Use of ICD-9 coding as a proxy for stage of disease in lung cancer

• Published in: Pharmacoepidemiology and drug safety Vol. 11; no. 8; pp. 709 - 713
• Main Authors: Thomas, Simu K., Brooks, Sandra E., Daniel Mullins, C., Baquet, Claudia R., Merchant, Sanjay
• Format: Journal Article
• Language: English
• Published: Chichester, UK John Wiley & Sons, Ltd 01-12-2002
• Subjects: cancer staging; claims; Humans; International Classification of Diseases; lung neoplasms; Lung Neoplasms - pathology; Retrospective Studies; Severity of Illness Index; tumor registry; validity
• ISSN: 1053-8569; 1099-1557
• DOI: 10.1002/pds.759

Objective
To determine the validity of using ICD‐9‐CM codes indicating metastases as a proxy to classify lung cancer patients by stage of disease.
Research Design
This retrospective database analysis used diagnosis codes to classify subjects to either localized or advanced‐stage disease and then compared this classification to the tumor registry staging, which was considered as the ‘gold standard.’
Subjects and Measures
Study subjects included all lung cancer patients treated at our academic institution during 1996–1997, who were also members of a large insurance company. Data were derived from inpatient cancer‐related claims linked with our institution's tumor registry data. Advanced‐stage disease (stages II to IV) was defined by claims indicating lymph node involvement or metastases (ICD‐9 codes 196–199.1). The tumor registry stagings of the disease for these patients were clustered into two groupings, stages 0–I (localized) and stages II–IV (advanced).
Results
Tumor registry entries were identified for 66/77 (85.7%) patients. A total of 19 out of 22 local disease patients (sensitivity of 86.4%) and 30 out of 44 advanced disease patients (sensitivity of 68.2%) were classified correctly by ICD‐9 code. A total of 19 out of 33 patients with local disease codes (PPV of 57.6%) and 30 out of 33 patients with advanced disease codes (PPV = 90.9%) were properly identified.
Conclusions
For a population of lung cancer patients in an academic institution who were under a private insurance plan, the ICD‐9 coding was associated with a sensitivity and positive predictive values that were consistent with previously reported estimates using Medicare–SEER data. The use of such data to classify patients to disease stages should be executed with caution as under‐reporting might exist. Continued attention to discharge ing will be needed to improve the validity of this technique. Copyright © 2002 John Wiley & Sons, Ltd.