Katherine Stevens

Developing a descriptive system for a new preference-based measure of health-related quality of life for children.

ObjectivesThe use of preference-based measures (PBMs) of health-related quality of life (HRQoL) is increasing in health care resource allocation decisions. Whilst there are measures widely used for this purpose in adults, research in the paediatric field is more limited. This paper reports on how the descriptive system for a new paediatric generic PBM of HRQoL was developed from dimensions identified in previous research.MethodsExisting scales from the paediatric literature were reviewed for suitability, and scales were also developed empirically, based on qualitative interview data from children, by taking adverbial phrases and confirming the ordinality by a ranking exercise with children. The resulting scales were applied to the dimensions from the previous research.ResultsNo suitable scales were found in the paediatric literature, so the empirically derived scales were used resulting in seven different types. Children were successfully able to rank these to determine the ordinality, and these types were applied to the dimensions.ConclusionsThis work has empirically developed a descriptive system for the dimensions of HRQoL identified in previous research. Further research is needed to test the descriptive system on a paediatric population and reduce the number of dimensions to be amenable to health state valuation.

Valuation of the Child Health Utility 9D Index

Background and ObjectivesThe aim of this study was to test the feasibility of estimating preference weights for all health states defined by the Child Health Utility 9D (CHU9D), a new generic measure of health-related quality of life for children aged 7–11 years. The estimation of preference weights will allow the calculation of QALYs for use in paediatric economic evaluation.MethodsValuation interviews were undertaken with 300 members of the UK adult general population to obtain preference weights for a sample of the health states in the CHU9D descriptive system. Both standard gamble and ranking valuation methods were used. Regression modelling was undertaken to estimate models that could predict a value for every health state defined by the system. A range of models were tested and were evaluated based on their predictive performance.ResultsModels estimated on the standard gamble data performed better than the rank model. All models had a few inconsistencies or insignificant levels and so further modelling was done to estimate a parsimonious consistent regression model using the general-to-specific approach, by combining inconsistent levels and removing non-significant levels. The final preferred model was an ordinary least squares (OLS) model. All the coefficients in this model were significant, there were no inconsistencies and the model had the best predictive performance and a low mean absolute error.ConclusionThis research has demonstrated it is feasible to value the CHU9D descriptive system, and preference weights for each health state can be generated to allow the calculation of QALYs. The CHU9D can now be used in the economic evaluation of paediatric healthcare interventions. Further research is needed to investigate the impact of children’s preferences for the health states and what methods could be used to obtain these preferences.

Valuing Child Health Utility 9D health states with a young adolescent sample: a feasibility study to compare best-worst scaling discrete-choice experiment, standard gamble and time trade-off methods.

QALYs are increasingly being utilized as a health outcome measure to calculate the benefits of new treatments and interventions within cost-utility analyses for economic evaluation. Cost-utility analyses of adolescent-specific treatment programmes are scant in comparison with those reported upon for adults and tend to incorporate the views of clinicians or adults as the main source of preferences. However, it is not clear that the views of adults are in accordance with those of adolescents on this issue. Hence, the treatments and interventions most highly valued by adults may not correspond with those most highly valued by adolescents. Ordinal methods for health state valuation may be more easily understood and interpreted by young adolescent samples than conventional approaches. The availability of young adolescent-specific health state values for the estimation of QALYs will provide new insights into the types of treatment programmes and health services that are most highly valued by young adolescents.The first objective of this study was to assess the feasibility of applying best-worst scaling (BWS) discrete-choice experiment (DCE) methods in a young adolescent sample to value health states defined by the Child Health Utility 9D (CHU9D) instrument, a new generic preference-based measure of health-related quality of life developed specifically for application in young people. The second objective was to compare BWS DCE questions (where respondents are asked to indicate the best and worst attribute for each of a number of health states, presented one at a time) with conventional time trade-off (TTO) and standard gamble (SG) questions in terms of ease of understanding and completeness.A feasibility study sample of consenting young adolescent school children (n = 16) aged 11–13 years participated in a face-to-face interview in which they were asked to indicate the best and worst attribute levels from a series of health states defined by the CHU9D, presented one at a time. Participants were also randomly allocated to receive additional conventional TTO or SG questions and prompted to indicate how difficult they found them to complete.The results indicate that participants were able to readily choose ‘best’ and ‘worst’ dimension levels in each of the CHU9D health states presented to them and provide justification for their choices. Furthermore, when presented with TTO or SG questions and prompted to make comparisons, participants found the BWS DCE task easier to understand and complete.The results of this feasibility study suggest that BWS DCE methods are potentially more readily understood and interpretable by vulnerable populations (e.g. young adolescents). These findings lend support to the potential application of BWS DCE methods to undertake large-scale health state valuation studies directly with young adolescent population samples.

Assessing the performance of a new generic measure of health-related quality of life for children and refining it for use in health state valuation

BackgroundPrevious research to develop a new generic paediatric health-related quality of life (HR-QOL) measure generated 11 dimensions of HR-QOL, covering physical, emotional and social functioning. These dimensions and their response scales were developed from interviews with children. Some of these dimensions have alternative wording choices. The measure is intended to be preference based so that it can be used in paediatric economic evaluation.ObjectivesThe aims of this research were to assess the performance of this new descriptive system in a general and clinical paediatric population, to determine the most appropriate wording for the dimensions and to refine the descriptive system to be amenable to health state valuation to make it suitable for use in economic evaluation.MethodsA sample of 247 children was recruited from general and clinical paediatric populations. Each child completed the descriptive system and data were collected to allow assessment of practicality (including response rates, completion rates and time to complete), content, face and construct validity, whether the child could self-complete and preferences for alternative wordings that could be used for dimensions. These data were used to inform the final choice of wording for dimensions, the scales used for each dimension and the reduction of dimensions to meet the constraints of health state valuation.ResultsThe descriptive system demonstrated good practicality and validity in both the general and clinical paediatric samples. The completion rates were excellent (>98%), the mean time to complete was low (3.8 minutes for the general and 5.3 minutes for the clinical sample) and there was evidence of face, content and construct validity. The descriptive system was able to demonstrate significant differences between the general and clinical samples and according to the level of health of children. 96% of the school sample and 85% of the clinical sample were able to self-complete. The final choice of wording for the 11 dimensions was determined by the preferences and comments of the children. To make it amenable for health state valuation, the number of dimensions was reduced from 11 to 9 by removing the dimensions ‘jealous’ and ‘embarrassed’.ConclusionsThe descriptive system performed well in both the general and the clinical populations, and the final descriptive system generates health states that are feasible for health state valuation. Further research is needed to value the final descriptive system by obtaining preference weights for each health state, thereby making the measure suitable for use in paediatric economic evaluation.

Outcome at 6 Months After Admission for Pediatric Intensive Care: A Report of a National Study of Pediatric Intensive Care Units in the United Kingdom

OBJECTIVE. The goal was to measure, by using the Health Utilities Index, the health status of children 6 months after admission to PICUs in the United Kingdom. METHODS. All PICUs in the United Kingdom were invited to participate. Children who were ≥6 months of age at admission and were discharged alive from participating units during a 1-year period were eligible for this study. Children with completed consent forms who had survived to 6 months after admission received the Health Utilities Index questionnaire. RESULTS. Admission data on 7214 admissions to 22 units were collected between March 2001 and February 2002. Of those patients, 6786 survived to unit discharge, and consent for follow-up contact was obtained for 2642. At 6 months after admission, the mortality rate was estimated at 11.1% and 2034 children were still alive and contactable. Of those children, Health Utilities Index questionnaires were returned by 1455. No impairment was indicated for 767, 951, 940, 919, 962, and 939 children with respect to the Health Utilities Index sensation, cognition, emotion, pain, mobility, and self-care attributes, respectively. The mean ± SE Health Utilities Index utility score was 0.73 ± 0.01, with 397 children (27.3%) in full health. CONCLUSIONS. Death after pediatric intensive care in the United Kingdom is uncommon, making assessment of health status important. At 6 months after admission, there is significant morbidity. To assess the impact of strategies to improve or to optimize care on longer-term outcomes, standardized collection of data on preexisting comorbidities and illness severity is required.

Developing adolescent-specific health state values for economic evaluation: an application of profile case best-worst scaling to the Child Health Utility 9D

BackgroundThe way that health is measured and valued is fundamental to economic evaluation. To date, adult health state values have been routinely used in the calculation of QALYs for the economic evaluation of healthcare treatment and preventive programmes, including those targeted at adolescents.ObjectivesThe main objective of this study was to apply profile case best-worst scaling (BWS) discrete-choice experiment (DCE) methods to obtain adolescent-specific values for the Child Health Utility 9D (CHU9D), a new generic preference-based measure of health-related quality of life developed specifically for application in cost-effectiveness analyses of treatments and interventions targeted at young people. A secondary aim was to assess the feasibility of a web-based method of data collection for the valuation of health states defined by the CHU9D.MethodsA web-based survey was developed including the CHU9D instrument and a series of BWS DCE questions. Specifically, respondents were asked to indicate the best and worst attribute levels from a series of ten health states defined by the CHU9D, presented one at a time. The survey was administered to a community-based sample of consenting adolescents (n=590) aged 11–17 years. A conditional logistic regression model was applied to estimate values (part-worth utilities) for each level of the nine attributes relating to the CHU9D. A marginal utility matrix was then estimated to generate an adolescent-specific scoring algorithm on the full health=1 and dead =0 scale required for the calculation of QALYs.ResultsThe results indicate that participants were able to readily choose ‘best’ and ‘worst’ attribute levels for the CHU9D health states. Large differences in value were found between the first and fifth levels (indicating ‘no problems’ and ‘severe problems’, respectively) for all nine attributes relating to the CHU9D. In general, there was little differentiation between the middle levels of all attributes indicating only limited additional value for adolescents of moving between these levels. Comparison of the adolescent-specific algorithm and the existing adult scoring algorithm for the CHU9D revealed some significant differences in values for identical health states, which may have important implications for the application of the CHU9D to value adolescent treatment and service programmes particularly for mental health. In general, adolescents appeared to place more weight upon the CHU9D attributes relating to mental health (worried, sad and annoyed) than would be implied by application of the existing algorithm based upon adult values.ConclusionThis study provides preliminary indications that there may be potentially important and systematic differences in the valuations attached to identical health states by adolescents in comparison with adult population groups. The study findings lend support to the potential future application of profile case BWS DCE methods to undertake large-scale health state valuation studies directly with young adolescent population samples and provide support for the feasibility and acceptability of a web-based mode of administration for this purpose.

The development of a preference-based measure of health in children with atopic dermatitis

Background  Outcomes of health care interventions can be measured in many different ways, but there has been growing interest in the role that quality‐adjusted life‐years (QALYs) can play in informing priorities in health care. While existing generic preference‐based measures can be used to obtain QALYs, these measures may often be inappropriate for some conditions such as atopic dermatitis (AD). These measures are also usually developed for adults and may not capture the domains pertinent to children.

Use of visual analog scales in economic evaluation

Visual analog scales or rating scales are commonly used in economic evaluation to elicit preferences in order to estimate quality-adjusted life years. Values obtained from visual analog scales have been used on their own or via a transformation to map them onto one of the choice-based methods for elicting preferences, namely standard gamble or time trade-off. The arguments against using visual analog scales in economic evaluation directly or indirectly via a transformation are reviewed. It is concluded that it canonly ever provide a second best solution compared with the direct use of a choice-based technique.

The Use of Qualitative Methods in Developing the Descriptive Systems of Preference-Based Measures of Health-Related Quality of Life for Use in Economic Evaluation

OBJECTIVES To describe how qualitative methods can be used in the development of descriptive systems of preference-based measures (PBMs) of health-related quality of life. METHODS The requirements of the National Institute for Health and Clinical Excellence and other agencies together with the increasing use of patient-reported outcome measures has led to an increase in the demand for PBMs. Recently, interest has grown in developing new PBMs and while previous research on PBMs has mainly focused on the methods of valuation, research into the methods of developing descriptive systems is an emerging field. RESULTS Traditionally, descriptive systems of PBMs were developed by using top-down methods, where content was derived from existing measures, the literature, or health surveys. A contrasting approach is a bottom-up methodology, which takes the views of patients or laypeople on how their life is affected by their health. This approach generally requires the use of qualitative methods. Qualitative methods lend themselves well to the development of PBMs. They also ensure that the measure has appropriate language, content validity, and responsiveness to change. While the use of qualitative methods in the development of non-PBMs is fairly standard, their use in developing PBMs was until recently nonexistent. CONCLUSIONS In this article, we illustrate the use of qualitative methods by presenting two case studies of recently developed PBMs, one generic and one condition specific. We outline the stages involved, discuss the strengths and weaknesses of the approach, and compare with the top-down approach used in the majority of PBMs to date.

The incremental cost effectiveness of withdrawing pulmonary artery catheters from routine use in critical care

ObjectiveThe objective of this study was to conduct an economic evaluation to identify any differences in the expected costs and outcomes between patients treated with pulmonary artery catheters (PACs) and those without, in order to better inform healthcare decision makers.MethodThe evaluation was carried out alongside a clinical trial investigating the use of PACs in intensive care units (ICUs) in the UK. It was conducted from the perspective of the UK NHS, in which PACs are an established intervention. Treating patients without using a PAC was characterised as the new intervention.The primary outcome measure was QALYs. The secondary outcome measure was hospital mortality. NHS costs per patient were calculated for the financial year 2002/03.The bootstrap method was used to characterise the uncertainty of the results and to construct cost-effectiveness acceptability curves.ResultsThe cost per QALY and per life gained from the withdrawal of PACs were £2892 and £21 164, respectively.ConclusionThe results of this study indicate that withdrawal of PACs from routine clinical use in ICUs within the NHS would be considered cost effective in the current decision-making climate.