Defining Cohorts Using Current Procedural Terminology Codes in Metastatic Bone Disease: Accuracy and Implications
Program/Major
MD/MPH-Epidemiology
Student Level
MD/MPH
Presentation Type
Poster
Start Date
April 2019
End Date
April 2019
Persistent Identifier
https://archives.pdx.edu/ds/psu/30950
Abstract
INTRODUCTION:
Cohorts from the electronic health record (EHR) are often defined by ontologies, including Current Procedural Terminology (CPT) codes. Research in other fields has investigated the accuracy of CPT codes relative to the procedure performed, but the accuracy of CPT codes as surrogates for diagnoses has not been adequately discussed.
METHODS:
The target population was patients with metastatic cancer to the femur. We identified all cases of prophylactic femoral or hip stabilization between 2010-2015 at Oregon Health & Science University via CPT codes. A detailed chart review was conducted to determine the procedure performed as documented in the operative note and the patient diagnosis as documented in the pathology report, operative note, and office visit notes.
RESULTS:
A total of 215 cases were identified based on CPT codes (27495 and 27187). After removing duplicate cases, 171 cases remained. The error prevalence was 7 cases out of 171 (4.1%). Of the 164 cases that were coded correctly, 87 (53.0%) were in our target population.
DISCUSSION & CONCLUSION:
In this study, CPT coding was subject to an error prevalence of 4.1%. Inaccuracies in coding have implications regarding the validity of research conducted without direct verification. In addition, CPT codes for prophylactic femur and hip stabilization are inadequate to define a cohort of patients with metastatic disease at an academic medical center. Of those cases that were coded correctly, only 53% were in our target population. This suggests a major limitation in the use of administrative databases for research on this patient population.
Defining Cohorts Using Current Procedural Terminology Codes in Metastatic Bone Disease: Accuracy and Implications
Smith Memorial Student Union, Room 296/8
INTRODUCTION:
Cohorts from the electronic health record (EHR) are often defined by ontologies, including Current Procedural Terminology (CPT) codes. Research in other fields has investigated the accuracy of CPT codes relative to the procedure performed, but the accuracy of CPT codes as surrogates for diagnoses has not been adequately discussed.
METHODS:
The target population was patients with metastatic cancer to the femur. We identified all cases of prophylactic femoral or hip stabilization between 2010-2015 at Oregon Health & Science University via CPT codes. A detailed chart review was conducted to determine the procedure performed as documented in the operative note and the patient diagnosis as documented in the pathology report, operative note, and office visit notes.
RESULTS:
A total of 215 cases were identified based on CPT codes (27495 and 27187). After removing duplicate cases, 171 cases remained. The error prevalence was 7 cases out of 171 (4.1%). Of the 164 cases that were coded correctly, 87 (53.0%) were in our target population.
DISCUSSION & CONCLUSION:
In this study, CPT coding was subject to an error prevalence of 4.1%. Inaccuracies in coding have implications regarding the validity of research conducted without direct verification. In addition, CPT codes for prophylactic femur and hip stabilization are inadequate to define a cohort of patients with metastatic disease at an academic medical center. Of those cases that were coded correctly, only 53% were in our target population. This suggests a major limitation in the use of administrative databases for research on this patient population.