff need to be trained to use SNOMED CT
  to maintain the list
• May increase consistency or accuracy because it
  is pre-coded
• Data may be collected in real time
• Clinician selects the diagnosis or procedure from
  the list
• Simple queries may be done based on the pick list
• More in-depth queries provided by VMDB

4
Benefits of Collecting Procedural Data

• Collecting procedural data along with the
  diagnoses may enhance query results
• May present a more pared down list for the
  searcher to utilize
• Lymphoma where treated with chemotherapy
• Some queries are only performed on procedural
  data
• Drug studies
• Procedural data may be utilized for control
  studies
• Normal examinations

5
Mapping Only to Billing Information

• Does not allow for capturing diagnoses
• Cushings
• Urinary infections
• Some procedures may not be able to be captured by
  billing information alone
• Colic surgeries (too general)
• Procedures are not performed at every patient
  encounter
• We would like to collect both diagnosis and
  procedural data to get a better picture of the
  patient

6
How it works
7
VMDB SNOMED CT Master Diagnosis List

• Developed with Missouri / Dr. Al Hahn
• vmdb sample master listing SNOMED.xlsm
• UVIS and Florida internal ID
• Diagnoses text
• UVIS standard code
• SNOMED CT code
• SNOMED CT fully specified name

8
Distribution of Master List

• Diagnoses may not be changed, only retired or
  inactivated from the list
• Diagnoses may be reworded by the institution as
  long as it does not change the meaning
• Poisoning vs Toxicity
• But not substituting benign tumor for a
  malignant tumor
• A content expert should interpret the SNOMED
  mapping when a diagnoses wording is changed

9
Requests to Add Codes

• Request additions through VMDB
• Should be unambiguous
• So the clinician is absolutely sure of the
  meaning of the term(s) being chosen
• Pannus
• Cornea or third eyelid
• Bloat
• Rumen or non-ruminant
• Desmitis Suspensory
• Equine or non-equine

10
Requests to Add Codes

• Preference towards less detail
• Fracture femur, rather than Fracture patellar
  surface femur
• Synonyms
• Preference is not to add synonyms or duplicate
  concepts
• When concepts are revised, synonyms must also be
  revised
• Duplicate concepts within a patient visit may
  skew the data
• (When data is submitted, duplicates should be
  purged)

11
Maintenance of Mapped Lists

• Periodically reviewed for needed updates
• Retired codes
• Rules changes

12
Submitting Data to VMDB

• http//www.vmdb.org/Transmission_information/v
  mdb_transmission_file_guide.htm
• Weed out duplicate entries within a visit
• Mark a concept as a recheck
• Mark a concept as suspect (optional)
• Submit the concept code ID for Necropsy, Died and
  Euthanasia
• Keep the sort order for post-coordinated concepts

13
Submitting Data to VMDB

• Resend cases that are updated within the record
• Diagnoses/procedures added or deleted
• Patient signalment revised
• Visit information revised
• Resend cases where the concept IDs are revised
  within the mapping (such as with a code being
  retired or in error)
• Export records in chronologic order by admit or
  discharge date
• If possible, do error checking before submission,
• e.g., discharge date is not before admit date,
  DOB not after admit date, gender changes

14
Demo Concept IDs used Internally in HIS

• Diagnosis text added
• Diagnosis mapped to SNOMED
• Clinician chooses diagnosis
• Mapped codes stored in medical records screen for
  quality assurance

15
Using Mapped List Outside the HIS

• Some institutions build their own tool outside of
  their HIS to link their diagnoses to the mapped
  list
• Will work with institutions on a case by case
  basis

16
XML exampletest xml data.txt

• ltvisitsgt
• ltvisitgt
• ltinst_idgt13lt/inst_idgt
• lttrans_typegtClt/trans_typegt
• ltpat_nogt039875lt/pat_nogt
• ltaccession_numgt20051293lt/accession_numgt
• ltspecies/gt
• ltbreedgt44696006lt/breedgt
• ltdate_of_birthgt19950920lt/date_of_birthgt
• ltgendergtClt/gendergt
• ltpostal_codegt60416lt/postal_codegt
• ltweight_valuegt30.9lt/weight_valuegt
• ltweight_unitgtlbslt/weight_unitgt
• ltadmit_dtgt20051123101953lt/admit_dtgt
• ltcliniciangt59lt/cliniciangt
• ltdischarge_dtgt20051123lt/discharge_dtgt
• ltdischarge_dispositiongt 0lt/discharge_disposition
  gt
• ltsnomed_groupgt
• ltconcept_id suspect"0" recheck"0"
  order"1"gt68566005lt/concept_idgt

17
Acknowledgements

• Hahn, Allen (University of Missouri)
• Laskowski, Natalie (University of Illinois)
• Siegel, Arthur (University of Illinois/VMDB)