The Relationship between Health Literacy and Non-Recommended Screening for Prostate Cancer Relative to other Malignancies
Madeline Rutan, BA1, Jesse D. Sammon, DO2, David-Dan Nguyen, DES3, Kerry L. Kilbridge, MD4, Peter Herzog, MPH5, Quoc-Dien Trinh, MD5.
1Tufts University School of Medicine, Boston, MA, USA, 2Division of Urology, Maine Medical Center, Portland, ME, USA, 3Center for Surgery and Public Health, Brigham and Women's Hospital, Boston, MA, USA, 4Lank Center for Genitourinary Oncology, Dana Farber Cancer Institute, Boston, MA, USA, 5Division of Urological Surgery, Brigham and Women's Hospital, Boston, MA, USA.
Background: Cancer screening risks overdiagnosis of indolent tumors, and subsequent intervention introduces cost and poor health outcomes. Shared decision making encourages screening choices that align with patient values; however, these conversations are difficult for patients with low health literacy (HL). We hypothesized that higher HL would support guideline-concordant screening. We sought to assess the effect of HL on non-recommended prostate cancer (PCa) screening relative to other cancers.
Methods: We examined the 2016 BRFSS, which includes HL modules. Respondents self-reported their ability to obtain and understand health information, resulting in 4 HL rankings. We assessed the effect of HL on non-recommended PCa screening and compared it to the effect on breast cancer (BCa) and cervical cancer (CC) screening. We calculated the population weighted proportion of patients in each HL category that underwent screening against USPSTF guidelines. The odds ratios of non-recommended screening for each malignancy was calculated, with the referent category of low HL.
Results: Individuals with higher HL exhibited higher rates of non-recommended screening for PCa and similar findings were seen for BCa and CC. Non-recommended PCa screening was performed in 27.4% (CI 23.7%-31.4%; p<0.001) and 47.7% (CI 44.1%-51.3%; p<0.001) of respondents with low and high HL, respectively. This compared favorably to BCa, which had the highest rates of non-recommended screening: 46.8% (CI 42.6%-51.1%; p=0.002) and 67.7% (CI 64.2%-71.1%; p=0.002) of respondents with low and high HL. Non-recommended CC screening was performed in 33.8% (CI 31.1%-36.5%; p<0.001) and 48.4% (CI 46.3%-50.5%; p<0.001) of low and high HL respondents. Individuals with high HL were significantly more likely than those with low HL to screen against recommendations for PCa (OR 1.73; CI 1.34-2.23; p<0.001), CC (OR 1.533; CI 1.31-1.80; p<0.001) and BCa (OR 8.213; CI 4.90-13.76; p<0.001).
Conclusions: Patients with higher HL undergo increased rates of non-recommended PCa screening, contrary to our initial hypothesis. While this compared favorably to very high rates of non-recommended BCa screening in the high HL populations, the association was less pronounced in CC screening. 
Back to 2020 Abstracts