The Journal of Pediatrics
Volume 155, Issue 1 , Pages 7-9, July 2009

Current Challenges and Future Research in Measuring Preferences for Pediatric Health Outcomes

  • Lisa A. Prosser, PhD

      Affiliations

    • Corresponding Author InformationReprint requests: Lisa A. Prosser, PhD, Child Health Evaluation and Research Unit, Division of General Pediatrics, University of Michigan Health System, 300 North Ingalls, Bldg 6E14, Ann Arbor, MI, 48109.

Child Health Evaluation and Research Unit, Division of General Pediatrics, University of Michigan Health System, Ann Arbor, Michigan

Article Outline

 

See related article, p 21

The methodological and practical challenges of measuring preferences for health outcomes in children are receiving increased attention.1, 2, 3 Methodological challenges include the limitations of applying standardized methods developed for adults to children and the dependence of children on others for care and decision making.1, 2, 4 Practical challenges include the inability of young children to value changes in health directly and the potential biases of using proxy respondents. The study by Carroll and Downs5 in this issue of The Journal is a welcome addition to the sparse literature describing preferences for pediatric health states. Their study presents parent-derived health utility weights for 29 health conditions in children and represents a comprehensive list of common pediatric illnesses.

Back to Article Outline

Lack of Standardized Methods for Valuing Child Health for Cost-Effectiveness Analyses 

Preference-based measures, such as quality-adjusted life-years derived by using health utility weights, are recommended for inclusion in cost-effectiveness analysis.6 Substantial variation in elicitation methods and variation in the sources of preferences used to calculate quality-adjusted life-years for children in cost-effectiveness studies has been identified in several reviews.3, 7, 8 This variation may be attributable to the challenges of valuing childhood outcomes within the cost-effectiveness framework, and variations from standard methods may be the result of attempts to overcome the limitations of using methods developed for adults.7, 9 The greatest variation in methods is associated with health states in infants and very young children, likely because of the particular challenges of applying existing valuation methods to this age group.

Back to Article Outline

Limitations of Currently Available Approaches for Valuing Child Health 

There are 2 main categories of approaches used to identify preference-based health utility weights: health status instruments and direct valuation. Health status instruments, such as the EQ-5D and Health Utilities Index, can be used to value health states with a specified set of attributes. The attribute levels that correspond to a particular health state are then used in a mathematical model to calculate utility weights on the basis of previously collected community preferences.10 Direct valuation approaches include the use of time trade-off or standard-gamble questions to elicit health utility weights for either hypothetical or experienced health states.

Currently available health status instruments vary in their applicability to children. Most were not developed to capture health domains specific to children, although some existing instruments considered older children and adolescents during development11 and adaptations for children are underway.12 None of the currently available instruments has been validated for use in children younger than 5 years. Another limitation of existing instruments is the assumption that health utility weights do not vary by age.

Direct valuation approaches, although more resource-intensive, are likely to be better-suited for valuing child health because these techniques allow the investigator to tailor the description of the health state to include relevant health domains and child age. A key challenge of direct valuation approaches is potential bias associated with the use of parent proxies. Carroll and Downs used a direct valuation approach with both standard-gamble and time trade-off questions to value hypothetical health states, which avoids the limitations associated with the use of health status instruments. In their study, however, age was not included as part of the health state description nor were results stratified by the age of the “index child” considered by the parent during the valuation exercise.

Back to Article Outline

Potential Issues with Using Parents as Proxy Respondents 

One significant methodological issue for valuing child health is the necessity of using proxy respondents for younger children. A sixth-grade reading level has been identified as the minimum for reliable completion of standard-gamble or time trade-off questions.13 Studies that have assessed both parent and adolescent measures of the child's quality-of-life demonstrate significant differences between parent and child ratings.14 Factors such as the effect of childhood illness on parent quality-of-life may affect time trade-off or standard-gamble responses. Little research has focused on how these and other factors might influence parent proxy responses.

It is unclear where a respondent sample of parent proxies fits into the usual taxonomy of respondent groups for valuing health states for cost-effectiveness analysis. For public policy decisions, the Panel on Cost-effectiveness in Health and Medicine4 recommends the use of health-state ratings elicited from a representative community sample.4 The other common respondent group consists of patients who have experienced the health state being valued. Ratings on the basis of a patient sample are not recommended for a “reference case” analysis unless community ratings are unavailable. A parent proxy sample does not fall clearly into either of these groups because parents will not be a representative sample of the US general population and will not represent a patient or experienced population unless parents of children who have experienced a particular health state are specifically recruited for the elicitation exercise. Using a community sample could pose significant challenges that have not been well researched. In a community sample, a non-parent respondent could be required to imagine a hypothetical child and then value a hypothetical illness in the hypothetical child. Alternative approaches include asking non-parent respondents to imagine themselves as a small child during the valuation exercise. Research has yet to address questions on whose values should be used and best practices for elicitation and framing approaches for valuing child health.

Back to Article Outline

Family Perspective May be More Appropriate for Childhood Illnesses 

Another consideration is whether a family-based model of cost-effectiveness is more appropriate for assessing childhood interventions. Standard utility assessment uses an individual perspective, but the effect of a child's illness on a parent or other family members could result in substantial quality-of-life changes. The dependency of children on others for decision-making and caregiving is a key consideration in the relevance of a household perspective.15 This dynamic is likely to change as the child ages; a parent's involvement in decision-making will gradually shift to the child/adolescent, and, at the same time, the effect of a child's illness on a parent's quality- of-life is likely to decrease for illnesses in older children/adolescents. A family-based perspective for economic evaluations that includes changes in a parent's quality-of-life is likely to be most salient for infants and young children and would require the measurement of changes in health utility for both parent and child.

Back to Article Outline

Summary 

Substantial research remains to be done in the field of measuring preferences for child health. Future research should include studies to improve methods for valuing child health in addition to empirical studies, such as that by Carroll and Downs, that provide useful values with the methods currently available. Carroll and Downs contribute to the field in their latest study by applying standard health utility assessment methods to pediatric health states with parents as proxy respondents for children. This study has many strengths, including the large sample size, the comprehensiveness of pediatric health outcomes valued, and the use of well-established elicitation methods. As the authors note, one potential limitation is their use of a parent convenience sample in Indiana that is not consistent with guidelines for cost-effectiveness analysis and may limit generalizability. The study also does not report results by age; stratifying results by the age of the “index child” parents considered when valuing the scenario would strengthen the usefulness of these results. The study does not consider any effects on the parent's quality-of-life. Despite these limitations, this set of utility weights represents an important resource for cost-effectiveness analysts as methodological research continues. Methodological research in valuing child health for cost-effectiveness analyses should explore potential biases of parent proxies, consider domains specific to child health and whether values vary by age, and the appropriateness of a family-based model.

Back to Article Outline

References 

  1. Prosser LA, Hammitt JK, Keren R. Measuring health preferences for use in cost-utility and cost-benefit analyses of interventions in children: theoretical and methodological considerations. PharmacoEconomics. 2007;25:713–726
  2. Petrou S. Methodological issues raised by preference-based approaches to measuring the health status of children. Health Econ. 2003;12:697–702
  3. Petrou S, Gray R. Methodological challenges posed by economic evaluations of early childhood intervention programmes. Appl Health Econ Health Policy. 2005;4:175–181
  4. Petrou S, Henderson J. Preference-based approaches to measuring the benefits of perinatal care. Birth. 2003;30:217–226
  5. Carroll AE, Downs SM. Improving decision analyses: parent preferences (utility values) for pediatric health outcomes. J Pediatr. 2009;155:21–25
  6. In:  Gold MR,  Siegel JE,  Russell LB,  Weinstein MC editor. Cost-effectiveness in health and medicine. New York: Oxford University Press; 1996;
  7. Griebsch I, Coast J, Brown J. Quality-adjusted life-years lack quality in pediatric care: a critical review of published cost-utility studies in child health. Pediatrics. 2005;115:e600–e608
  8. Ladapo JA, Neumann PJ, Keren R, Prosser LA. Valuing children's health: a comparison of cost-utility analyses for adult and paediatric health interventions in the US. PharmacoEconomics. 2007;25:817–828
  9. Ungar WJ, Santos MT. The pediatric quality appraisal questionnaire: an instrument for evaluation of the pediatric health economics literature. Value Health. 2004;6:522–531
  10. Dolan P. Output measures and valuation in health. In:  Drummond MF,  McGuire A editor. Economic evaluation in health care. New York: Oxford; 2001;p. 46–67
  11. Feeny D, Furlong W, Boyle M, Torrance GW. Multi-attribute health status classification systems: Health Utilities Index. PharmacoEconomics. 1995;7:490–502
  12. Hennessy S, Kind P. Measuring health status in children: developing and testing a child-friendly version of EQ-5D. In: Scientific proceedings of the EuroQol Group.York. United Kingdom: EuroQol Group. 2002;291–310
  13. Juniper EF, Guyatt GH, Feeny DH, Griffith LE, Ferrie PJ. Minimum skills required by children to complete health-related quality of life instruments for asthma: comparison of measurement properties. Eur Respir J. 1997;10:2285–2294
  14. Saigal S, Stoskopf BL, Feeny D, Furlong D, Burrows E, Rosenbaum PL, et al. Differences in preferences for neonatal outcomes amoung health care professionals, parents, and adolescents. JAMA. 1999;281:1991–1997
  15. Keren R, Pati S, Feudtner C. The generation gap: differences between children and adults pertinent to economic evaluations of health interventions. PharmacoEconomics. 2004;22:71–81

PII: S0022-3476(09)00220-0

doi:10.1016/j.jpeds.2009.03.007

Refers to article:

  • Improving Decision Analyses: Parent Preferences (Utility Values) for Pediatric Health Outcomes , 27 April 2009

    Aaron E. Carroll, Stephen M. Downs
    The Journal of Pediatrics July 2009 (Vol. 155, Issue 1, Pages 21-25.e5)

The Journal of Pediatrics
Volume 155, Issue 1 , Pages 7-9, July 2009

Article Tools