Symptom clusters and their influencing factors among Vietnamese women after cancer treatment

Abstract

This study aimed to (1) identify symptom clusters in Vietnamese women with cancer and (2) examine the factors influencing those identified clusters. A cross-sectional study was conducted in 5 hospitals across Vietnam from September to December 2023. A total of 217 valid data sets from women with cancer were included. Exploratory factor analysis was applied to identify symptom clusters, followed by structural equation modeling to confirm the underlying structure. Fatigue and appetite loss were recognized as the most common symptoms. The exploratory factor analysis showed two distinct groups of factors, occupying 54.66% of total variance: fatigue, appetite loss, pain, sleep issues, hair loss, nausea, and sexual issues (Factor 1-physical cluster) and mood issues, personal stress, depression, and anxiety (Factor 2-psychological cluster). Multiple linear regression analysis revealed that place of residence (B = 0.318; p < 0.05) and occupation (B = 0.263; p < 0.05) were significant predictors for the physical cluster. For the psychological cluster, physical activity (B = - 0.599, p < 0.001) and the presence of chronic diseases (B = - 0.328, p < 0.05) were significant influencing factors, with physical activity demonstrating a strong negative association. The physical and psychological symptom clusters underlie the multidimensional nature of symptom burden in women with cancer and highlight the need for integrative, gender-responsive care models. Culturally tailored, cluster-based interventions are required to enhance survivorship care and patients' outcomes and quality of life.

Keywords: Cancer; Influencing factors; Symptom clusters; Vietnam; Women.

Fig. 1

Latent variables (Physical symptom cluster, Psychological symptom cluster) are represented as ovals, and observed symptom indicators are shown as rectangles with standardized factor loadings. Residual variances are indicated by e1 to e11. The double-headed arrow between the two latent variables represents the estimated correlation between clusters (ρ = 0.545). Model fit indices are presented within the figure, χ²/df = 3.480, GFI = 0.896, AGFI = 0.840, CFI = 0.884, TLI = 0.852, RMSEA = 0.107 values, which together suggest that although some convergence toward the hypothesized model is observed, the overall fit remains limited, and further validation is warranted.