Uncertainty in Illness and Health Literacy in Pancreatic Cancer Patients

Monday, 17 September 2018

Rae Brana Reynolds, PhD, RN, ANP-BC
Cizik School of Nursing, The University of Texas - Houston, Houston, TX, USA
Geri Wood, PhD, MS, BS, RN, FAAN
Nursing Systems Department, The University of Texas School of Nursing at Houston, Houston, TX, USA
Terri S. Armstrong, PhD, NP, MS, BSN, RN, ANP-BC, FAANP, FAAN
Center for Cancer Research, National Cancer Institute, Bethesda, MD, USA
Nikhil S. Padhye, PhD
Center for Nursing Research, University of Texas Health Science Center at Houston, Houston, TX, USA
Matthew Katz, MD, FACS
Surgical Oncology, MD Anderson Cancer Center, Houston, TX, USA

Uncertainty in Illness and Health Literacy in Pancreatic Cancer Patients

Background: Despite a shared link to cognitive processing suggested by their definitions, information on the association between uncertainty (UC) and health literacy (HL) is scarce. Their relationship has not been studied in pancreatic cancer patients whose complex illness experience predispose to uncertainty and require proficient health literacy to manage effectively.

Aims: To evaluate UC and HL in pancreatic cancer patients, examine their bivariate correlation, and determine significant predictors.

Methods: This descriptive, cross-sectional study was conducted in a comprehensive cancer center. UC was measured with the Mishel Uncertainty in Illness Scale -Community and HL with the Cancer Health Literacy Test 30. Spearman’s rho tested correlation and linear regression models tested for predictors. Bias corrected, accelerated bootstrap was used when regression residuals violated normality.

Results: The sample (N=82) was predominantly male (55%), White/Caucasian (79%), married 74%), and receiving neo-adjuvant treatment in anticipation of potential surgical resection (49%). Mean age was 64.59 years ranging from 30 to 80. A significant but weak correlation was noted between UC and HL (rs = -.24, p = .032). HL was not a significant predictor of uncertainty after adjusting for age, gender, education, race/ethnicity, and phase of care. Education was a significant predictor of UC (p = .001; ηp 2 = .217) and HL (p =.003; ηp 2 = .174). Phase of care was a significant predictor of UC (p = .001; ηp 2 = .221).

Conclusion: UC and HL had a significant albeit weak correlation. HL was not an significant predictor of UC when accounting for other clinical and demographic predictors. Education was a significant predictor of UC and HL. Significant differences in the ability to interpret health events were found through the different phases of the pancreatic cancer experience. Sample homogeneity restricted inferences and generalizability on effects of race/ethnicity.