TABLE 2.

Procedures to promote rigour

Criteria	Definition	Description of steps taken in this study
Credibility	The extent to which an interpretation of data is representative of the experiences of participants	Discussion of data and themes with co‐author group (peer debriefing) to check that interpretations was representative of experiences. Co‐author group are from multidisciplinary backgrounds including: Psychology (KH, FW), Nursing (SC, DC), Medicine (ET), Physiotherapy (AF). Study findings were also presented to a public and patient involvement group consisting of five members (recruited from a local older people's action and support group). Participants suggested that findings relating to ward culture and staff shortages resonated with their own experiences. The group was also glad to see isolation and lack of stimulation was included as they felt strongly that this was a key factor in older peoples decline during a hospital stay.
Transferability	The extent to which findings might be applied or generalized to other participants in similar contexts	To inform readers judgements about transferability, we have included relevant contextual information about sites and participants in the findings.
Dependability	The extent to which a researcher's interpretation of data would be consistent if repeated	We used NVivo software to provide a clear audit trail for the analysis.
Confirmability	The extent to which the findings of the study are free from bias	Data were initially coded line‐by‐line using terminology which stayed close to the original data (and thus participant's experiences). In developing the themes, we actively explored atypical experiences to refine our interpretations. Two researchers coded a sample of transcripts (see method for further details) to ensure there was agreement on the coding of risk factors.