<b>The food environment in three South African public tertiary hospitals and legal doctrinal analysis of policy levers for a healthier hospital food environment</b>

Data collection for the survey was via a 5-10-minute, fieldworker-administered, semi-structured electronic questionnaire (see appendix xyz) to recruited and consenting participants. The questionnaire had been pilot-tested to assess and improve timing, clarity, utility and data quality prior to implementation. To improve representativeness, data were collected on different days of the week and at various times of the day (morning, afternoon, and evening).</p><h3 dir="ltr">Interviews</h3><p dir="ltr">Interviews were conducted by trained researchers with no prior relationships, and debriefings helped document reflections.

Trustworthiness was enhanced through trained interviewers, verbatim transcripts, coding frameworks, and detailed descriptions to support transferability. We developed the interview guide based on the study objectives. Its purpose was to explore participants' motivations, perceptions, and experiences regarding food choices within hospitals.

The guide was designed to examine reasons for purchasing or bringing specific foods, factors influencing food and beverage choices, satisfaction with available food options, perceived affordability, convenience, quality, and variety of foods, views regarding healthier food alternatives and suggested improvements to the hospital food environment. Questions were intentionally open-ended to allow participants to elaborate on their experiences.

Follow-up probes were included to obtain richer information on topics such as perceptions of healthy food, and barriers to making healthier food choices. The guide was also reviewed and refined during fieldworker training and pilot-testing to improve clarity, relevance, and consistency in administration. We purposively sampled thirty (30) participants (staff, outpatients and visitors) from participants enrolled for the intercept survey across the three hospitals.

We conducted in–person semi-structured interviews on hospital premises. All interviews were audio–recorded and transcribed and translated into English.</p><h3 dir="ltr">Food inventory audit</h3><p dir="ltr">A bespoke food inventory tool was developed to systematically characterise the consumer food environment across participating public hospitals. The audit was designed to document all food retail outlets operating within hospital premises, including cafeterias, kiosks, general shops, restaurants, fast-food outlets and vending machines.

For each outlet, trained fieldworkers recorded availability of food and beverage items, price ranges and promotional activities using an electronic audit tool administered on tablet devices (see appendix xyz). The audit tool was pilot tested before data collection commenced, and minor revisions were made to improve clarity and consistency. All fieldworkers received standardised training on the use of the audit tool prior to data collection.

To ensure consistency in data collection, inter-rater reliability was assessed by having two trained fieldworkers independently complete the audit for the same randomly selected food outlet. The completed audits were compared, discrepancies were discussed, and the audit procedures were refined to improve validity of the audit tool.</p>