The Data Characterization data model was designed to facilitate rapid querying of data characterization output from the data quality and characterization activities of distributed research networks. The data model is based on the output of the Mini-Sentinel data quality review process for the Mini-Sentinel Distributed Database and similar data quality review outputs against the PCORnet Common Data Model, but it is generalizable and extensible for use with other data quality review output.
The Data Characterization data model is directly queryable in the PopMedNet System by the Data Checker requests.
The Data Characterization data model includes thirteen tables containing information regarding the distribution and missingness of variables among the data held by each data partner:
|Age||Distribution of patients' ages stratified by data partner|
Counts of diagnoses codes stratified by code, code type, and data partner
|Height||Distribution of patients' heights stratified by data partner|
|Hispanic||Ethnicity distribution stratified by data partner|
|Metadata||Completeness and availability of data stratified by table, minimum and maximum date, and data partner|
Presence of National Drug Codes stratified by data partner
|PDX||Distribution of discharge diagnosis types stratified by encounter type and data partner|
|Procedures||Counts of procedure codes stratified by code, code type, and data partner|
|Race||Race distribution stratified by data partner|
|RxAmt||Distribution of pharmacy dispensing amount supply stratified by data partner|
|RxSup||Distribution of pharmacy dispensing days supply stratified by data partner|
|Sex||Distribution of patients' sex stratified by data partner|
|Weight||Distribution of patients' weights stratified by data partner|