Findings from this study are consistent with previous studies documenting the primary care provider and practice characteristics that families of CSHCN value most (DeCamp et al., 2013; Han et al., 2013; Zickafoose et al., 2013; Bechtel & Ness, 2010; Nelson et al., 2005; Sia et al., 2004). With regard to providers, these characteristics included having confidence that the physician knew and cared about their child and family; had expertise or sought out information related to their child's special needs; listened to them; respected their knowledge, opinions, and values; and involved them in decision-making. Valued characteristics related to the practice as a whole included having continuity in the care they received, feeling welcomed by well-informed staff, receiving care management support (including information sharing with specialists and other providers), having superb access to the care team and dedicated care coordination staff, and being linked to community supports. Parent-leaders in our sample discussed the ways in which a number of these characteristics worked together to create a culture of patient-centeredness in PCMHs. Parent-leaders also emphasized that provision of care management and coordination services was an important and valued feature of PCMH models. Comments offered by parent-leaders showed that they are strong supporters of PCMH initiatives, especially for CSHCN and their families due to the multiple providers and complex coordination required to meet these children's needs. Nonetheless, they offered a variety of suggestions for enhancing PCMH initiatives and models so that they further benefit CSHCN and their families. Suggestions included expanding the reach of PCMHs into additional arenas important to CSHCN, devising a larger role for specialists, increasing parents' awareness of the PCMH concept, and including parents' input during recognition processes and ongoing quality improvement activities.
This study was designed to be exploratory. The design and methods have three primary limitations. First, study participants represent small convenience samples of parents in two states and an even smaller number of PCMHs. This limitation precludes generalization of findings to parents of CSHCNs generally. Second, all respondents participated voluntarily in the study rather than being randomly drawn from a full population. Voluntary participation creates self-selection bias. Parents of children who receive primary care at PCMHs were recruited based on their response to a flier given to them by their physician. This could bias the data in two ways. For one, providers could have distributed fliers only to those parents with who they knew to have a favorable view of his or her practice, thus overstating the quality and characteristics of care at PCMHs generally. Also, parents who responded to the flier may be significantly different (potentially higher functioning and generally more positive) than parents who did not respond to the flier. Parent-leaders at advocacy or policy organizations also volunteered to participate in the study and were identified based on the recommendations of experts in the area of CSHCN. It is possible that those who were recommended and chose to participate in the study were different in important ways from those leaders who were not recommended and did not participate. For example, those who agreed to participate may have had stronger positive or negative feelings about medical home initiatives and activities. Third, the study lacks a counterfactual against which to compare preferences of parents of CSHCN who receive services from pediatric practices not recognized as medical homes. In spite of these limitations, the study yielded compelling information which can be used to draw findings about parents' experiences with and perspectives on medical home models.