Objectives: To test the Ohio American Cancer Society (ACS) model of patient navigation (PN) as it relates to reducing time to diagnostic resolution among persons with abnormal breast, cervical, or colorectal cancer screening tests or symptoms.

Methods: 862 patients from 18 primary care and specialty clinics within the Ohio State University (OSU) Medical Center and Columbus Neighborhood Health Centers with one of these abnormal findings were eligible to participate in this group-randomized study. Clinics were paired and randomized within pairs to either usual care or PN. Patients were informed about the study, consented, and completed a baseline survey. The navigators were not clinic-based, but followed the Ohio ACS-PN model where navigators mainly used phone contact from a non-clinic location. Chart review was done to capture medical information on the abnormality as well as the date of resolution and follow-up tests received. The primary analysis focused on time to resolution using shared gamma frailty models used to test for effect of navigation on time to resolution. Crude Hazard Ratios (HR) were used, as adjustment for OSU and national variables did not significantly change the hazard ratios. Assessment of effect modification was also conducted.

Results: Crude HR's deviated starting at 6 months, so that by 15 months, participants in the PN arm were 61% more likely to have their abnormality resolved (p=0.004 for difference in risks across arms; p=0.005 for increase in relative risk over time). Effect modifier analysis showed a trend for income (<$50K vs. $50K+; p=0.07) in that lower income participants benefited earlier from PN.

Participants in the navigation arm (N=475) were further assessed in terms of reported barriers and navigator actions documented in the navigation encounters. Almost half reported no barriers (n=226; 46.6%), while the remainder reported one barrier (n=113; 23.8%), or 2+ barriers (n=136; 28.6%). The most frequently reported barriers were perception/beliefs about tests or treatment (n=89), communication with providers (n=84), scheduling problems (n=65), insurance/co-pay problems (n=49), and medical/mental health co-morbidities (n=44). Most frequent action taken by navigators were providing support (n=175), referrals (n=125), and education (n=74). On average, navigated participants with barriers received 3 encounters with a PN, and 90.9% of the encounters lasted <15 minutes. The most frequent actions taken by a PN included: support (n=175); referrals (n=125); and education (n=74).

Conclusions: The OPNRP demonstrated that patients with abnormal cancer screening tests or symptoms are more likely to resolve their abnormality, if assigned to a PN program compared to those not assigned to a PN program, even when navigators are not embedded in the clinic.

Citation Information: Cancer Epidemiol Biomarkers Prev 2011;20(10 Suppl):B27.