Understanding quality of life for children after liver transplantation: A work in progress

Abstract

Liver transplantation is standard-of-care treatment for children with end-stage liver disease caused by biliary atresia, fulminant liver failure, and many metabolic liver diseases.1 Current 1-year patient survival rates are over 95% and 5-year rates are greater than 85% at experienced centers. It is increasingly clear that the success of liver transplantation in children is defined by more than just excellent patient or graft survival rates. Although life-saving, liver transplantation is not curative: a fatal disease has been replaced by a chronic condition with its own associated morbidities, which may be present despite well-preserved allograft function.2 The ensuing posttransplant course often creates new stressors for both the pediatric recipient and his or her parents. These stressors include lingering fears, both medical and psychological, as well as uncertainties regarding the potential immediate and long-term medical complications that may impact their lives. This consideration highlights the fact that the outcomes of liver transplantation in children must consider not only the quantity but also the quality of life years survived and restored.3 CHQ, Child Health Questionnaire; HRQOL, health-related quality of life; ITQOL, Infant Toddler Quality of Life Instrument; SPLIT, Studies of Pediatric Liver Transplantation. Current understanding of health-related quality of life (HRQOL) after pediatric liver transplantation is largely derived from single-center experiences. A recent systematic review of 11 published single-center studies assessing HRQOL in children and adolescents after liver transplantation suggests that pediatric liver transplantation recipients have poorer quality of life in areas of physical, psychological, social, and family functioning when compared to healthy peers.4 However, there is equal or better HRQOL than that seen in available comparison groups of children and adolescents with other chronic illnesses, such as pediatric juvenile arthritis, asthma, and epilepsy.5-7 Recommendations from both this review4 and the accompanying editorial3 were that, moving forward, investigators should use only measures that are well validated for a broad pediatric age range, are available in multiple languages, and include a self-report format for older children.3, 4 Indeed, multicenter collaborations will help overcome some of these hurdles, given the relative infrequency of liver transplantation in children even at the largest of pediatric liver transplant centers, by facilitating evaluations of more heterogeneous populations of survivors of pediatric liver transplantation and encouraging the utilization of similar outcome measures of HRQOL to allow comparison across centers of the overall impact of pediatric liver transplantation. Since its inception in 1996, the primary objectives of the Studies of Pediatric Liver Transplantation (SPLIT) database have been to characterize and follow trends in patient and graft survival, rejection incidence, growth parameters, and immunosuppressive therapy and to identify potential risk factors for patient mortality and morbidity and graft failure among children undergoing liver transplantation.1 SPLIT provides a prospective collection of comprehensive pretransplant and posttransplant medical data from pediatric liver transplant candidates and recipients from multiple pediatric liver transplant programs. This comprehensive and multicenter approach helps to overcome the single-center bias concerns of earlier work. Currently, SPLIT has registered and prospectively follows at least 80% of all patients under the age of 18 years who receive a liver transplant in Canada and the United States8 from a current complement of 45 pediatric liver transplant programs in North America. As a result, the SPLIT database registry is uniquely positioned to facilitate ancillary studies aimed at optimizing the outcomes of survivors of pediatric liver transplantation. In this issue of Liver Transplantation, Estella Alonso and a subgroup of SPLIT colleagues9 conducted a 5-center cross-sectional analysis of HRQOL and family function dynamics in a homogeneous population of children at 20 to 28 months (herein referred to as 2-year survivors) after liver transplantation. Two generic HRQOL measures were used: the Infant Toddler Quality of Life Instrument (ITQOL)10 was completed by 67 parents of “younger” 2-year survivors (that is, study participants less than 5 years of age at the time of this study), whereas Child Health Questionnaire (CHQ) Parent Form 5011 was completed by 35 parents of “older” 5-year survivors (study participants greater than 5 years of age at the time of assessment). A third questionnaire, the Family Assessment Device,12 which assesses family function via examination of 6 dimensions in the McMaster Model of Family Functioning (problem solving, communication, roles, behavior control, affective responsiveness, and involvement), was completed by all adult family members living in the same household as the study subjects. Several observations are noteworthy from this multicenter study. First, the younger survivors of pediatric liver transplantation had lower scores in the domains of global health, general health perceptions, and change in health in comparison with the older transplant recipients, who had lower scores in the domains of physical health and general health. Median scores from the parent proxy report suggest decreased issues with discomfort and pain in the younger survivors than a control group created for this study. Parents of the older survivors reported increased emotional stress and disruption of family activities. Mean Family Assessment Device scores indicated that families of a liver transplant recipient do not have increased levels of family dysfunction. The authors are to be congratulated for directly addressing some of the limitations inherent in previous single-center studies of HRQOL. Their collaboration with one of the developers of both the ITQOL and the CHQ reflect their attention to using well-validated measures to improve the generalizability of the results. The CHQ is a well-established research tool that has robust normative data with which to make comparisons across disease cohorts.11 The ability to make these comparisons is especially important when describing the disability associated with a given condition or when advocating for further allocation of limited health-care resources. The ITQOL, being a more recently developed tool,10 lacks a similar level of normative data. This required the authors to develop comparative “control” HRQOL data for those under 5 years of age, using patients drawn from general pediatric practices. Although lack of normative data is a limitation, at the time the study was conducted, no ideal alternative tool was available to conduct a generic HRQOL assessment in children under 5 years of age. However, there are now several options of generic tools currently in use to allow comparisons between healthy populations and the pediatric liver transplant population.13, 14 A limitation of all generic tools is their potential insensitivity to important clinical changes due to either the inclusion of items that are not relevant to the individual patient with a specific disease or the omission of key HRQOL items related to a specific disease.15 For these reasons, a disease-specific HRQOL instrument for survivors of pediatric liver transplantation might be useful to address these deficiencies. To date, however, there is no published disease-specific HRQOL tool for use in pediatric liver transplantation patients. Nonetheless, work is ongoing toward this goal, so that, in the near future, such a tool is likely to become available.16-18 Despite the participation of 5 centers, the ability to recruit a sufficient sample size remains the biggest limitation of this study. Although 102 is not an insignificant number of participants, the study cohort was stratified into the two age groups (<5 years of age and ≥5 years of age), primarily because no single generic HRQOL tool adequately addresses quality-of-life issues across the pediatric age spectrum. This stratification brought numbers down to 35 and 67 in the two age groups. As a result, the authors recognize that there was insufficient power to conduct meaningful analyses of potential medical or demographic risk factors that might influence the HRQOL outcomes. For similar reasons, the failure to demonstrate factors that significantly impacted family function may also have been due to a type II statistical error. In this study, we hear the voice of the parents as they describe their perception of their children's HRQOL. In pediatric HRQOL assessments, the use of a parent proxy is often necessary because of the age, developmental level, or underlying medical condition of the child. However, given the age of some of the respondents in this study, self-report would have been possible with validated tools. Is the lack of self-report HRQOL data a limitation? Not completely, but it is important to appreciate what information we may fail to capture adequately when we depend solely on parent proxy reporting. Of concern with proxy-reported measures is the low to modest agreement with the child's self-reported quality-of-life assessment, particularly for emotional and social domains of the HRQOL index.19 Nonetheless, as acknowledged by Alonso et al., the liver transplantation experience impacts the entire family, and the lack of self-report in this study lays the groundwork for future studies. Finally, the outstanding challenge remains how to make HRQOL outcomes relevant to the health-care providers for children post liver transplant. Researchers must endeavor to translate HRQOL scores into meaningful terms so they can be understood and acted upon by the practitioner. For example, understanding what items make up a “social functioning” domain is important, if this is a domain scoring low in a poor HRQOL for an individual patient, so that an intervention could be targeted to address the specific areas of concern. Successful liver transplantation in a child is often a hard-won victory, requiring all the combined expertise of a dedicated multidisciplinary team. Assessment of HRQOL permits us to hear the voice and perspective of the patient, which can predict, often better than physiologic measures, future health status, mortality, and resource utilization.20, 21 Given the varying perceptions and expectations between individuals, two people with the same health status may have very different responses in domains of HRQOL. Factors contributing to such discrepancies in HRQOL between patients receiving the same treatment can then be targeted for interventions. Given the limited numbers of liver transplant patients in any one pediatric program, broad collaborative efforts are now required. The SPLIT collaboration affords such an effective collaboration. We thank Professors Anne-Marie Griffiths and Philip Sherman for their thoughtful reviews of this editorial.