feat!: switch encounter categories from 'type' to 'class'

Because this changed the name of a column in c_resource_count, the data package version and Python package major version are both getting bumped: data package version is now 3 Python package version is now 8 This change follows the same switch in the upstream qualifier spec.
smart-on-fhir · Feb 10, 2025 · 8d2015d · 8d2015d
1 parent 5012592
commit 8d2015d
Show file tree

Hide file tree

Showing 14 changed files with 45 additions and 43 deletions.
diff --git a/README.md b/README.md
@@ -4,7 +4,7 @@ A Cumulus-based implementation of the [qualifier metrics](https://github.com/syn
 
 ## Implemented Metrics
 
-The following qualifier metrics are implemented (per January 2025 qualifer definitions).
+The following qualifier metrics are implemented (per February 2025 qualifer definitions).
 
 - [c_attachment_count](https://github.com/sync-for-science/qualifier/blob/master/metrics.md#c_attachment_count)
 - [c_content_type_use](https://github.com/sync-for-science/qualifier/blob/master/metrics.md#c_content_type_use)

diff --git a/cumulus_library_data_metrics/__init__.py b/cumulus_library_data_metrics/__init__.py
@@ -1,3 +1,3 @@
 """Data Metrics study for Cumulus Library"""
 
-__version__ = "7.0.2"
+__version__ = "8.0.0"
diff --git a/cumulus_library_data_metrics/base.py b/cumulus_library_data_metrics/base.py
@@ -152,7 +152,7 @@ def render_sql(self, template: str, **kwargs) -> str:
             kwargs["dates"] = self.date_fields.get(src)
             kwargs["patient_field"] = resource_info.PATIENTS.get(src)
             kwargs["schema"] = self.schemas.get(src)
-            kwargs.update(resource_info.CATEGORIES.get(src, {}))
+            kwargs["cat_info"] = resource_info.CATEGORIES.get(src, {})
 
         # See how we should combine counts.
         kwargs["output_mode"] = self.output_mode

diff --git a/cumulus_library_data_metrics/c_resource_count/c_resource_count.jinja b/cumulus_library_data_metrics/c_resource_count/c_resource_count.jinja
@@ -1,17 +1,16 @@
 {% import 'utils.jinja' as utils %}
 
 CREATE TABLE {{ study_prefix }}__count_c_resource_count_{{ src|lower }}_{{ period|lower }} AS (
-{%- set orig_cat_field = cat_field %}
 WITH
 src_status AS {{ utils.extract_status(src) }},
-src_category AS {{ utils.extract_category_codes(src, cat_field, cat_systems) }},
+src_category AS {{ utils.extract_category_codes(src, cat_info) }},
 
 simplified AS (
     SELECT
         src.id,
 
-        {%- if cat_field %}
-        {{ utils.coalesce_missing('src_category.category') }} AS {{ cat_field }},
+        {%- if cat_info %}
+        {{ utils.coalesce_missing('src_category.category') }} AS {{ cat_info["cat_field"] }},
         {%- endif %}
 
         -- Parse string datetime into a single period string
@@ -30,8 +29,8 @@ simplified AS (
 )
 
 {% call utils.make_counts('simplified', output_mode) %}
-    {%- if cat_field %}
-    {{ cat_field }},
+    {%- if cat_info %}
+    {{ cat_info["cat_field"] }},
     {%- endif %}
 
     {%- if dates %}

diff --git a/cumulus_library_data_metrics/c_resources_per_pt/c_resources_per_pt.jinja b/cumulus_library_data_metrics/c_resources_per_pt/c_resources_per_pt.jinja
@@ -12,11 +12,7 @@ patient_refs AS (
 
 {% for resource, field_name in patient_fields.items() %}
 cat_slice_{{ resource|lower }}_codes AS
-{{ utils.extract_category_codes(
-    resource,
-    categories.get(resource, {}).get("cat_field"),
-    categories.get(resource, {}).get("cat_systems"),
-) }},
+{{ utils.extract_category_codes(resource, categories.get(resource, {})) }},
 {% endfor %}
 
 counts_by_pt AS (

diff --git a/cumulus_library_data_metrics/c_system_use/c_system_use.jinja b/cumulus_library_data_metrics/c_system_use/c_system_use.jinja
@@ -6,7 +6,7 @@ WITH
 src_status AS {{ utils.extract_status(src) }},
 
 {% if use_category %}
-src_categories AS {{ utils.extract_codes(src, cat_field, system=cat_systems, is_array=true) }},
+src_categories AS {{ utils.extract_category_codes(src, cat_info) }},
 {% endif %}
 
 src_systems_flat AS (
@@ -58,7 +58,7 @@ simplified AS (
         {% endif %}
 
         {% if use_category %}
-        {{ utils.array_to_string('src_categories.codes') }} AS category,
+        {{ utils.coalesce_missing('src_categories.category') }} AS category,
         {% endif %}
 
         {{ utils.array_to_string('src_systems.systems') }} AS systems,

diff --git a/cumulus_library_data_metrics/meta/version.jinja b/cumulus_library_data_metrics/meta/version.jinja
@@ -1,2 +1,2 @@
 CREATE TABLE {{ study_prefix }}__meta_version AS
-SELECT 2 AS data_package_version;
+SELECT 3 AS data_package_version;
diff --git a/cumulus_library_data_metrics/resource_info.py b/cumulus_library_data_metrics/resource_info.py
@@ -31,9 +31,10 @@
         "cat_systems": [systems.USCORE_DOCREF_CATEGORY],
     },
     "Encounter": {
-        "cat_field": "type",
-        # https://hl7.org/fhir/us/core/STU4/ValueSet-us-core-encounter-type.html
-        "cat_systems": [systems.CPT, systems.SNOMED],
+        "cat_field": "class",
+        # https://www.hl7.org/fhir/R4/v3/ActEncounterCode/vs.html
+        "cat_systems": [systems.ENCOUNTER_CLASS],
+        "is_coding": True,
     },
     "MedicationRequest": {
         "cat_field": "category",
@@ -50,7 +51,7 @@
 # Date fields in preference order.
 # This prefers "interaction with health system" dates, then administrative dates like "issued",
 # then best effort start dates like onsetDateTime.
-# See https://github.com/smart-on-fhir/cumulus-library-data-metrics/issues/16 for more.
+# See https://github.com/sync-for-science/qualifier/blob/master/metrics.md#by-date
 DATES = {
     "AllergyIntolerance": ["recordedDate", "onsetDateTime", "onsetPeriod.start", "onsetPeriod.end"],
     "Condition": [

diff --git a/cumulus_library_data_metrics/systems.py b/cumulus_library_data_metrics/systems.py
@@ -14,6 +14,7 @@
 # Resource-specific systems
 CONDITION_CATEGORY = "http://terminology.hl7.org/CodeSystem/condition-category"
 DIAGNOSTIC_SECTION = "http://terminology.hl7.org/CodeSystem/v2-0074"
+ENCOUNTER_CLASS = "http://terminology.hl7.org/CodeSystem/v3-ActCode"
 MEDREQ_CATEGORY = "http://terminology.hl7.org/CodeSystem/medicationrequest-category"
 OBSERVATION_CATEGORY = "http://terminology.hl7.org/CodeSystem/observation-category"
 USCORE_CONDITION_CATEGORY = "http://hl7.org/fhir/us/core/CodeSystem/condition-category"

diff --git a/cumulus_library_data_metrics/utils.jinja b/cumulus_library_data_metrics/utils.jinja
@@ -1,18 +1,18 @@
-{% macro _compare_systems(system) -%}
+{% macro _compare_systems(coding_field, system) -%}
     {#- Support both a list of systems or a single system #}
     {%- if system is string %}
-        c.coding.system = '{{ system }}'
+        {{ coding_field }}.system = '{{ system }}'
     {%- elif system %}
         (
             FALSE -- just here so rest of these can start with OR
         {%- for sys_info in system %}
             {#- sys_info can either by a string or a list of string then codes #}
             {%- if sys_info is string %}
-            OR c.coding.system = '{{ sys_info }}'
+            OR {{ coding_field }}.system = '{{ sys_info }}'
             {%- else %}
             OR (
-                c.coding.system = '{{ sys_info[0] }}'
-                AND c.coding.code in ('{{ sys_info[1]|join("', '") }}')
+                {{ coding_field }}.system = '{{ sys_info[0] }}'
+                AND {{ coding_field }}.code in ('{{ sys_info[1]|join("', '") }}')
             )
             {%- endif %}
         {%- endfor %}
@@ -48,7 +48,7 @@
         {%- endif %}
     WHERE
         c.coding.code IS NOT NULL
-        AND {{ _compare_systems(system) }}
+        AND {{ _compare_systems("c.coding", system) }}
 
     GROUP BY id
 )
@@ -72,7 +72,7 @@
         {%- endif %}
     WHERE
         c.coding.code IS NOT NULL
-        AND {{ _compare_systems(system) }}
+        AND {{ _compare_systems("c.coding", system) }}
 )
 {%- endmacro %}
 
@@ -117,9 +117,17 @@
 
 
 -- Extracts the category codes and returns a select with (id (string), category (string)).
-{% macro extract_category_codes(src, cat_field, cat_systems) -%}
+{% macro extract_category_codes(src, cat_info) -%}
 (
-{% if cat_systems %}
+{%- set cat_systems = cat_info.get("cat_systems") %}
+{%- set cat_field = cat_info.get("cat_field") %}
+{%- set is_coding = cat_info.get("is_coding", False) %}
+{% if is_coding %}
+    {# TODO: handle array & no-systems cases. But this works for now. #}
+    SELECT id, {{ cat_field }}.code AS category
+    FROM {{ src }}
+    WHERE {{ _compare_systems(cat_field, cat_systems) }}
+{% elif cat_systems %}
     WITH
     array_codes AS {{ extract_codes(src, cat_field, system=cat_systems, is_array=true) }}
     SELECT id, {{ array_to_string('codes') }} AS category

diff --git a/tests/data/c_resource_count/general/encounter/0.ndjson b/tests/data/c_resource_count/general/encounter/0.ndjson
@@ -1,6 +1,5 @@
-{"resourceType": "Encounter", "id": "with-date", "type": [{"coding": [{"code": "abcd", "system": "http://www.ama-assn.org/go/cpt"}]}], "period": {"start": "2023-10-04"}}
-{"resourceType": "Encounter", "id": "no-date", "type": [{"coding": [{"code": "1234", "system": "http://snomed.info/sct"}]}]}
-{"resourceType": "Encounter", "id": "multiple-codings", "type": [{"coding": [{"code": "abcd", "system": "http://www.ama-assn.org/go/cpt"}, {"code": "1234", "system": "http://snomed.info/sct"}]}], "period": {"start": "2023-10-04"}}
-{"resourceType": "Encounter", "id": "unrelated-category", "type": [{"coding": [{"code": "nope", "system": "nope"}]}]}
+{"resourceType": "Encounter", "id": "with-date", "class": {"code": "AMB", "system": "http://terminology.hl7.org/CodeSystem/v3-ActCode"}, "period": {"start": "2023-10-04"}}
+{"resourceType": "Encounter", "id": "no-date", "class": {"code": "EMER", "system": "http://terminology.hl7.org/CodeSystem/v3-ActCode"}}
+{"resourceType": "Encounter", "id": "unrelated-category", "class": {"code": "nope", "system": "nope"}}
 {"resourceType": "Encounter", "id": "no-category", "period": {"start": "2022-06-04"}}
 {"resourceType": "Encounter", "id": "nothing"}
diff --git a/tests/data/c_resource_count/general/expected_encounter_month.csv b/tests/data/c_resource_count/general/expected_encounter_month.csv
@@ -1,6 +1,5 @@
-cnt,type,month,status
+cnt,class,month,status
 2,cumulus__none,cumulus__none,cumulus__none
 1,cumulus__none,2022-06,cumulus__none
-1,abcd,2023-10,cumulus__none
-1,1234; abcd,2023-10,cumulus__none
-1,1234,cumulus__none,cumulus__none
+1,EMER,cumulus__none,cumulus__none
+1,AMB,2023-10,cumulus__none
diff --git a/tests/data/c_resource_count/general/expected_encounter_year.csv b/tests/data/c_resource_count/general/expected_encounter_year.csv
@@ -1,6 +1,5 @@
-cnt,type,year,status
+cnt,class,year,status
 2,cumulus__none,cumulus__none,cumulus__none
 1,cumulus__none,2022,cumulus__none
-1,abcd,2023,cumulus__none
-1,1234; abcd,2023,cumulus__none
-1,1234,cumulus__none,cumulus__none
+1,EMER,cumulus__none,cumulus__none
+1,AMB,2023,cumulus__none
diff --git a/tests/data/meta/general/expected_version.csv b/tests/data/meta/general/expected_version.csv
@@ -1,2 +1,2 @@
 data_package_version
-2
+3