aimedicalcoding 0.3.1__tar.gz → 0.3.4__tar.gz

{aimedicalcoding-0.3.1/aimedicalcoding.egg-info → aimedicalcoding-0.3.4}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: aimedicalcoding
-Version: 0.3.1
+Version: 0.3.4
 Summary: Detect missing/broken clinical codes (LOINC / SNOMED CT / RxNorm / ICD-10-CM / CVX) in CCDA and FHIR documents, with recovery evidence
 Author: Aakash Kag, Vinay Shankar Miryala, Siva Karthik, Mohammed Rayaan
 License: MIT
@@ -194,7 +194,7 @@ to fix it:
 | `ICD_ONLY` | ICD present, SNOMED missing | official ICD ↔ SNOMED map |
 | `SNOMED_ONLY` | SNOMED present, ICD-10-CM missing | official SNOMED → ICD-10-CM map |
 | `NDC_ONLY` | NDC present, RxNorm/CVX missing | NDC ↔ RxNorm / CDC NDC ↔ CVX table |
-| `RXNORM_ONLY` | RxNorm vaccine product, CVX missing | CVX ↔ RxNorm map |
+| `RXNORM_ONLY` | RxNorm present, target missing (vaccine → CVX; drug allergen → SNOMED) | CVX ↔ RxNorm map / RxNorm ↔ SNOMED via UMLS |
 | `MULTUM_ONLY` / `FDB_ONLY` / `MEDISPAN_ONLY` | proprietary drug vocabulary, RxNorm missing | vendor ↔ RxNorm via UMLS |
 
 ---

{aimedicalcoding-0.3.1 → aimedicalcoding-0.3.4}/README.md

@@ -56,11 +56,20 @@ Run tests:
 ```bash
 pytest
 ```
-Build distributables (version comes from the git tag):
-```bash
-python -m build
+
+### Publish to PyPI
+
+See **`PUBLISHING.md`** for the full release guide (version bump in `setup.py`,
+clean build, `twine check`, upload, verify, troubleshooting). Short version:
+
+```powershell
+# bump version in setup.py first — PyPI rejects re-uploads of the same version
+Remove-Item -Recurse -Force dist; python -m build
+python -m twine check dist/*
+python -m twine upload dist/*      # credentials from ~/.pypirc
 ```
-Publish privately to AWS CodeArtifact (the repo is on CodeCommit):
+
+Alternative: private publish to AWS CodeArtifact (the repo is on CodeCommit):
 ```bash
 aws codeartifact login --tool twine --domain <domain> --repository <repo>
 twine upload --repository codeartifact dist/*
@@ -187,7 +196,7 @@ a FHIR Bundle, and which gap patterns each can produce:
 | Target | CCDA source (slot) | FHIR source (slot) | Gap patterns it can emit |
 |---|---|---|---|
 | **LOINC** | results / vitals / surveys — observation `code` (question) | `Observation.code` (category `laboratory`/`vital-signs`/`social-history`/`survey`) | `LOCAL_CODE_ONLY` · `NULLFLAVOR` · `DISPLAY_ONLY` · `NARRATIVE_ONLY` · `BAD_CODE` |
-| **SNOMED CT** | problems / family hx / social hx / reactions — `value`; procedures / allergen substance — `code` | `Condition.code` / `FamilyMemberHistory.condition[].code` / `Observation.valueCodeableConcept` / `reaction[].manifestation[]` — value slot; `Procedure.code` / `AllergyIntolerance.code` — code slot | `LOCAL_CODE_ONLY` · **`ICD_ONLY`** · `NULLFLAVOR` · `DISPLAY_ONLY` · `NARRATIVE_ONLY` · `BAD_CODE` |
+| **SNOMED CT** | problems / family hx / social hx / reactions — `value`; procedures / allergen substance — `code` | `Condition.code` / `FamilyMemberHistory.condition[].code` / `Observation.valueCodeableConcept` / `reaction[].manifestation[]` — value slot; `Procedure.code` / `AllergyIntolerance.code` — code slot | `LOCAL_CODE_ONLY` · **`ICD_ONLY`** · **`RXNORM_ONLY`** · `NULLFLAVOR` · `DISPLAY_ONLY` · `NARRATIVE_ONLY` · `BAD_CODE` |
 | **RxNorm** | medications — `manufacturedMaterial/code` | `Medication.code` via `medicationReference` (or inline `medicationCodeableConcept`) | **`NDC_ONLY`** · **`MULTUM_ONLY`** · **`FDB_ONLY`** · **`MEDISPAN_ONLY`** · `LOCAL_CODE_ONLY` · `NULLFLAVOR` · `DISPLAY_ONLY` · `NARRATIVE_ONLY` |
 | **ICD-10-CM** | problems / encounters — diagnosis `value` | `Condition.code` (value slot) | **`SNOMED_ONLY`** · `LOCAL_CODE_ONLY` · `NULLFLAVOR` · `DISPLAY_ONLY` · `NARRATIVE_ONLY` · `BAD_CODE` |
 | **CVX** | immunizations — `manufacturedMaterial/code` | `Immunization.vaccineCode` | **`RXNORM_ONLY`** · **`NDC_ONLY`** · `LOCAL_CODE_ONLY` · `NULLFLAVOR` · `DISPLAY_ONLY` · `NARRATIVE_ONLY` · `BAD_CODE` |
@@ -217,7 +226,7 @@ deterministically to the target. Full schema + per-pattern JSON in
 | `ICD_ONLY` | ICD-9/10 present, no SNOMED | ICD↔SNOMED map (NLM) |
 | `SNOMED_ONLY` | SNOMED present, no ICD-10-CM | SNOMED→ICD-10-CM map (NLM) |
 | `NDC_ONLY` | NDC present, no RxNorm/CVX | NDC↔RxNorm / CDC NDC↔CVX table |
-| `RXNORM_ONLY` | RxNorm vaccine product, no CVX | CVX↔RxNorm map |
+| `RXNORM_ONLY` | RxNorm code present, target missing (vaccine → CVX; drug allergen → SNOMED) | CVX↔RxNorm map / RxNorm↔SNOMED via UMLS |
 | `MULTUM_ONLY` / `FDB_ONLY` / `MEDISPAN_ONLY` | proprietary drug vocab, no RxNorm | vendor↔RxNorm via UMLS CUI |
 
 Skipped (never emitted): a valid code for the target (incl. via `<translation>`),

{aimedicalcoding-0.3.1 → aimedicalcoding-0.3.4}/README_PYPI.md

@@ -157,7 +157,7 @@ to fix it:
 | `ICD_ONLY` | ICD present, SNOMED missing | official ICD ↔ SNOMED map |
 | `SNOMED_ONLY` | SNOMED present, ICD-10-CM missing | official SNOMED → ICD-10-CM map |
 | `NDC_ONLY` | NDC present, RxNorm/CVX missing | NDC ↔ RxNorm / CDC NDC ↔ CVX table |
-| `RXNORM_ONLY` | RxNorm vaccine product, CVX missing | CVX ↔ RxNorm map |
+| `RXNORM_ONLY` | RxNorm present, target missing (vaccine → CVX; drug allergen → SNOMED) | CVX ↔ RxNorm map / RxNorm ↔ SNOMED via UMLS |
 | `MULTUM_ONLY` / `FDB_ONLY` / `MEDISPAN_ONLY` | proprietary drug vocabulary, RxNorm missing | vendor ↔ RxNorm via UMLS |
 
 ---

{aimedicalcoding-0.3.1 → aimedicalcoding-0.3.4}/aimedicalcoding/serialize.py

@@ -121,13 +121,18 @@ def _dedupe(items) -> list:
     return out
 
 
-# SNOMED search-term enrichment: sections whose gaps also get the CODE-slot
-# labels appended (Problems 11450-4, Allergies 48765-2, Social History 29762-2),
-# matched by section LOINC or by the contexts those sections route to
-# (FHIR resources often carry no meta.tag/section code).
-_SNOMED_LABEL_SECTIONS = {"11450-4", "48765-2", "29762-2"}
-_SNOMED_LABEL_CONTEXTS = {"problem", "allergy_substance", "allergy_reaction",
-                          "social_history"}
+# SNOMED search-term enrichment (Problems 11450-4, Allergies 48765-2, Social
+# History 29762-2), matched by section LOINC or by the contexts those sections
+# route to (FHIR resources often carry no meta.tag/section code). Which slot's
+# labels get appended differs by section:
+#   - Social History  -> CODE tag labels (the question, e.g. "Gender identity" —
+#     the value label is often a placeholder like "Not on file")
+#   - Problems / Allergies -> VALUE tag labels (the concept; the code tag is a
+#     generic marker like ASSERTION / "Problem" — appending it is noise)
+_SNOMED_CODE_LABEL_SECTIONS = {"29762-2"}
+_SNOMED_CODE_LABEL_CONTEXTS = {"social_history"}
+_SNOMED_VALUE_LABEL_SECTIONS = {"11450-4", "48765-2"}
+_SNOMED_VALUE_LABEL_CONTEXTS = {"problem", "allergy_substance", "allergy_reaction"}
 
 
 def _recovery(record: NormalizedRecord, slot: str, system: str) -> dict:
@@ -141,13 +146,18 @@ def _recovery(record: NormalizedRecord, slot: str, system: str) -> dict:
         terms = [stmt.display_name, stmt.original_text]
     # add any normalized term the handler produced
     terms += [ev.get("analyte_normalized"), ev.get("term_normalized"), ev.get("ingredient")]
-    # SNOMED only, Problems/Allergies/Social History only: append the CODE
-    # tag's displayName/originalText (never the value's) as trailing context —
-    # e.g. the question label when the value label is a placeholder.
-    if system == System.SNOMEDCT.value and (
-            stmt.section_code in _SNOMED_LABEL_SECTIONS
-            or stmt.context in _SNOMED_LABEL_CONTEXTS):
-        terms += [stmt.display_name, stmt.original_text]
+    # SNOMED only, Problems/Allergies/Social History only: append trailing
+    # label context from the section-appropriate slot (see constants above).
+    if system == System.SNOMEDCT.value:
+        if (stmt.section_code in _SNOMED_CODE_LABEL_SECTIONS
+                or stmt.context in _SNOMED_CODE_LABEL_CONTEXTS):
+            terms += [stmt.display_name, stmt.original_text]
+        elif (stmt.section_code in _SNOMED_VALUE_LABEL_SECTIONS
+                or stmt.context in _SNOMED_VALUE_LABEL_CONTEXTS):
+            if stmt.value:
+                terms += [stmt.value.coded_value.display
+                          if stmt.value.coded_value else None,
+                          stmt.value.original_text]
 
     rec = {
         "search_terms": _dedupe(terms),

{aimedicalcoding-0.3.1 → aimedicalcoding-0.3.4}/aimedicalcoding/systems/snomed/patterns.py

@@ -31,6 +31,7 @@ ICD_SYSTEMS = {System.ICD10CM, System.ICD9CM, System.ICD10PCS}
 class SnomedGapPattern(str, Enum):
     LOCAL_CODE_ONLY = "LOCAL_CODE_ONLY"   # a local/other clinical code (incl. CPT), no SNOMED
     ICD_ONLY = "ICD_ONLY"                 # ICD-9/10 only -> ICD↔SNOMED crosswalk
+    RXNORM_ONLY = "RXNORM_ONLY"           # RxNorm-coded allergen -> RxNorm↔SNOMED (UMLS)
     NULLFLAVOR = "NULLFLAVOR"
     DISPLAY_ONLY = "DISPLAY_ONLY"
     NARRATIVE_ONLY = "NARRATIVE_ONLY"
@@ -107,13 +108,17 @@ def classify(stmt: ClinicalStatement) -> Optional[SnomedGapPattern]:
 
     if cc and cc.code and cc.system in ICD_SYSTEMS:
         return SnomedGapPattern.ICD_ONLY
+    if cc and cc.code and cc.system == System.RXNORM:
+        return SnomedGapPattern.RXNORM_ONLY   # standard vocab, not "local"
     if cc and cc.code and cc.system not in STRUCTURAL:
         return SnomedGapPattern.LOCAL_CODE_ONLY   # CPT / local / other
 
-    # no usable concept code
+    # no usable concept code. The label must come from the CONCEPT's slot
+    # (value display/originalText for value-slot contexts) — raw_text alone
+    # misses e.g. reaction observations whose code tag is a bare ASSERTION.
     if _has_nullflavor(stmt):
         return SnomedGapPattern.NULLFLAVOR
-    if stmt.raw_text:
+    if concept_term(stmt):
         if "/text" in (stmt.source_ref or ""):
             return SnomedGapPattern.NARRATIVE_ONLY
         return SnomedGapPattern.DISPLAY_ONLY
@@ -152,6 +157,12 @@ def parse_gap(stmt: ClinicalStatement, pattern: SnomedGapPattern) -> Gap:
         gap.evidence["icd_system"] = (cc.system.value if cc.system else None)
         gap.note = ("Use the official ICD↔SNOMED map (NLM/UMLS) — deterministic; "
                     "LLM only to choose among multiple map targets.")
+    elif pattern == SnomedGapPattern.RXNORM_ONLY and cc:
+        gap.evidence["rxnorm_code"] = cc.code
+        gap.evidence["rxnorm_system"] = cc.raw_system
+        gap.note = ("RxNorm-coded allergen (drug allergy) — crosswalk RxNorm → "
+                    "SNOMED substance/product via UMLS CUI; LLM only if the "
+                    "ingredient maps ambiguously.")
     elif pattern == SnomedGapPattern.BAD_CODE and cc:
         gap.evidence["malformed_code"] = cc.code
         gap.note = "SCTID-shaped but Verhoeff check failed; resolve via text + hierarchy."

{aimedicalcoding-0.3.1 → aimedicalcoding-0.3.4/aimedicalcoding.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: aimedicalcoding
-Version: 0.3.1
+Version: 0.3.4
 Summary: Detect missing/broken clinical codes (LOINC / SNOMED CT / RxNorm / ICD-10-CM / CVX) in CCDA and FHIR documents, with recovery evidence
 Author: Aakash Kag, Vinay Shankar Miryala, Siva Karthik, Mohammed Rayaan
 License: MIT
@@ -194,7 +194,7 @@ to fix it:
 | `ICD_ONLY` | ICD present, SNOMED missing | official ICD ↔ SNOMED map |
 | `SNOMED_ONLY` | SNOMED present, ICD-10-CM missing | official SNOMED → ICD-10-CM map |
 | `NDC_ONLY` | NDC present, RxNorm/CVX missing | NDC ↔ RxNorm / CDC NDC ↔ CVX table |
-| `RXNORM_ONLY` | RxNorm vaccine product, CVX missing | CVX ↔ RxNorm map |
+| `RXNORM_ONLY` | RxNorm present, target missing (vaccine → CVX; drug allergen → SNOMED) | CVX ↔ RxNorm map / RxNorm ↔ SNOMED via UMLS |
 | `MULTUM_ONLY` / `FDB_ONLY` / `MEDISPAN_ONLY` | proprietary drug vocabulary, RxNorm missing | vendor ↔ RxNorm via UMLS |
 
 ---

{aimedicalcoding-0.3.1 → aimedicalcoding-0.3.4}/docs/GAP_SCHEMA.md

@@ -40,7 +40,7 @@ Each record separates four concerns:
 | `id` | sequence number |
 | `target.system` | `LOINC` \| `SNOMED-CT` — which terminology this gap recovers |
 | `target.slot` | `code` (question/analyte) \| `value` (answer) — which slot is being filled |
-| `pattern` | `LOCAL_CODE_ONLY` \| `ICD_ONLY` \| `NULLFLAVOR` \| `DISPLAY_ONLY` \| `BAD_CODE` \| `NARRATIVE_ONLY` |
+| `pattern` | `LOCAL_CODE_ONLY` \| `ICD_ONLY` \| `RXNORM_ONLY` \| `NULLFLAVOR` \| `DISPLAY_ONLY` \| `BAD_CODE` \| `NARRATIVE_ONLY` |
 | `payload_type` | `codeable` (code is the concept, value is a measurement) \| `qa` (code is a question, value is a coded answer) |
 | `section` | `{ code, name, system }` — the source section |
 | **`observation`** | the faithful facts of BOTH slots (see below) |
@@ -89,10 +89,12 @@ Each record separates four concerns:
 - `target.slot = "code"` (LOINC) → search `observation.code.display_text` / `.original_text`.
 - `target.slot = "value"` (SNOMED value-slot) → search `observation.value.display_text` / `.original_text`.
 `recovery.search_terms` already gives you the right ones, deduped.
-For **SNOMED-CT** targets in **Problems (11450-4), Allergies (48765-2) and
-Social History (29762-2)**, the CODE tag's `display_text`/`original_text`
-are additionally appended at the end (never the value's) — search context
-for when the value label is a placeholder like "Not on file".
+For **SNOMED-CT** targets, extra label context is appended at the end,
+section-specific: **Social History (29762-2)** appends the CODE tag's
+`display_text`/`original_text` (the question — useful when the value label
+is a placeholder like "Not on file"); **Problems (11450-4)** and
+**Allergies (48765-2)** append the VALUE tag's labels (the concept — the
+code tag there is a generic marker like ASSERTION/"Problem").
 
 ---
 
@@ -102,6 +104,7 @@ for when the value label is a placeholder like "Not on file".
 |---|---|---|---|
 | `LOCAL_CODE_ONLY` | both | a local/in-house code (or CPT for SNOMED proc), no target code | crosswalk first, else text+filter |
 | `ICD_ONLY` | SNOMED | value coded in ICD-9/10 only | **ICD↔SNOMED map** (deterministic) |
+| `RXNORM_ONLY` | SNOMED | allergen coded in RxNorm only (drug allergy) | **RxNorm↔SNOMED via UMLS CUI** |
 | `NULLFLAVOR` | both | code blank; label recovered from narrative | text+filter on search_terms |
 | `DISPLAY_ONLY` | both | a display/text but no usable code | text+filter |
 | `BAD_CODE` | both | shaped right, check digit fails (Luhn/Verhoeff) | LOINC: `checkdigit_correction`; SNOMED: re-resolve |

{aimedicalcoding-0.3.1 → aimedicalcoding-0.3.4}/setup.py

@@ -2,7 +2,7 @@ from setuptools import setup, find_packages
 
 setup(
     name="aimedicalcoding",
-    version="0.3.1",
+    version="0.3.4",
     author="Aakash Kag, Vinay Shankar Miryala, Siva Karthik, Mohammed Rayaan",
     description=(
         "Detect missing/broken clinical codes (LOINC / SNOMED CT / RxNorm / "