Use pd NaT when using deidentified samples

deploy/cttso-ica-to-pieriandx-cdk/Changelog.md

@@ -19,6 +19,8 @@ are nested under deploy/cttso-ica-to-pieriandx-cdk.
   * Resolves https://github.com/umccr/cttso-ica-to-pieriandx/issues/190
 * Fix deidentified samples where wrong keys used for study id and participant id (https://github.com/umccr/cttso-ica-to-pieriandx/pull/194)
   * Resolves https://github.com/umccr/cttso-ica-to-pieriandx/issues/193
+* Fix deidentified samples where NAT instead of NaT null value was used (https://github.com/umccr/cttso-ica-to-pieriandx/pull/198)
+  * Resolves https://github.com/umccr/cttso-ica-to-pieriandx/issues/197
 
 ### Workarounds
 * Don't submit NTC samples (https://github.com/umccr/cttso-ica-to-pieriandx/pull/195)

utils/accession.py

@@ -538,7 +538,7 @@ def sanitise_data_frame(input_df: pd.DataFrame) -> pd.DataFrame:
     input_df["date_of_birth"] = input_df.apply(
         lambda x: x.date_of_birth
         if hasattr(x, "date_of_birth")
-        else pd.NAT,
+        else pd.NaT,
         axis="columns"
     )
     input_df["first_name"] = input_df.apply(