Self-efficacy difference among patients with cancer with different socioeconomic status: application of latent class analysis and standardization and decomposition analysis.

Abstract Introduction Although the relationship between partial socioeconomic status (SES) and self-efficacy has been studied in previous studies, few research have examined self-efficacy difference among patients with cancer with different SES. Methods A cross-sectional survey involving 764 patients with cancer was completed. Latent class analysis (LCA) was applied to identify distinct groups of patients with cancer using four SES indicators (education, income, employment status and health insurance status). Standardization and decomposition analysis (SDA) was then used to examine differences in patients’ self-efficacy among SES groups and the components of the differences attributed to confounding factors, such as gender, age, anxiety, depression and social support. Results Participants were classified into four distinctive SES groups via using LCA method, and the observed self-efficacy level significantly varied by SES groups; as theorized, higher self-efficacy was associated with higher SES. The self-efficacy differences by SES groups were decomposed into “real” group differences and factor component effects that are attributed to group differences in confounding factor compositions. Conclusion Self-efficacy significantly varies by SES. Social support significantly confounded the observed differences in self-efficacy between different SES groups among Chinese patients with cancer.