School of Medicine and Health Sciences Poster Presentations

Poster Number

244

Document Type

Poster

Publication Date

3-2016

Abstract

Identifying Venous Thromboembolism in Cancer Patients Using Veterans Affairs Administrative Data

Noah Ravenborg1, Najeebah A. Bade, M.D.2, Dalia Abdelaziz Mobarek, M.D.1, 2, 3

1 George Washington University School of Medicine and Health Sciences, Washington, D.C. 2 Department of Medicine, Section of Hematology, Veterans Affairs Medical Center, Washington, D.C., 3 Department of Medicine, Section of Hematology & Medical Oncology, George Washington University Medical Center, Washington, D.C. Noah Ravenborg Cell phone: (310)-804-0848, Email: nravenborg@gwu.edu, MS2

IMPORTANCE: The validity of administrative data in identifying diagnoses within the Veterans Affairs (VA) database, including viral hepatitis, cirrhosis, H. pylori, and cancer metastasis has been reported. International Classification of Diseases-9 (ICD-9) validity for venous thromboembolism (VTE) in cancer patients within the VA database is unknown.

OBJECTIVES: Determine the validity of ICD-9 codes for VTE in cancer patients in a local VA database.

DESIGN: We conducted a retrospective study utilizing data from the Washington, DC VA Cancer Registry and the Electronic Health Records (EHR). VTE diagnosis was identified using the ICD-9 codes for Pulmonary Embolism and Thrombosis, with subsequent confirmation via comprehensive chart reviews.

SETTING: Veterans Affairs Medical Center, Washington, DC.

PARTICIPANTS: 6,678 patients with cancer were identified from 1999-2015 using the cancer registry. We applied the algorithms above and identified subjects with VTE in the database.

EXPOSURES: N/A

MAIN OUTCOME(S) AND MEASURE(S): The primary outcome study measurement was the validity of ICD-9 for VTE in cancer patients. Positive (PPV), negative predictive (NPV) values, sensitivity, specificity, and likelihood ratios were calculated. Our hypothesis, that ICD-9 codes alone are not predicative of VTE diagnosis in our cancer registry, was formulated during data collection after initial chart reviews yielded false positives.

RESULTS: Initial application of ICD-9 codes for VTE among 6,678 subjects yielded 616 VTE. Chart reviews confirmed the presence of VTE among 403/616. The ICD-9 codes had a 65% PPV, 95% NPV, 57% sensitivity and 96.4% specificity. Estimated prevalence of VTE in 6,678 subjects is 10.6%. Positive and negative likelihood ratios were 15.8 and 0.45, respectively.

CONCLUSIONS AND RELEVANCE: Within our local VA database, ICD-9 codes for VTE are not sensitive for identifying patients with VTE. Accurate ICD coding by physicians is paramount for patient care and research purposes. There is a lack of data on physician coding education. A systematic literature review revealed variable ICD-9 code validity based on the population of interest making larger studies challenging with added need for manual abstraction for validation. Provider education on proper use of ICD code is important for health outcomes research perspective and would allow for more accurate retrospective research.

REGISTRATION: N/A

Creative Commons License

Creative Commons License
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License.

Open Access

1

Comments

Presented at: GW Research Days 2016

 

Identifying Venous Thromboembolism in Cancer Patients Using Veterans Affairs Administrative Data

Identifying Venous Thromboembolism in Cancer Patients Using Veterans Affairs Administrative Data

Noah Ravenborg1, Najeebah A. Bade, M.D.2, Dalia Abdelaziz Mobarek, M.D.1, 2, 3

1 George Washington University School of Medicine and Health Sciences, Washington, D.C. 2 Department of Medicine, Section of Hematology, Veterans Affairs Medical Center, Washington, D.C., 3 Department of Medicine, Section of Hematology & Medical Oncology, George Washington University Medical Center, Washington, D.C. Noah Ravenborg Cell phone: (310)-804-0848, Email: nravenborg@gwu.edu, MS2

IMPORTANCE: The validity of administrative data in identifying diagnoses within the Veterans Affairs (VA) database, including viral hepatitis, cirrhosis, H. pylori, and cancer metastasis has been reported. International Classification of Diseases-9 (ICD-9) validity for venous thromboembolism (VTE) in cancer patients within the VA database is unknown.

OBJECTIVES: Determine the validity of ICD-9 codes for VTE in cancer patients in a local VA database.

DESIGN: We conducted a retrospective study utilizing data from the Washington, DC VA Cancer Registry and the Electronic Health Records (EHR). VTE diagnosis was identified using the ICD-9 codes for Pulmonary Embolism and Thrombosis, with subsequent confirmation via comprehensive chart reviews.

SETTING: Veterans Affairs Medical Center, Washington, DC.

PARTICIPANTS: 6,678 patients with cancer were identified from 1999-2015 using the cancer registry. We applied the algorithms above and identified subjects with VTE in the database.

EXPOSURES: N/A

MAIN OUTCOME(S) AND MEASURE(S): The primary outcome study measurement was the validity of ICD-9 for VTE in cancer patients. Positive (PPV), negative predictive (NPV) values, sensitivity, specificity, and likelihood ratios were calculated. Our hypothesis, that ICD-9 codes alone are not predicative of VTE diagnosis in our cancer registry, was formulated during data collection after initial chart reviews yielded false positives.

RESULTS: Initial application of ICD-9 codes for VTE among 6,678 subjects yielded 616 VTE. Chart reviews confirmed the presence of VTE among 403/616. The ICD-9 codes had a 65% PPV, 95% NPV, 57% sensitivity and 96.4% specificity. Estimated prevalence of VTE in 6,678 subjects is 10.6%. Positive and negative likelihood ratios were 15.8 and 0.45, respectively.

CONCLUSIONS AND RELEVANCE: Within our local VA database, ICD-9 codes for VTE are not sensitive for identifying patients with VTE. Accurate ICD coding by physicians is paramount for patient care and research purposes. There is a lack of data on physician coding education. A systematic literature review revealed variable ICD-9 code validity based on the population of interest making larger studies challenging with added need for manual abstraction for validation. Provider education on proper use of ICD code is important for health outcomes research perspective and would allow for more accurate retrospective research.

REGISTRATION: N/A

 

To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.