Understanding the implementation work needed to support Rescue for Emergency Surgery Patients Observed to uNdergo acute Deterioration (RESPOND): a qualitative study

Background Failure to recognise and respond to acute deterioration in hospital patients remains a persistent international challenge and is the most common identifiable treatment‑related factor in avoidable in‑hospital deaths in the UK. Improvement initiatives have typically focused on standardised monitoring and escalation processes, with variable outcomes. Few have addressed broader human‑factors issues, including the interplay between systems, cultures and technologies that shape rescue processes. Little is also known about the implementation work required to support such multifaceted approaches. Methods The RESPOND programme comprised of a suite of interventions to strengthen the identification of, and response to, deterioration in patients following abdominal surgery. The four intervention strands: included a patient‑led escalation system, team‑strengthening activities, redesigned escalation procedures and work to improve cross‑departmental relationships. Implementation strategies comprised education, engagement, and audit and feedback. We undertook an independent qualitative process evaluation of the feasibility stage of the RESPOND programme, using ethnographic methods to assess implementation across three English hospital pilot sites between November 2022 and June 2023. Data included interviews with clinical and managerial staff ( n = 30), patients and relatives ( n = 15), and over 100 h of ward‑based observation. Analysis was informed by Normalization Process Theory (NPT) to examine how contextual factors shaped the sense‑making, enrolment and operational work required to implement RESPOND. Results Whilst RESPOND was implemented across all sites, variation in uptake between strands and sites was illuminated through the NPT constructs of coherence, cognitive participation and collective action. The wider context of surgical care also intersected with the collaborative work required for implementation. Our findings revealed the significance of readiness for respond, in particular workplace turbulence; lack of coherence as to the nature of the problem (and appropriate solution); ease of integration of the different strands into workflows; and leadership support and local collaborative efforts for successful implementation of the programme. Conclusions This feasibility study highlighted the scope of implementation work required to support RESPOND. Clearly articulating the role of human factors in strengthening rescue processes, alongside the purpose of individual intervention strands, was critical. Site readiness - including resource capacity, senior clinical support, situated project management and local tailoring to existing workflows - was identified as essential for successful roll‑out.