Authentication of Algorithm to Detect Metastases in Men with Prostate Cancer Using ICD-9 Codes

Matthew T Dolan1, Sung Kim1,2,3, Yu-Hsuan Shao2,4

  • 1Department of Radiation Oncology, University of Medicine and Dentistry of New Jersey, Robert Wood Johnson Medical School, New Brunswick, NJ 08901, USA.

Insights

A new algorithm using specific ICD-9 codes accurately detects prostate cancer (PCa) metastasis. This method offers high sensitivity and specificity for identifying metastatic PCa in patient records.

Area of Science:

  • Oncology
  • Medical Informatics
  • Health Services Research

Background:

  • Metastasis is a critical outcome for prostate cancer (PCa) patients.
  • Current methods for detecting PCa metastasis using claims data are not validated.
  • Accurate identification of metastasis is essential for patient management and research.

Purpose of the Study:

  • To develop and validate a claims-based algorithm for detecting prostate cancer (PCa) metastasis.
  • To utilize International Classification of Diseases, Ninth Revision (ICD-9) codes for accurate metastasis reporting.
  • To compare the performance of two distinct ICD-9 code-based algorithms.

Main Methods:

  • Reviewed medical records of 300 hospitalized prostate cancer (PCa) patients.
  • Established chart review as the gold standard for metastatic PCa presence.
  • Compared Algorithm A (ICD-9 codes 198.5, 197.0, 197.7, 198.3) against Algorithm B (ICD-9 code 198.5).
  • Calculated sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV).
  • Assessed agreement using Kappa statistics.

Main Results:

  • Algorithm A achieved 95% sensitivity, 100% specificity, 100% PPV, and 98.7% NPV.
  • Algorithm B showed 90% sensitivity, 100% specificity, 100% PPV, and 97.5% NPV.
  • Agreement rates were 96.8% for Algorithm A and 93.5% for Algorithm B.

Conclusions:

  • The algorithm using ICD-9 codes 198.5, 197.0, 197.7, or 198.3 effectively detects prostate cancer (PCa) metastasis.
  • This claims-based approach demonstrates high performance metrics (sensitivity, specificity, PPV, NPV).
  • The developed algorithm provides a validated tool for identifying PCa metastasis in administrative data.
Abstract

Related Concept Videos