Abstract
In the United Kingdom, hospital patients suffer preventable deaths ("failure to rescue") and delayed admission to the Intensive Care Unit because of poor illness recognition. This problem has consistently been identified in care reviews. Strategies to improve deteriorating ward patient care, such as early warning systems and specialist care teams (Critical Care Outreach or Rapid Response), have not reliably demonstrated reductions to patient deaths. Current research focuses on failure to rescue, but further reductions to patient deaths are possible, by examining care of unwell hospital patients who are rescued (successfully treated). Our primary objective is to develop a framework of care escalation success factors that can be developed into a complex intervention to reduce patient mortality and unnecessary admissions to the Intensive Care Unit (ICU).
Introduction: Shifting the Paradigm to "Safety-II"
Traditional patient safety research focuses on "Failure to Rescue" (FTR)—the inability to recognize and respond to reversible complications. While identifying errors is crucial, it often overlooks the human and organizational factors that lead to successful outcomes. The SUFFICE study adopts a "Safety-II" philosophy: instead of just trying to eliminate errors, it seeks to understand and reinforce the conditions that allow for successful "rescues"—when a patient becomes severely unwell but is successfully treated on the ward without needing the ICU.
The Problem: Beyond Early Warning Scores
In the UK, an estimated 11,000 preventable deaths occur annually due to FTR. While Early Warning Scores (EWS) and Rapid Response Teams are standard, they haven't consistently reduced mortality. There is a significant gap in understanding the "soft signals" of deterioration and the expert decision-making processes that algorithmic tools often miss. Escalation is a social and communicative process involving situational awareness, leadership, and staffing dynamics.
Methodology
The SUFFICE study is a multicentre mixed-methods, exploratory sequential study conducted across two National Health Service (NHS) Trusts. The research is structured into four distinct phases designed to build a comprehensive knowledge base.
- Phase 1 (Ethnographic Observations): Researchers shadow clinical staff to capture real-time communication and team dynamics during escalation events.
- Phase 2 (Notes Review): A retrospective deep dive into patient records using the Structured Judgement Review (SJR) method to identify care success factors in patients who avoided ICU admission.
- Phase 3 (Expert Interviews): Utilizing Applied Cognitive Task Analysis (ACTA) to elicit "hidden" knowledge from expert clinicians (doctors, nurses, and AHPs) regarding how rescue is achieved under pressure.
- Phase 4 (Data Integration): Synthesis of findings to generate a theoretical framework of care escalation success factors.
Study Targets and Data Collection
The study aims to collect a robust dataset to ensure the validity of the resulting framework, targeting specific volumes of observations and reviews to ensure statistical and qualitative depth.
- Observational Data: 200–400 witnessed escalation events.
- Clinical Records: 400 retrospective care records from high-risk patients (EWS ≥7).
- Qualitative Data: Transcripts from 30 ACTA interviews with staff having >4 years of clinical experience.
Expected Outcomes and Future Work
The primary output will be a Framework of Success Factors, categorizing what worked, the specific outcomes facilitated, the context of success, and potential balancing measures to avoid unintended consequences.
Following the development of this framework, the researchers plan to model a complex intervention and conduct a feasibility study. The ultimate goal is to provide a toolkit or protocol that can be implemented across the NHS to improve patient rescue rates and reduce unnecessary ICU admissions.
Limitations
The study acknowledges potential challenges, including the Hawthorne effect during observations (where staff may change behavior while being watched), the limitations of retrospective record documentation, and the site-specific nature of the research (two UK-based Trusts).
Summary updated