7

Are there any data sets of medical coding examples available online?

Eg. Doctor's notes and resulting (correct/accepted) procedure and diagnosis codes for individual encounters or samples of incorrectly coded ICD10 and CPT code sequences/bunches and they are associated with corrected sequences/bunches (by saying, ing medical coders or reviewers).

Or something of the form, for example, of what a medical billing coder might see on a charge review screen. Something like says a dataset where each sample contains

[<patient demographics>, <Diagnoses>, <Charges (ie. ICD10 codes and the associated diagnoses)>, etc]

I'm aware of the MIMIC-III data set, but need ICD10 (the latest medical code standard, implemented on 2015-10-01) while the latest version of MIMIC uses ICD9. Also aware of this very similar question, but was hoping that users may have suggestions in this case since the data I need is slightly more specific. Would like to look at differences in what doctors record as diagnoses and procedure codes and what the actual correct/acceptable code combinations should be as edited by a medical coder before the charge was sent to an insurer.

Improve this question
1
  • This information exists primarily in the general practioneers / specialists computer and most likely not anywhere else. E.g. in Europe due to GDPR this is usually not shared with sick funds / insurances. All handling /statistics / anomaly dection is then already based on coded values, AFAIK. You may find patient registries that provide a deeper level of information (than codes) or some other commercial offering of RWD might have that information in other jurisdictions like the US. Hope some others can help you with information on that.
    – Grimaldi
    Apr 12, 2023 at 13:53

1 Answer 1

Reset to default
0

I believe this file represents the data you seek: https://raw.githubusercontent.com/reymor/meical/master/medical_icd10/data/diseases.xml

Improve this answer
1
  • Not quite. Was looking more for datapoints that give a picture of a patient-doctor encounter. Something of the form, for example, of what a medical billing coder might see. Something like say a dataset where each sample is contains [<patient demographics>, <Diagnoses>, <Charges (ie. ICD10 codes and the associated diagnoses)>, etc]. This questions' been quiet for some time, but will update it to include this info.
    – lampShadesDrifter
    Aug 15, 2018 at 22:33

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.