The Relationships Among Symptom Clusters, Perceived Self-Efficacy And Quality Of Life: A Model Estimation In Patients With Primary Liver Cancer

Objectives: The objectives of this study were to develop a symptom module specific toprimary liver caner (PLC) through comparing with The M. D. Anderson SymptomInventory-PartⅠ(MDASI-PartⅠ), which would be used to identify symptom clusters inPLC patients; to test a hypothesis that perceived self-efficacy (PSE) mediates betweensymptom clusters and quality of life(QOL), to explore self-management behaviours andtheir influencing factors apart from PSE, and finally to propose a relationship model ofsymptom clusters, PSE and QOL in PLC patients.Methods: A three-stage design was used in this study. StageⅠ: The content validity of thesymptom module specific to primary liver caner (TSM-PLC) was developed throughliterature and expert reviews, and other psychological properties of TSM-PLC were testedin a convenient sample of196PLC patients; moreover, a model of symptom clusters inPLC patients was proposed by exploratory factor analysis. StageⅡ: A convenient sampleof388PLC patients from a hepatobiliary hospital in Shanghai were questionnaired byMDASI-PartⅠ, TSM-PLC, Stratigies Used by Patients to Promote Health, FunctionalAssessment of Cancer Therapy-Heptobiliary Carcinoma and Hospital Anxiety andDepression Scale; accoding to the collected data, the proposed model regarding symptomclusters was verified and revised by comfirmatory factor analysis, the influencing fators ofsymptom clusters and PSE were identified by multiple stepwise regression model, and theproposed hypothesis was tested by structure equation model. Stage Ⅲ: Eight patients withPLC from the above hospital were interviewed by semi-structure interviews.Results:①TSM-PLC was consisted of6symptom items including abdominal distention,diarrhea, jaundice, pruritus, weight loss, and fever, all of which constituted a factor basedon the criteria of Eigenvalues greater than1.00; its content validity and internal reliabilitywere0.911and0.835respectively; the criterion validity related to MDASI-PartⅡwas0.403(P＜0.001), and it was sensitive to performance status, liver function status, cancerstage and the number of current treatments(P＜0.01).②Three symptom clusters wereidentified as gastrointestinal sickness, liver dysfunction and neuropsychological symptomclusters. Performance status, anxiety state, current transcatheter arterialchemoembolization, cancer stage and the number of current treatments were the factorsinfluencing the gastrointestinal sickness symptom cluster; performance status, anxiety state,cancer stage, depression state and employment status were the factors influencing the liverdysfuntion symptom cluster, and performance status, anxiety state, previous other treatment and the number of current treatments were the factors influencing theneuropsychological symptom cluster.③The mean item score of the SUPPH in this samplewas3.15, indicating a moderate lever of PSE, however, the affective domain of PSE was abit better than the performance domain of PSE. Anxiety and depression states were thesame factors of the above two domains of PSE; Previous treatment for protecting liverfunction and previous traditional Chinese medicine were the factors related to the affectivedomain of PSE while liver function status and EOOG were the factors related to theperformance domain of PSE.④The mean item score of FACT-Hep was2.72, indicating amoderate lever of QOL. In respect to the five domains of FACT-Hep, the mean item scoreswere3.19,2.81,2.74,2.40,2.35respectively for social and family well-being,hepatobiliary module, emotional well-being, functional well-being and physical well-being.⑤The QOL were directly predicted by the symptom clusters (β=-0.537, P＜0.001) and thePSE (β=0.472, P＜0.001), and noticeably, the PSE mediated between the symptom clustersand the QOL (β=-0.146, P＜0.001).⑥Two main kinds of self-management behaviorswere identified as explicit and implicit behaviors. The former included treatment-relatedbehaviors, healthy life styles, seeking support and relieving stress, while the latter includedaffective regulation and cognitive processing. Furthermore, apart from PSE, individual,behavior-related, and environmental factors were found to influence the behaviors.Conclusions:①TSM-PLC showes a good reliability, validity and sensitivity, which canbe applied together with MDASI-PartⅠto assess symptoms in patients with PLC.②Thereare three symptom clusters in patients with PLC, and their influencing factors refer tophysiological, psychological and environmental aspects, both of which have verified theUnpleasant Symptom Theory.③Improving PSE can partly palliated the negative effect ofsymptom clusters on QOL in patients with PLC, and the affective domain of PSE couldtotally mediates between symptom clusters and the social well-being domain of QOL.④Arelationship model of symptom clusters, PSE and QOL has been proposed based on thefindings of the quantitative and qualitative studies. This model will probably play animportant part in studying interventions of symptom clusters, advancing stratigies ofimproving PSE, and implementing self-management programs for patients with PLC, all ofwhich will then constitute to better QOL in patients with PLC.