Many ICD-9 codes are incorrectly labeled #2
Description
Here are a few rows of all_codes_mapping.parquet:
┌────────────┬──────────────────────────────┬──────────────────────────────────────────────────────────────┐
│ code │ med_code_node_mapping[:4] │ mt."desc"[:60] │
│ varchar │ int64[] │ varchar │
├────────────┼──────────────────────────────┼──────────────────────────────────────────────────────────────┤
│ icd9:00.25 │ [63160, 39991] │ Intravascular Imaging of Renal Vessels - This diagnostic pro │
│ icd9:00.28 │ [63160, 39991] │ Intravascular Imaging of Other Specified Vessels - This proc │
│ icd9:00.29 │ [63160, 39991] │ Paratyphoid Fever, Unspecified - A bacterial infection cause │
ICD-9 00.29 is actually "Intravascular imaging, unspecified vessel(s)". The present description is incorrectly taken from 002.9.
In ICD-9, codes like NN.xxx are procedures and NNN.xxx are diagnoses. As seen in some of the following rows, some procedures are currently mislabeled as diagnoses:
select mt.med_code, med_code_node_mapping[:4], mt.desc[:60] from mt where mt.code_system = 'icd9' and med_code[3] = '.';
┌──────────┬──────────────────────────────┬──────────────────────────────────────────────────────────────┐
│ med_code │ med_code_node_mapping[:4] │ mt."desc"[:60] │
│ varchar │ int64[] │ varchar │
├──────────┼──────────────────────────────┼──────────────────────────────────────────────────────────────┤
│ 00.0 │ [3529, 6739, 10620] │ Therapeutic Ultrasound - A non-invasive treatment using soun │
│ 00.01 │ [63160, 63139, 63444] │ Therapeutic Ultrasound of Vessels of Head and Neck - This pr │
│ 00.02 │ [63426, 64134] │ Therapeutic Ultrasound of Heart - A non-invasive procedure u │
│ 00.03 │ [63160] │ Therapeutic Ultrasound of Peripheral Vascular Vessels - This │
│ 00.1 │ [63907] │ Pharmaceuticals - This category encompasses the use of medic │
│ 00.10 │ [28454] │ Cholera due to Vibrio cholerae - An infection caused by the │
│ 00.16 │ [63907] │ Pressurized Pharmaceutical Treatment of Venous Bypass Graft │
│ 00.19 │ [63240, 63203, 63430] │ Cholera, Unspecified - An acute infectious disease caused by │
│ 00.2 │ [63160, 63240, 64421, 39991] │ Intravascular Imaging of Blood Vessels - A diagnostic proced │
│ 00.21 │ [63160, 39991] │ Paratyphoid Fever A - A bacterial infection caused by Salmon │
│ 00.22 │ [63160, 39991] │ Paratyphoid Fever B - A bacterial infection caused by Salmon │
│ 00.23 │ [63160, 39991] │ Paratyphoid Fever C - A bacterial infection caused by one of │
│ 00.24 │ [63160, 67703, 39991] │ Intravascular Imaging of Coronary Vessels - A diagnostic pro │
│ 00.25 │ [63160, 39991] │ Intravascular Imaging of Renal Vessels - This diagnostic pro │
│ 00.28 │ [63160, 39991] │ Intravascular Imaging of Other Specified Vessels - This proc │
│ 00.29 │ [63160, 39991] │ Paratyphoid Fever, Unspecified - A bacterial infection cause │
│ 00.35 │ [74956] │ Computer Assisted Surgery with Multiple Datasets - A surgica │
│ 00.40 │ [63160, 68719] │ Shigella dysenteriae Infection - A bacterial infection cause │
│ 00.41 │ [63160, 68719] │ Shigella flexneri - A bacterial infection caused by the bact │
│ 00.42 │ [63160, 68719] │ Shigella boydii Infection - A bacterial infection caused by │
│ · │ · │ · │
│ · │ · │ · │
│ · │ · │ · │
│ 99.55 │ [5015] │ Prophylactic Vaccination Against Other Diseases - This code │
│ 99.56 │ [63694, 32494] │ Administration of Tetanus Antitoxin - This procedure involve │
│ 99.57 │ [32729] │ Other Adverse Food Reactions, Not Elsewhere Classified - Thi │
│ 99.6 │ [63426, 363, 64134, 9391] │ Conversion of Cardiac Rhythm - A medical procedure aimed at │
│ 99.60 │ [63426, 28454, 64134, 64487] │ Cardiopulmonary Resuscitation (CPR) - A life-saving emergenc │
│ 99.62 │ [63426, 64134] │ Mechanical Complication of Nervous System Device, Implant, a │
│ 99.63 │ [63890] │ Closed Chest Cardiac Massage - A life-saving procedure perfo │
│ 99.64 │ [65970] │ Carotid Sinus Stimulation - A medical procedure involving th │
│ 99.74 │ [63240, 9391, 5938, 6739] │ Surgical Complications - Digestive System - This code refers │
│ 99.76 │ [69063] │ Extracorporeal Immunoadsorption - A medical procedure that f │
│ 99.81 │ [24100] │ Hypothermia - A medical condition characterized by an abnorm │
│ 99.82 │ [6739, 10620, 11747] │ Accidental Puncture or Laceration During a Procedure, NEC - │
│ 99.84 │ [62075] │ Foreign Body Accidentally Left During a Procedure - This cod │
│ 99.85 │ [35612] │ Hyperthermia for Treatment of Cancer - A therapeutic procedu │
│ 99.86 │ [52155, 41975] │ Persistent Postoperative Fistula - A complication characteri │
│ 99.88 │ [28454, 69063, 11369, 9391] │ Therapeutic Photopheresis - A medical procedure involving th │
│ 99.94 │ [4917, 12150] │ Anaphylactic Shock Due to Serum - A severe, life-threatening │
│ 99.95 │ [70362, 68782] │ Serum Reaction NEC (Not Elsewhere Classified) - This diagnos │
│ 99.96 │ [104164] │ ABO Incompatibility Reaction - An adverse immune response oc │
│ 99.98 │ [63266, 64354] │ Transfusion Reaction NEC (Not Elsewhere Classified) - This c │
├──────────┴──────────────────────────────┴──────────────────────────────────────────────────────────────┤
│ 3443 rows (40 shown) 3 columns │
└────────────────────────────────────────────────────────────────────────────────────────────────────────┘
It seems this problem goes beyond the decimal placement error. Consider the following entry:
│ 99.81 │ [24100] │ Hypothermia - A medical condition characterized by an abnorm │
ICD-9 99.81 is therapeutic hypothermia. The description pertains to the diagnosis of hypothermia. ICD-9 998.1 is unrelated ("Hemorrhage or hematoma complicating a procedure not elsewhere classified".
Is it possible to provide the code which generated med_code_descriptions.pkl, so it can be verified? Thank you for the interesting work.