Self-sampling tools to increase cancer screening among underserved patients: a pilot randomized controlled trial

Abstract

Background Screening can reduce cancer mortality, but uptake is suboptimal and characterized by disparities. Home-based self-sampling can facilitate screening for colorectal cancer (with stool tests, eg, fecal immunochemical tests [FIT]) and for cervical cancer (with self-collected human papillomavirus [HPV] tests), especially among patients who face barriers to accessing healthcare. Additional data are needed on feasibility and potential effects of self-sampling tools for cancer screening among underserved patients. Methods We conducted a pilot randomized controlled trial with patients (female, ages 50-65, out-of-date with colorectal and cervical cancer screening) recruited from federally-qualified health centers in rural and racially-segregated counties in Pennsylvania. Participants in the Standard of care arm (n = 24) received screening reminder letters. Participants in the Self-sampling arm (n = 24) received self-sampling tools for FIT and HPV testing. We assessed uptake of screening (10-week follow-up), self-sampling screening outcomes, and psychosocial variables. Analyses used Fisher’s exact tests to assess the effect of study arm on outcomes. Results Cancer screening was higher in the Self-sampling arm than the Standard of care arm (colorectal: 75% vs 13%, respectively, odds ratio [OR]=31.32, 95% confidence interval [CI]=5.20-289.33; cervical: 79% vs 8%, OR = 72.03, 95% CI = 9.15-1141.41). Among participants who returned the self-sampling tools, the prevalence of abnormal findings was 24% for colorectal and 18% for cervical cancer screening. Cancer screening knowledge was positively associated with uptake (p<.05). Conclusions Self-sampling tools can increase colorectal and cervical cancer screening among unscreened, underserved patients. Increasing the use of self-sampling tools can improve primary care and cancer detection among underserved patients. Clinical Trials Registration Number STUDY00015480.