ASSESSING THE QUALITY OF HEALTH-CARE PROVIDED TO CHILDREN

Objective. To present a conceptual framework for evaluating quality of care for children and adolescents, summarize the key issues related t o developing measures to assess pediatric quality of care, examine som e existing measures, and present evidence about their current level of performance. Principal Findings. Assessing the quality of care for ch ildren poses many challenges not encountered when making these measure ments in the adult population. Children and adolescents (from this poi nt forward referred to collectively as children unless differentiation is necessary) differ from adults in two clinically important ways (Ja meson and Wehr 1993): (1) their normal developmental trajectory is cha racterized by change, and (2) they have differential morbidity. These factors contribute to the limitations encountered when developing meas ures to assess the quality of care for children. The movement of a chi ld through the various stages of development makes it difficult to est ablish what constitutes a ''normal'' outcome and by extension what con stitutes a poor outcome. Additionally, salient developmental outcomes that result from poor quality of care may not be observed for several years. This implies that poor outcomes may be observed when the child is receiving care from a delivery system other than the one that provi ded the low-quality care. Attributing the suboptimal outcome to the ne w delivery system would be inappropriate. Differential morbidity refer s to the fact that the type, prevalence, and severity of illness exper ienced by children is measurably different from that observed in adult s. Most children experience numerous self-limited illnesses of mild se verity. A minority of children suffer from markedly more severe diseas es. Thus, condition-specific measures in children are problematic to i mplement for routine assessments because of the extremely low incidenc e and prevalence of most severe pediatric diseases (Halfon 1996). Howe ver, children with these conditions are potentially the segment of the pediatric population that can be most affected by variations in the q uality of care. Improving the care provided to these children is likel y to have the largest impact on quality of life and longevity. The low prevalence of most severe pediatric diseases also makes it difficult to evaluate the effectiveness of new treatment modalities; multi-cente r trials or long enrollment periods are usually required to obtain a l arge enough patient sample to conduct the necessary randomized control led trials or cohort studies. Another challenge encountered when measu ring quality of care for children is that, in most cases, they depend on adults to both obtain care and to report on the outcomes of that ca re. Parents and their children may have different perceptions of what defines health or have different levels of satisfaction with the care they receive. Children, particularly those with special needs, also de pend on a broad range of services including the medical system, commun ity intervention programs, social programs, and school-based services. Dependency on these various services adds to the difficulty of measur ing and appropriately attributing health outcomes observed in children to a particular service delivery entity. Adolescents also depend on a dults for access to some of their care; however, they have special nee ds related to confidentiality and parent-child information sharing. Ad olescents commonly seek care at facilities, such as school-based clini cs, that allow them to obtain confidential care. These facilities usua lly provide out-of-health plan care for these children, which raises s pecial issues related to information availability for quality assessme nts and for assessing utilization patterns in this population. If the source of poor health outcomes is not known, quality improvement is no t possible. The many challenges faced when constructing pediatric (thi s term will be used to refer to both children and adolescents) quality of care measures have resulted in few of these instruments being deve loped specifically for children. Most of the measures developed to dat e have either a very limited pediatric component or still require the process or outcome validation step. Although several practice guidelin es and indicators of quality have been constructed, a conceptual frame work to guide the development of such tools for quality assessment in the pediatric population is lacking. Conclusions. Pediatric health ser vices researchers and the organizations that fund this work need to fo cus on developing a set of quality assessment tools that will address several challenging issues. Working within the context of the conceptu al framework presented, we draw several conclusions related to issues that should be considered in developing quality of care measures for c hildren.