Merge pull request #50 from mmcdermott/33_track_code_descriptions

Very preliminary code to fix #33
mmcdermott · Jul 23, 2024 · 3285875 · 3285875
2 parents a7cdaaf + 1aaf037
commit 3285875
Show file tree

Hide file tree

Showing 10 changed files with 1,387 additions and 57 deletions.
diff --git a/MIMIC-IV_Example/configs/event_configs.yaml b/MIMIC-IV_Example/configs/event_configs.yaml
@@ -45,6 +45,10 @@ hosp/diagnoses_icd:
     hadm_id: hadm_id
     timestamp: col(hadm_discharge_time)
     timestamp_format: "%Y-%m-%d %H:%M:%S"
+    _metadata:
+      d_icd_diagnoses:
+        description: "long_title"
+        parent_code: "ICD{icd_version}CM/{icd_code}" # Single strings are templates of columns.
 
 hosp/drgcodes:
   drg:
@@ -79,6 +83,11 @@ hosp/hcpcsevents:
     hadm_id: hadm_id
     timestamp: col(chartdate)
     timestamp_format: "%Y-%m-%d"
+    _metadata:
+      # These are not all CPT codes, unfortunately
+      d_hcpcs:
+        description: "long_description"
+        possibly_cpt_code: "code"
 
 hosp/labevents:
   lab:
@@ -92,6 +101,11 @@ hosp/labevents:
     numerical_value: valuenum
     text_value: value
     priority: priority
+    _metadata:
+      d_labitems_to_loinc:
+        description: ["omop_concept_name", "label"] # List of strings are columns to be collated
+        itemid: "itemid (omop_source_code)"
+        parent_code: "{omop_vocabulary_id}/{omop_concept_code}"
 
 hosp/omr:
   omr:
@@ -152,6 +166,12 @@ hosp/procedures_icd:
     hadm_id: hadm_id
     timestamp: col(chartdate)
     timestamp_format: "%Y-%m-%d"
+    _metadata:
+      d_icd_procedures:
+        description: "long_title"
+        parent_code: # List of objects are string labels mapping to filters to be evaluated.
+          - "ICD{icd_version}Proc/{icd_code}": { icd_version: 9 }
+          - "ICD{icd_version}PCS/{icd_code}": { icd_version: 10 }
 
 hosp/transfers:
   transfer:
@@ -193,6 +213,16 @@ icu/chartevents:
     text_value: value
     hadm_id: hadm_id
     icustay_id: stay_id
+    _metadata:
+      meas_chartevents_main:
+        description: ["omop_concept_name", "label"] # List of strings are columns to be collated
+        itemid: "itemid (omop_source_code)"
+        parent_code: "{omop_vocabulary_id}/{omop_concept_code}"
+      # TODO: I don't know if this is necessary...
+      d_labitems_to_loinc:
+        description: ["omop_concept_name", "label"] # List of strings are columns to be collated
+        itemid: "itemid (omop_source_code)"
+        parent_code: "{omop_vocabulary_id}/{omop_concept_code}"
 
 icu/procedureevents:
   start:
@@ -204,6 +234,15 @@ icu/procedureevents:
     timestamp_format: "%Y-%m-%d %H:%M:%S"
     hadm_id: hadm_id
     icustay_id: stay_id
+    _metadata:
+      proc_datetimeevents:
+        description: ["omop_concept_name", "label"] # List of strings are columns to be collated
+        itemid: "itemid (omop_source_code)"
+        parent_code: "{omop_vocabulary_id}/{omop_concept_code}"
+      proc_itemid:
+        description: ["omop_concept_name", "label"] # List of strings are columns to be collated
+        itemid: "itemid (omop_source_code)"
+        parent_code: "{omop_vocabulary_id}/{omop_concept_code}"
   end:
     code:
       - PROCEDURE
@@ -213,3 +252,77 @@ icu/procedureevents:
     timestamp_format: "%Y-%m-%d %H:%M:%S"
     hadm_id: hadm_id
     icustay_id: stay_id
+    _metadata:
+      proc_datetimeevents:
+        description: ["omop_concept_name", "label"] # List of strings are columns to be collated
+        itemid: "itemid (omop_source_code)"
+        parent_code: "{omop_vocabulary_id}/{omop_concept_code}"
+      proc_itemid:
+        description: ["omop_concept_name", "label"] # List of strings are columns to be collated
+        itemid: "itemid (omop_source_code)"
+        parent_code: "{omop_vocabulary_id}/{omop_concept_code}"
+
+icu/inputevents:
+  input_start:
+    code:
+      - INFUSION_START
+      - col(ordercategorydescription)
+      - col(itemid)
+      - col(rateuom)
+    timestamp: col(starttime)
+    timestamp_format: "%Y-%m-%d %H:%M:%S"
+    hadm_id: hadm_id
+    icustay_id: stay_id
+    order_id: orderid
+    link_order_id: linkorderid
+    numerical_value: rate
+    _metadata:
+      inputevents_to_rxnorm:
+        description: ["omop_concept_name", "label"] # List of strings are columns to be collated
+        itemid: "itemid (omop_source_code)"
+        parent_code: "{omop_vocabulary_id}/{omop_concept_code}"
+        rateuom: null # A null column means this column is needed in pulling from the metadata.
+  input_end:
+    code:
+      - INFUSION_END
+      - col(ordercategorydescription)
+      - col(itemid)
+      - col(amountuom)
+      - col(statusdescription)
+    timestamp: col(endtime)
+    timestamp_format: "%Y-%m-%d %H:%M:%S"
+    hadm_id: hadm_id
+    icustay_id: stay_id
+    order_id: orderid
+    link_order_id: linkorderid
+    numerical_value: amount
+    _metadata:
+      inputevents_to_rxnorm:
+        description: ["omop_concept_name", "label"] # List of strings are columns to be collated
+        itemid: "itemid (omop_source_code)"
+        parent_code: "{omop_vocabulary_id}/{omop_concept_code}"
+  patient_weight:
+    code:
+      - PATIENT_WEIGHT_AT_INFUSION
+      - KG
+    timestamp: col(starttime)
+    timestamp_format: "%Y-%m-%d %H:%M:%S"
+    numerical_value: patientweight
+
+icu/outputevents:
+  output:
+    code:
+      - PATIENT_FLUID_OUTPUT
+      - col(itemid)
+      - col(valueuom)
+    timestamp: col(charttime)
+    timestamp_format: "%Y-%m-%d %H:%M:%S"
+    hadm_id: hadm_id
+    icustay_id: stay_id
+    numerical_value: value
+    _metadata:
+      outputevents_to_rxnorm:
+        description: ["omop_concept_name", "label"] # List of strings are columns to be collated
+        itemid: "itemid (omop_source_code)"
+        parent_code: "{omop_vocabulary_id}/{omop_concept_code}"
+        valueuom: unitname
diff --git a/pyproject.toml b/pyproject.toml
@@ -33,6 +33,7 @@ MEDS_extract-shard_events = "MEDS_polars_functions.extract.shard_events:main"
 MEDS_extract-convert_to_sharded_events = "MEDS_polars_functions.extract.convert_to_sharded_events:main"
 MEDS_extract-merge_to_MEDS_cohort = "MEDS_polars_functions.extract.merge_to_MEDS_cohort:main"
 MEDS_transform-aggregate_code_metadata = "MEDS_polars_functions.aggregate_code_metadata:main"
+MEDS_extract-extract_code_metadata = "MEDS_polars_functions.extract.extract_code_metadata:main"
 
 [project.urls]
 Homepage = "https://github.com/mmcdermott/MEDS_polars_functions"

diff --git a/src/MEDS_polars_functions/configs/extract.yaml b/src/MEDS_polars_functions/configs/extract.yaml
@@ -35,6 +35,7 @@ stages:
   - convert_to_sharded_events
   - merge_to_MEDS_cohort
   - aggregate_code_metadata
+  - extract_code_metadata
 
 stage_configs:
   aggregate_code_metadata: