- Top of page
- Patients And Methods
During the past decades, medical technology has improved dramatically and many otherwise fatal diseases have become chronic. Consequently, increasing attention has been paid to health-related quality of life (HRQoL) to complement clinical outcomes. With the possibility of liver transplantation and the increased success of medication treatment for many liver diseases, this has also been the case in hepatology. HRQoL is often defined as the impact of disease and/or treatment on a patient's physical, emotional and social functioning.1 Numerous studies have shown that HRQoL is impaired in patients with chronic liver disease (CLD), and within this group, patients with chronic hepatitis C experience the lowest HRQoL.2–18 Despite the many studies that have shown a reduced HRQoL in hepatology, relatively few studies have investigated what factors influence liver patients’ HRQoL. Disease severity, as indicated by stage of fibrosis (absent, early or advanced) or Child-Pugh scores, seems to determine HRQoL.4, 5, 15 Such a relationship between disease severity and HRQoL seems fairly self-evident. Nevertheless one study did not find this relationship.19 Marchesini et al. found itch and muscle cramps to be of major concern in patients with cirrhosis.20 Besides these mainly physical aspects of the illness, the predictive value of psychosocial aspects on HRQoL has also received some attention. Decreased energy and emotional reactions were found to be related to HRQoL in patients with primary biliary cirrhosis,17 although it remains unclear exactly which emotional reactions the authors refer to. Depression, anxiety and illness understanding were all related to HRQoL in patients with hepatitis C and liver patients with various disease aetiologies.3, 14, 17, 19, 21 Depression, anxiety and illness understanding are typically generic features of chronic disease. Other psychosocial factors, which are more specific to suffering from a liver disease, may influence HRQoL as well. These have not been studied previously.
Indeed, a recent study on the development of a disease-specific HRQoL questionnaire in hepatology emphasized that there are many more physical and psychological factors important in determining HRQoL in CLD patients.1 In that study the Liver Disease Symptom Index 2.0 (LDSI 2.0) was developed based on the results from prior studies and interviews with CLD patients about liver disease-specific symptoms and health-related disabilities. In conjunction with important predictors such as depression and anxiety, domains such as itch, joint pain, fatigue, pain in the right upper abdomen, memory problems, change of personality, money problems and problems in sexual functioning were found to be of particular importance to CLD patients. The development of the LDSI 2.0 offers the opportunity to determine which specific symptoms for liver disease may influence HRQoL. Therefore, the current study proposed to assess the predictive value of these patient-based items in determining HRQoL. These factors were studied in a large sample of patients who presented with a spectrum of diseases, symptoms and signs, which is broader than that of an in-hospital patient population only.
Besides the dearth of research on factors influencing HRQoL in chronic liver patients, another shortcoming of HRQoL research in CLD patients is that all studies have operationalized HRQoL as a multidimensional, unweighted outcome. Such multidimensional unweighted outcomes make it impossible to compare a burden in different dimensions of HRQoL. For instance, with unweighted scores on different dimensions, it is impossible to ascertain whether a particular decrease in mobility is worse or less of a problem than a particular increase in pain. Typically, these articles present the observations for all HRQoL dimensions measured, which results in a presentation of results which is difficult to interpret. Moreover, the conclusions drawn from such multiple outcomes become contestable, as there is no way of telling if one result is more clinically relevant than the other. This means that it is yet undetermined which variables have the highest impact on the overall HRQoL in hepatology. In this study we use the SF-6D HRQoL questionnaire, which allows for a weighted overall score of HRQoL. The present study will be the first to make use of such ‘utility scores’ of HRQoL in relation to physical and psychosocial predictors.
- Top of page
- Patients And Methods
This is the first study in which a weighted overall score of HRQoL was related to different physical and psychosocial issues of CLD patients. Regression analyses showed that HRQoL of patients with CLD (excluding hepatitis C) was most strongly determined by joint pain and disease severity. An increase of one point on the joint pain scale or the disease severity scale, predicted a 0.023 and 0.029 point decrease in HRQoL respectively. Depression, pain in the right upper abdomen, fatigue, decreased appetite, memory problems, change of personality and daily time management also predicted HRQoL significantly in the overall group of chronic liver patients, but to a smaller extent. Regarding HCV patients, HRQoL was most strongly related to disease severity and depression. Use of interferon, fatigue, joint pain and hindrance in financial affairs were also significantly related to HRQoL in patients with HCV. Note that because of the high number of patients involved in this study, we were able to show that even less obvious items may have a relationship with quality of life. Indeed, many of the extra items of the LDSI 2.0 added on the basis of suggestions of the NLV have a statistically significant relationship with quality of life.
When interpreting the results, it is important to realize that we are looking at associations between symptoms and HRQoL in a population receiving best clinical care. The remaining variance in both symptoms and HRQoL is the variance for which this clinical practice could not control. This explains why, for instance, a variation in disease severity has only a limited influence on HRQoL: most of that variance is controlled for by an apparently successful treatment. What is left of the variance in symptoms and HRQoL and the relation between them suggests room for clinical improvements.
Comparing the results of the present study with those presented in the literature, several considerations may be derived. Some findings were in accordance with the findings from previous studies, such as for the relationship of depression and fatigue with HRQoL.14, 15, 19 However, contradictory to other studies, no relationship was found between HRQoL and disease-related worries21 or itch.20 With regard to ‘worry’, it is possible that the difference in measurement instruments accounts for this discrepancy. In addition, the ‘worry’ items could be different. Unfortunately it is not clear from Hauser's study which particular items investigated this dimension.21 Regarding ‘itch’, the difference in findings could be attributed to a sampling issue; the current study contained relatively few patients who experienced itch (2.7%), whereas Marchesini et al. investigated patients with cirrhosis who were therefore more prone to experiencing itch.20
Most of the factors that were assessed in this study were neither supported nor contradicted by previous research simply because they have not previously been investigated. Of these factors, some significantly predicted HRQoL, namely: ‘joint pain’, ‘pain in the right upper abdomen’, ‘decreased appetite’, ‘memory problems’, ‘change of personality’ and ‘daily time management’. Factors that failed to determine HRQoL significantly were ‘jaundice’, ‘hindrance in financial affairs’, ‘decreased sexual activity’ and ‘decreased sexual interest’. As explained above this does not mean that, for example, jaundice could not have a influence on HRQoL, but in this patient population no such associations were found, quite possibly because the treatment has already reduced the problems caused by the disease.
A few limitations must be considered. First, this study was conducted in a group of liver patients from the NLV. We concede that becoming a member of a patient association is an action which could possibly induce a selection bias. Indeed, compared with a Dutch in-hospital population which was used in a validation study of the LDSI 2.0, this population differed significantly with respect to gender (more women), disease stage (less severe) and disease aetiology (less viral hepatitis, more liver transplant patients).22 However, our aim was to include patients with a spectrum of disease and symptoms and signs which was broader than an in-hospital patient population only.
A second possible limitation of this study which is inherent in the selection of the patient population is that respondents reported the clinical characteristics, disease stage and aetiology of their disease themselves. However, a prior pilot study at the outpatient clinic demonstrated that liver patients are very much aware of the clinical symptoms they have or have had, and the type of liver disease they suffer from.15 Therefore, we are confident that this study provided reliable insight into the HRQoL of chronic liver patients.
A third limitation to this study is that the study population included patients who were transplanted or cured from their liver disease. Considering their limited number (n = 71), these patients were included in this study all the same. To control for any bias these patients could cause, additional analyses were conducted without these two groups of patients. No significant differences were found between analyses that included all CLD patients and analyses that excluded transplanted and cured patients, except for the subscale ‘change of personality (due to liver disease)’, which did not relate statistically significantly to HRQoL when ‘cured’ and transplanted patients were excluded. This could be explained by the exclusion of transplanted patients who were rejoiced by the life-saving treatment and the new life opportunities given.
A fourth limitation is that the physical and psychosocial factors used in this study to predict HRQoL consisted of only one item per dimension, so when it is stated in this study that depression predicts HRQoL significantly, it should be kept in mind that depression was not measured with a questionnaire specifically validated for the purpose of measuring depression. Nonetheless, a question asking to what extent they have felt down during the past 4 weeks gives a good indication, certainly for research which is based on data from a large sample of patients. Indeed, as emphasized earlier, the results concerning the relationship between depression and HRQoL in CLD patients are in accordance with previous research.14, 19 It should be noted that the reliability and validity of the LDSI 2.0, from which the items are derived, are good.22
With the results of the present study, several suggestions for application in clinical practice can be made. In conjunction with slowing down the progression of the disease, treatment should focus on those aspects that are possibly modifiable and of significant influence on HRQoL, beyond the present treatment level. Strikingly, the factors that were found to determine HRQoL were in fact reasonably modifiable; joint pain, pain in the right upper abdomen, depression, decreased appetite, fatigue, daily time management, memory problems and change of personality. Providing information on presence or absence of these symptoms to the doctor by administering a short checklist to patients right before each visit may prove useful. Consequently, treating the symptoms might obviously be a way of enhancing quality of life. For joint pain, institution of (drug) therapy may prove beneficial. Regarding decreased appetite, dietary advice to prevent malnutrition can be considered. For clinically depressed patients, some form of drug therapy or psychotherapy may be beneficial. An alternative, less obvious, intervention might consist of interventions that improve coping with the symptoms associated with CLD such as memory problems, pain in the right upper abdomen, change in daily time management, fatigue and change of personality. These are symptoms that are not instantly visible but can have a huge impact on daily functioning. As these cannot easily be treated, patients have to adapt to these problems. As a treating doctor, one of the approaches might be to offer participation in a cognitive behavioural programme. Cognitive behavioural treatment is based on the assumption that inappropriate thoughts and behaviours adversely affect well-being. Therefore, teaching patients to adjust their cognitions and behaviours will positively influence quality of life. Interventions based on cognitive behavioural therapy have already been proven to be effective in patients with other chronic diseases,26–29 and may therefore be of most benefit to this patient group.