Table 1.

Overview of multilevel data collection and analysis in Virtual Online Consultations: Advantages and Limitations (VOCAL) study.

Data source	Type and nature of data	First-order interpretation	Higher order categories
Macro-level study of the wider context for introducing video consulting	Accounts of national-level stakeholders (36 informal and 12 formal semistructured interviews); 50 national-level documents from 2000 onwards (including policies, guidance, and national-level announcements)	Historical and policy drivers for the move to video consultations; system-level blocks	External social structures such as political, regulatory and economic context; background and context to multilevel analysis
Meso-level study of organizational change	Accounts of 24 staff involved in delivering video consultations; approximately 300 hours of observations across 3 clinics; 16 documents (eg, operating procedures and meeting minutes) and researcher field notes about people and technologies delivering video consultations; diagrams and accounts of how people, technologies, and clinical work relate and interact	Key interactions and interdependencies; key organizational routines and how these are changing over time	External social structures (such as professional standards and definitions of excellence, symbolic meaning of illness); internal social structures (what actors “know” and how they interpret the strategic terrain, such as “scripts” held by patients and staff about how they should behave and how they change over time); assumptions built into the technology about, for example, capability of users, how people interact, privacy and consent, the nature of clinical work and routines and how all these interact
Micro-level study of virtual consultations	Video-recording and screen capture (at patient end and clinician end) of 30 virtual consultations (18 diabetes, 12 cancer); field notes from before or after the consultation at patient and clinician end	What is said and done in (video and face-to-face) consultations; unfolding interaction and strategies for communication; how technology shapes and constrains (video and face-to-face) consultations; how participants felt	External social structures (such as professional standards and definitions of excellence, symbolic meaning of illness); internal social structures (what actors “know” and how they interpret the strategic terrain, such as “scripts” held by patients and staff about how they should behave and how they change over time); assumptions built into the technology about, for example, capability of users, how people interact, privacy and consent, the nature of clinical work and routines and how all these interact
Micro-level study of matched face to face consultations	Video-recording of 17 face-to-face consultations (12 diabetes, 5 cancer); field notes from before or after the consultation	What is said and done in (video and face-to-face) consultations; unfolding interaction and strategies for communication; how technology shapes and constrains (video and face-to-face) consultations; how participants felt	External social structures (such as professional standards and definitions of excellence, symbolic meaning of illness); internal social structures (what actors “know” and how they interpret the strategic terrain, such as “scripts” held by patients and staff about how they should behave and how they change over time); assumptions built into the technology about, for example, capability of users, how people interact, privacy and consent, the nature of clinical work and routines and how all these interact
Descriptive and demographic data in the video consultation service	Number of patients offered video consultation option and proportion who accept and persist with it; start and finish time; DNA rate for video and face-to-face options; unscheduled encounters (eg, urgent care) for index condition	Acceptability/popularity of the service; demographic data (eg, uptake by age or ethnicity); failed encounter rate; risk of missing serious problems; consultation length	Background and context to multilevel analysis