Self-Collection for Cervical Cancer Screening in a Safety-Net Setting: The PRESTIS Randomized Clinical Trial

There are limited data on the effectiveness of mailed self-collection to increase cervical cancer screening (CCS) participation in underresourced health care settings.

To compare the effectiveness of mailed self-collection kits, with and without patient navigation, to telephone reminders to increase CCS in a safety-net health system.

This pragmatic, parallel, single-blinded, randomized clinical trial within a publicly funded safety-net health system in Houston, Texas, compared (1) telephone reminder (TR) for clinic-based screening, (2) TR with mailed self-collection (SC), and (3) TR with mailed SC and patient navigation among a random sample of CCS-eligible patients not up to date with CCS, including those with no CCS on record. The trial was conducted from February 20, 2020, to August 31, 2023.

All groups received a TR by a patient navigator to attend clinic-based CCS. In the SC and SC with patient navigation groups, participants were additionally mailed a self-collection kit to their home as an alternative to clinic-based CCS. In the SC with patient navigation group, the mailed kit was followed by a patient navigation telephone call.

CCS participation was defined as attendance for clinic-based screening or return of a mailed self-collection kit within 6 months of randomization and determined through electronic health record review.

Of the 2474 participants in the intent-to-screen analyses (median [IQR] age, 49 [39-57] years), 2325 (94.0%) were from racial or ethnic minoritized populations (1655 [66.9%] identifying as Hispanic or Latino, 82 [3.3%] as non-Hispanic Asian, 535 [21.6%] as non-Hispanic Black or African American, and 53 [2.1%] as other or unknown race, including American Indian or Alaska Native and Native Hawaiian or Other Pacific Islander), and 1388 (56.1%) were covered by the county's publicly funded financial assistance program. At 6 months, 144 of 828 participants (17.4%) in the TR group, 340 of 828 (41.1%) in the SC group, and 381 of 818 (46.6%) in the SC with patient navigation group had participated in CCS. Compared to TR, relative participation was 2.36 (95% CI, 1.99-2.80) times higher for SC and 2.68 (95% CI, 2.27-3.16) times higher for SC with patient navigation; screening difference was 23.7% (95% CI, 19.4%-27.9%) for SC and 29.2% (95% CI, 24.9%-33.5%) for SC with patient navigation.

In this randomized clinical trial in a safety-net health system, SC was effective for increasing CCS participation among underscreened patients; there were modest additional gains from SC with patient navigation. The large increase in CCS participation using SC compared to TR suggest that SC should be considered in safety-net settings with suboptimal CCS coverage.

ClinicalTrials.gov Identifier: NCT03898167.