(143) Validation of a Case-Finding Algorithm for Endometrial Cancer Using ICD10 Diagnostic Codes in Insurance Claims

Publication Number: 1188

Background: Although an algorithm to identify cases of endometrial cancer in insurance claims using International Classification of Disease version 9 Clinical Modification (ICD-9-CM) codes has been published, no such algorithm has been ascertained for ICD-10-CM codes.

Objectives: To evaluate the positive predictive value (PPV) of an endometrial cancer case identification algorithm using ICD-10-CM diagnosis codes from US insurance claims for use in a planned postmarketing safety study.

Methods: Provisional incident endometrial cancer cases were identified with the algorithm from 2016 through 2020, among women aged ≥50 years without prior hysterectomy or endometrial ablation. The algorithm used diagnostic codes for malignant neoplasms of uterine sites (C54.x), excluding C54.2 (malignant neoplasm of myometrium). A random sample of medical records was requested for adjudication independently by two clinical nurses, with discrepancy resolution by a physician. Confirmed cases showed biopsy evidence of endometrial cancer, documentation of cancer staging, or hysterectomy following diagnosis. We estimated the PPV of the case finding algorithm with 95% confidence intervals (CI), excluding cases remaining provisional.

Results: Among 3,143 provisional cases, 444 were randomly selected for review; 303 unique medical records were received during the 14-week collection period (response rate: 68.2%). Of 294 provisional cases adjudicated, 85% were from outpatient settings (n=249). Mean age at diagnosis was 69.3 years (standard deviation: 9.4 years), 42.7% (n=126) had type I or type II diabetes mellitus, and 27.5% (n=81) had endometrial hyperplasia. Of the 294 adjudicated records, 223 were confirmed as endometrial cancer cases, 2 were probable cases, 40 were non-cases, and 29 remained provisional cases due to inadequate information. The PPV (95% CI) of provisional cases with adequate medical record information was 84.2% (79.2%, 88.3%). Confirmed cases were mostly of Type I – endometrioid adenocarcinomas (78.5%) and of unknown estrogen receptor status (79.4%).

Conclusions: We found that an algorithm using ICD-10-CM diagnosis codes developed to identify endometrial cancer cases in health insurance claims had a sufficient PPV to use in a planned post marketing safety study.