Analysis

A 5-member research team, comprised 2 clinician-researchers (J.K.L. and C.E.M.), a parent consultant (C.A.S.), a research scientist (K.C.A.), and a qualitative methodologist (A.R.S.) conducted the analysis using a general inductive approach rooted in grounded theory.¹⁶ Interviews with clinicians and parents were analyzed concurrently, and emerging concepts were summarized in a jointly developed codebook and coding framework, adding concepts as needed. The same codebook was used to develop a conceptual model informed by multistakeholder perspectives for both parent and clinician interviews. After identifying initial concepts, we recognized parallels between our findings and the quality of care framework developed by Coyle and Battles¹⁷, which expands on Donabedian’s structure-process-outcome model to incorporate the “antecedents” of medical care (see Table 2, Supplemental Digital Content 2, http://links.lww.com/PQ9/A315 for definitions and adaptations). We also noted that participants’ descriptions of healthcare quality aligned with the Institute of Medicine’s quality of care domains.¹⁸ We, therefore, used these models to guide our subsequent analysis.

Three full interviews and selected text from five interviews were triple-coded by three members of the research team (A.R.S., C.A.S., and K.C.A.). Disagreements were resolved through in-depth discussions and revisions of the concepts, corresponding codes, and definitions. Following codebook finalization, these 3 research team members coded all transcripts using Dedoose.¹⁹ Through an iterative process of axial coding,²⁰ emergent themes and domains were summarized in a conceptual model. For validity, we member-checked our findings with a 12-member multistakeholder advisory board of parents and healthcare providers upon completion of analysis.²¹ This advisory board was convened to provide diverse perspectives on our direct admission study design and findings; advisory board members were not interview participants.