Evaluation of quality of life of childhood cancer survivors: a methodological conundrum.

QOL assessment in pediatric oncology is seriously understudied, especially compared with the adult population. The limited progress is due to the methodological complexity of the task, which should not be viewed as insurmountable. Given a precise study question, the methodological issues can be clarified simply, piece by piece. Researchers must consider very carefully the specific characteristics that define a study population in order to choose an instrument that is domain-appropriate and valid for the assessment paradigm. The first priority should be that a researcher must identify the means of accessing the information of interest. In the pediatric population, information about children's status may be elicited from parents, medical personnel, teachers, or the children themselves. Clearly, the type of instrument to be used for assessment is dependent on the choice of reporter. Researchers must also account for developmental age and disease; in assessing generic and disease-specific functioning, the "functional scale" against which an individual is compared must implicitly reflect the types of activities and/or levels of functioning that are realistic norms for the patient. Equally important is the analysis of independent domains in order to characterize the dynamics/divergence of clinical status and functional status. What are the merits of conducting QOL research for the pediatric cancer-survivor population? The policy implications are profound and pervasive both for the individual survivors (regarding treatment, care, and his/her ultimate ability to reintegrate into society) and for society (regarding resource allocation, cost planning, and productivity). Commensurate with the rapid advancement of oncologic therapy, there is now an expanding cohort of pediatric cancer survivors. Current estimates suggest that, by the turn of the century, 200,000 children will be in this category. The long-term survivorship of this cohort is still poorly defined. However, as the survivors mature, it is likely that their needs will evolve as well-whether for treatment of secondary malignancies, long-term morbidities, and fertility issues or for neuropsychological dysfunction, emotional counseling, or occupational issues. Children, as survivors, are unique, in that their future (the context within which long-term outcome is defined) spans decades. Based on a median age at diagnosis of 6 years, survivors can expect to live an additional 66 years. From a cost or policy perspective, children represent enormous future potential. The implications of children's long-term outcomes must be considered regarding the change in future potential secondary to survivorship. Pediatric QOL research plays a role both inside and outside the health care system. Clearly, in the provision of health care, QOL data may be used to improve or modify patient care by supplementing information about the clinical status of individual patients. Information about an individual's general functioning, particularly as it diverges from disease-specific functioning, complements clinical data to facilitate comprehensive care. Information about the long-term outcomes of pediatric cancer, as a whole, will influence the policies of health care institutions and the allocation of health care resources. By expanding the scope of survivorship (or cure) to include long-term clinical and general "costs" the "cost of cure" is shifted: this shift will ultimately impact estimations of cost effectiveness, with ramifications for the evaluation of hospital-wide protocols, utilization priorities, and cost policies. Outside of the hospital, the implications of QOL research are equally ubiquitous. Pediatric survivors will live an estimated 7 decades after "cure," during which time they will exist almost entirely outside the realm of health care; yet, their condition as a survivor, with or without the long-term clinical toxicities secondary to treatment, will continue to affect some or all of thei