Is physician-patient language discordance a barrier for clinical trial enrollment in patients with cancer? A real-life study.

Abstract

e18539 Background: Clinical trials are essential for enhancing cancer care. Participation is concerning in racial and ethnic minority groups historically under-represented in research. So, it is crucial to note that, despite Hispanics and African-Americans being 18.9% and 13.6% of the US population, each only represents 4% of the patients enrolled in clinical trials. We address effective doctor-patient communication as a fundamental clinical function in establishing an effective doctor-patient relationship, and is vital in delivering high-quality care. In this study, we question whether physician-patient language concordance affects clinical trial enrollment. Methods: The study evaluated 982 patients diagnosed with cancer who consented to experimental clinical trials in a private practice in Houston, Texas, from 2008-2022. All trials had approved language translations for English and Spanish. We used logistic regression to model the probability of treatment, while adjusting for the effects of cancer type, gender, race, ethnicity, and language (same or different as provider). Results: 982 patients with multiple cancer types (hematological, breast, thoracic/respiratory, genitourinary, gastrointestinal, head and neck and CNS) were included in the study. The most represented tumor types were GI (32%), Breast (24%) and Thoracic (23%). 66% of patients were successfully enrolled in an experimental clinical trial and started treatment, and 95% spoke the same language as their providers. 14% of patients spoke a different language than English, Spanish being the most commonly spoken language other than English spoken by the patient and provider. The study has adequate minority representation (Caucasian 45%, Hispanic 30%, African-American 18% and Asian 5%), and equal gender distribution (52% were female). After evaluating the results, it was found that there was no statistically significant association between physician-patient language and enrollment rates (p = 0.3). It was also found that there was no impact when the assessment was divided by tumor type, gender (p = 0.8) or ethnicity. It also evaluated the rate of consent withdrawal, with only 4% of patients withdrawing consent, showing no statistically significant association with language concordance. Conclusions: In conclusion, our study confirms no significant difference in cancer patients’ enrollment rate in clinical trials if there is language concordance between physician and patient. The efforts of the medical workforce to use translators and translated versions of informed consents, surveys or outcome assessments, when available, seem enough for our patients to collect all the information required to agree to continue enrollment. It is required to further evaluate more variables that impact enrollment in minorities to stop this disparity in cancer care.