PURPOSE: In this nationwide registry study, we investigated socioeconomic and structural patterns in referral to phase I cancer trials in a case-control study design.

METHODS: Personal identification numbers on all Danish patients referred to the Danish Phase I Unit at Rigshospitalet from 2005 to 2016, and a control group matched on age, sex, type of cancer, year of diagnosis, and time from diagnosis to referral ensured individual-level linkage between several registries. We examined the association between nonclinical factors-indicators of socioeconomic position and distance to the Phase I Unit-and referral using a conditional logistic regression analysis adjusted for several clinical factors. Association between nonclinical factors and enrollment once referred was examined with a Cox proportional hazards regression analysis in an historical cohort study design.

RESULTS: Complete data were available for 1,026 (84%) of 1,220 referred patients. Significantly decreased odds for referral were identified for patients with long distance to the Phase I Unit compared with short distance (adjusted odds ratio [OR], 0.35; 95% CI, 0.30 to 0.41), for less education (less than 9 years) compared with more (more than 12 years; OR, 0.69; 95% CI, 0.56 to 0.91), and for belonging to the lowest income quintile compared with the highest (OR, 0.78; 95% CI, 0.62 to 0.97). Medium education (9 to 12 years) compared with more, being outside the workforce compared with being within, and living alone compared with living with a partner were also negatively associated with referral. Among patients referred, 252 enrolled in a trial. Nonclinical factors were not associated with enrollment.

CONCLUSION: On the basis of individual long-term registry data from an unselected cohort, novel anticancer therapies seem to be tested on a socially selected group of patients with cancer.