Table 1. Methods used to pre-test and pilot the Hospital Consumer Assessment of Health Providers and Systems survey, Odisha, India, 2020.

Stepa	Purpose	Process	No. of participantsb
1. Cognitive testing	To refine translation of survey tool. To ensure variation in responses do not reflect differences in understanding of a given question, we aimed to identify how individuals interpret each survey item and how their cognitive processing relates to the construct intended by the researcher and original survey instrument	Focus groups discuss all survey items to assess if framing is logical and answerable, if response options are adequate, etc. We paired each item with structured verbal probes to elicit participants’ cognitive processes and assess their understanding and interpretation of each survey item	50
2. Item testing and exploratory factor analysis	Quantitively assess how survey items relate and if exposure to quality of care informs our overall variable of interest: patient satisfaction	Hospital-based exit interviews with eligible patients; responses anonymized and analysed using an exploratory factor analysis and series of ordinary least squares models with overall satisfaction posed as a dependent variable, controlling for patient complexity and interview characteristics, for example privacy and enumerator ID	507
3. Content validity indexing	Assess to which extent the tool items represent facets of the construct patient experience, that is, do the survey items represent what is important to patient-centredness in Odisha, India	One hour-long individual interviews, conducted in non-clinical settings with five patients, five health workers and five health-system experts. For each survey item, each interviewee rates the relevance to patients’ satisfaction and relevance given hospital environment, using a four-point Likert scale. Subsequently, interviewees describe the reasons for their ratings	15

^a Steps were conducted consecutively.

^b Participants partook only in one step, that is each group was distinct.