User login
Expansion of an Intervention to Ensure Accuracy and Usefulness of a SQL Code Identifying Oncology Patients for VACCR
Purpose
The Veterans Affairs Central Cancer Registry (VACCR) is a data management system for cancer surveillance and epidemiologic-based efforts, seeking to reduce the overall cancer burden. In 2024, the local VACCR successfully implemented a Structured Query Language (SQL) code, created to identify documents in the electronic medical record (EMR) with associated ICD-10 codes matching reportable cancer cases in the Surveillance, Epidemiology, and End Results (SEER) list. In 2025, code application expansion began at four additional VISN9 sites.
Outcomes Studied
Accuracy and usefulness of SQL code application in a significantly larger population and a diagnosis-specific population.
Methods
Local Cancer Program leadership collaborated with VISN9 leadership to expand the SQL code to the four sites’ EMR, identifying the Veteran’s name, social security number, location by city/state/county, and visit-associated data including location, ICD-10 code, and visit year. Data validation focused on ICD- 10-specific data and quality replication.
Results
After SQL code application to Mt Home TN VACCR data, 750 unique, randomized charts from 2015-2025 were selected for accuracy review. Data validation found that 90.5% (679) had a reportable cancer; 14.9% (112) were not entered into VACCR. 9.5% (71) were not reportable. The SQL code was applied to Lexington data to identify colorectal cancer (CRC) (ICD-10 codes C17-C21.9). 746 charts from 2015-2025 were identified. 88.9% (663) had a reportable CRC; 14.9% (111) of those were not entered into VACCR, and 11% (83) were not reportable. Most cases not entered into VACCR at both sites were cases in which the majority of care was provided through Care in the Community (CITC). Historically, identification of CITC-provided oncologic care has been manual and notoriously difficult.
Conclusions
This study demonstrated the feasibility and accuracy of the SQL code in the identification of Veterans with diagnoses matching the SEER list in a large population and at a diagnosis-specific level. VISN-wide use of the report will increase efficiency and timeliness of data entry into VACCR, especially related to care provided through CITC. An improved understanding of oncologic care in the VISN would provide critical data to VISN executive leadership, enabling them to advocate for resources, targeted interventions, and access to care.
Purpose
The Veterans Affairs Central Cancer Registry (VACCR) is a data management system for cancer surveillance and epidemiologic-based efforts, seeking to reduce the overall cancer burden. In 2024, the local VACCR successfully implemented a Structured Query Language (SQL) code, created to identify documents in the electronic medical record (EMR) with associated ICD-10 codes matching reportable cancer cases in the Surveillance, Epidemiology, and End Results (SEER) list. In 2025, code application expansion began at four additional VISN9 sites.
Outcomes Studied
Accuracy and usefulness of SQL code application in a significantly larger population and a diagnosis-specific population.
Methods
Local Cancer Program leadership collaborated with VISN9 leadership to expand the SQL code to the four sites’ EMR, identifying the Veteran’s name, social security number, location by city/state/county, and visit-associated data including location, ICD-10 code, and visit year. Data validation focused on ICD- 10-specific data and quality replication.
Results
After SQL code application to Mt Home TN VACCR data, 750 unique, randomized charts from 2015-2025 were selected for accuracy review. Data validation found that 90.5% (679) had a reportable cancer; 14.9% (112) were not entered into VACCR. 9.5% (71) were not reportable. The SQL code was applied to Lexington data to identify colorectal cancer (CRC) (ICD-10 codes C17-C21.9). 746 charts from 2015-2025 were identified. 88.9% (663) had a reportable CRC; 14.9% (111) of those were not entered into VACCR, and 11% (83) were not reportable. Most cases not entered into VACCR at both sites were cases in which the majority of care was provided through Care in the Community (CITC). Historically, identification of CITC-provided oncologic care has been manual and notoriously difficult.
Conclusions
This study demonstrated the feasibility and accuracy of the SQL code in the identification of Veterans with diagnoses matching the SEER list in a large population and at a diagnosis-specific level. VISN-wide use of the report will increase efficiency and timeliness of data entry into VACCR, especially related to care provided through CITC. An improved understanding of oncologic care in the VISN would provide critical data to VISN executive leadership, enabling them to advocate for resources, targeted interventions, and access to care.
Purpose
The Veterans Affairs Central Cancer Registry (VACCR) is a data management system for cancer surveillance and epidemiologic-based efforts, seeking to reduce the overall cancer burden. In 2024, the local VACCR successfully implemented a Structured Query Language (SQL) code, created to identify documents in the electronic medical record (EMR) with associated ICD-10 codes matching reportable cancer cases in the Surveillance, Epidemiology, and End Results (SEER) list. In 2025, code application expansion began at four additional VISN9 sites.
Outcomes Studied
Accuracy and usefulness of SQL code application in a significantly larger population and a diagnosis-specific population.
Methods
Local Cancer Program leadership collaborated with VISN9 leadership to expand the SQL code to the four sites’ EMR, identifying the Veteran’s name, social security number, location by city/state/county, and visit-associated data including location, ICD-10 code, and visit year. Data validation focused on ICD- 10-specific data and quality replication.
Results
After SQL code application to Mt Home TN VACCR data, 750 unique, randomized charts from 2015-2025 were selected for accuracy review. Data validation found that 90.5% (679) had a reportable cancer; 14.9% (112) were not entered into VACCR. 9.5% (71) were not reportable. The SQL code was applied to Lexington data to identify colorectal cancer (CRC) (ICD-10 codes C17-C21.9). 746 charts from 2015-2025 were identified. 88.9% (663) had a reportable CRC; 14.9% (111) of those were not entered into VACCR, and 11% (83) were not reportable. Most cases not entered into VACCR at both sites were cases in which the majority of care was provided through Care in the Community (CITC). Historically, identification of CITC-provided oncologic care has been manual and notoriously difficult.
Conclusions
This study demonstrated the feasibility and accuracy of the SQL code in the identification of Veterans with diagnoses matching the SEER list in a large population and at a diagnosis-specific level. VISN-wide use of the report will increase efficiency and timeliness of data entry into VACCR, especially related to care provided through CITC. An improved understanding of oncologic care in the VISN would provide critical data to VISN executive leadership, enabling them to advocate for resources, targeted interventions, and access to care.