For all tests of association, P values less than .05 were considered statistically significant. Analyses were performed at the University of North Carolina at Chapel Hill, which was the CanCORS site for all participants included in this study. We used CanCORS core data (version 1.16), medical record data (version 1.12), and participant survey data (version 1.12). All analyses were performed using SAS version 9.3 for Windows (SAS Institute, Cary, North Carolina).

Results

Table 1 presents descriptive statistics among the 347 participants for whom health literacy was assessed. Despite our goal of a sample with approximately equal numbers with adequate and marginal/inadequate levels of health literacy, 105 (30%) were categorized as marginal/ inadequate. Compared with those with an adequate level of health literacy, participants with marginal/inadequate health literacy were more frequently men, nonwhite, at least 65 years of age, not currently married or living with a partner, had not completed high school, and had annual household incomes below $40,000.


Continue Reading

Among participants with stage 3/4 disease having an adequate level of health literacy increased the odds of receiving chemotherapy compared with those with a marginal/inadequate level of health literacy (odds ratio [OR], 3.29; 95% confidence interval [CI], 1.23–8.80). However, across all stages, having an adequate level of health literacy did not increase the odds of presenting with early-stage compared with late-stage disease (OR, 1.11; 95% CI, 0.68–1.80).

For participants with stage 3/4 disease, Table 2 presents responses to survey questions about participant beliefs and preferences regarding chemotherapy, their roles in deciding whether to receive chemotherapy, and their communication with health care professionals about chemotherapy. Of the 130 participants with stage 3/4 disease, 89 (68%) had an adequate level of health literacy and 41 (32%) had inadequate/marginal levels of health literacy — percentages comparable with those for the sample across all stages. We found no statistically significant differences in participant responses by level of health literacy. Participants of all levels of health literacy thought that, after discussing chemotherapy with a health care professional, the treatment was likely to help them live longer (89% of those with an adequate level of health literacy and 100% of those with marginal/inadequate level of health literacy) and to help them with issues related to their cancers (82% of those with an adequate level of health literacy and 100% of those with a marginal/inadequate level of health literacy).

Although the differences were not statistically significant, participants with stage 3/4 disease and an adequate level of health literacy played a more prominent role in deciding whether to receive chemotherapy than those with inadequate/marginal levels of health literacy (see Table 2). Among those with informative responses, 75 of 83 (90%) participants with an adequate level of health literacy reported either making the decision to receive chemotherapy themselves or together with their health care professional (in contrast with the health care professional alone making the decision) compared with 20 of 25 (80%) among those with an inadequate/marginal level of health literacy. Three of the 83 participants (4%) with an adequate level of health literacy and informative responses reported that their health care professional made the decision about chemotherapy with little or no patient input compared with 5 of 25 (20%) among those with a marginal/inadequate level of health literacy and informative responses.

In the unadjusted regression of survival on health literacy, those with a marginal/inadequate level of health literacy had increased odds of being deceased at last observation compared with those with an adequate level of health literacy, but the effect was not statistically significant and disappeared when conditioned on cancer stage as well as demographic and socioeconomic covariates (Table 3). Male sex and stage 3/4 disease were associated with greater odds of being deceased at last observation, but no other notable effects were detected.