Background: Over the past decade, genomic testing has become standard of care for metastatic non-small cell lung cancer (NSCLC). These tests qualify patients for additional anti-cancer therapies and should be performed in all patients. Small scale studies at the institutional level have revealed that there may be disparities in genomic testing in NSCLC and not all patients may have similar access to care. In this study, we use the IBM Explorys Electronic Health Record (EHR) database to conduct a nationwide retrospective, observational study to understand how gender, race, insurance type, and spoken language impacts the rate of genomic testing in metastatic NSCLC patients. Methods: From Jan 1st, 2015 to Dec 31st, 2020, the IBM Explorys EHR database comprised 128,119 lung cancer patients using the SNOMED-CT concept of Primary Malignant Neoplasm of the Lung (CID 93880001). As structured staging information was not available, metastatic NSCLC patients were imputed by removing patients who received thoracic surgeries (presumably stage I or II) and those who received radiation therapy (presumably stage III). Following imputation, 120,470 patients with metastatic NSCLC were queried for testing for EGFR, ALK, ROS1, and/or RET. Odds ratios and chi-squared tests were computed for gender, race, insurance type, and spoken language comparing patients that received genomics testing to those who did not. Results: Genomic testing was taken significantly more by male patients (OR: 1.35, p<0.0001), and by Caucasian patients (OR: 1.39, p<0.0001). Compared to the public insurance plans, the genomic testing was significantly more in patients with private insurance plans (OR: 2.48, p<0.0001) and self-pay patients (OR: 2.84, p<0.0001). Patients speaking English as their first language significantly less likely took genomic testing (OR: 0.81, p<0.05). Conclusions: This study aims to identify gaps in health disparities in gender, race/ethnicity, and insurance type for genomic testing that should be standard practice. Future investigation and attention to this issue appears necessary to begin moving from documenting disparities, to understanding them, and ultimately to reducing them.Demographic
Category
Patients with Genomic Testing
Patients without Genomic Testing
Odds Ratio with 95% C.I.
(vs. reference category)
Gender**
Male
997 (48.40%)
48597 (41.07%)
1.346 ± 0.112
(vs. Female)
Female
1063 (51.60%)
69731 (58.93%)
Race/Ethnicity**
Caucasian
1352 (81.49%)
77849 (76.05%)
1.386±0.163
(vs. Non-Caucasian)
Non-Caucasian
307 (18.51%)
24493 (23.93%)
Insurance Type**
Public
437 (22.67%)
39267 (41.79%)
Private
1163 (60.32%)
42086 (44.79%)
2.483±0.260 (vs. Public)
Self-pay
303 (15.72%)
9594 (10.21%)
2.838±0.391 (vs. Public)
Language*
English
1837 (92.31%)
105654 (93.67%)
0.811±0.124
(vs. Non-English)
Non-English
153 (7.69%)
7135 (6.33%)