Self-Sampling Tools for Cancer Screening in Primary Care: Evidence from a Randomized Trial with Underserved Patients

Context. Implementation of routine screening for cervical cancer and colorectal cancer (CRC) is suboptimal. Primary care is an ideal venue for disseminating self-sampling screening tools, allowing patients to collect a sample at home and send it to a lab for testing. These tools can increase population-level cancer screening and reduce cancer disparities.

Objective. Assess the preliminary outcomes and acceptability of an intervention “package” to increase cancer screening using self-sampling tools among underserved patients. This package includes a novel tool (i.e., HPV self-sampling for cervical cancer screening) and a more established tool (i.e., fecal immunochemical test [FIT] for CRC screening).

Study Design and Analysis. Pilot randomized controlled trial, with telephone-administered surveys at baseline and 10-weeks follow-up. We used Fisher’s exact tests to assess differences in outcome measures by study arm.

Setting. Federally-qualified health centers in rural and racially-segregated counties in Pennsylvania.

Population Studied. Average-risk, female patients, ages 50-65, who were eligible but out-of-date with both cervical cancer screening and CRC screening (n=48).

Intervention. Participants in the intervention arm received an intervention package providing educational materials and self-sampling tools for cervical and CRC screening. Participants in the standard of care arm received a reminder to schedule cancer screening.

Outcome Measures. The primary outcomes were self-reported receipt of screening for cervical cancer and for CRC at follow-up. Additional outcomes among participants in the intervention arm were return, results, and acceptability of self-sampling tools.

Results. Cervical cancer screening at follow-up was higher for the intervention vs. standard of care arm (89% vs. 11%, p<.001). CRC screening was also higher for the intervention vs. standard of care arm (89% vs. 11%, p<.001). Among intervention participants, 71% returned both self-sampling tools, and of those, 18% had abnormal results on cervical cancer screening, and 24% had abnormal results on CRC screening. Acceptability of self-sampling tools was very high.

Conclusions. A bundled package of self-sampling tools improved screening for cervical cancer and CRC and was acceptable for among underserved patients in primary care. These tools have the potential to increase screening and reduce disparities in cancer outcomes.