Crossing the Line: How Threshold-Based Tools Shape Sepsis Decision-Making in the NHS

About the Project

Background

Across the NHS, threshold-based decision-support tools are widely used to structure clinical assessment and escalation. These tools simplify complex judgements by assigning numerical thresholds that trigger predefined actions, intended to standardise care and reduce missed deterioration. Identification of sepsis has been a key focus of such tools, as missed or delayed recognition remains a major patient-safety concern with significant risk of harm.

To support timely recognition and initiation of treatment, including antibiotics, quantitative early-warning scores and risk categories - most prominently based on NEWS2, and SIRS - have been adopted that embed threshold-based triggers for action (1). Threshold-based tools are designed to reduce uncertainty, standardise decision-making, and support communication about escalation, and are intended to be used alongside clinical judgement. However, NEWS scores have been found to have limited accuracy for predicting time-critical treatment needs (2), generate high alert rates relative to true events, produce false positives, and can encourage risk-averse behaviour with significant resource implications (3).

While individual clinician and patient decision-thresholds have been studied (4), less is known about how clinicians interpret, respond to, or override tool-generated thresholds such as NEWS2. Dual-process theories of decision-making and research on risk perception suggest thresholds shape tolerance for uncertainty, perceived responsibility, and defensive strategies. Threshold-based tools can act as a “bright-line” (5), with socio-organisational consequences, as responsibility and accountability are often tied to crossing the threshold.

Using social science theory and methods, this project will explore how clinicians respond to threshold-based scoring systems, and how these tools contribute to both over- and under-treatment in sepsis-related decisions, with implications for improving clinical decision support and mitigating unintended consequences.

Aim

To explore how threshold-based scoring systems, such as NEWS2, shape clinicians’ decision-making in sepsis care within a wider organisational context, critically evaluating how and why these tools contribute to both over- and under-treatment.

Objectives

• Explore how clinicians interpret and use threshold scoring tools in sepsis decision-making, including instances of threshold override or discounting.
• Apply psychological and decision-science frameworks (dual-process theory, heuristics and biases, signal detection theory (6)) to generate an empirically grounded account of threshold-based decision-making.
• Analyse how organisational risk cultures—policies, performance pressures, documentation requirements, and local norms—shape the use of threshold-based systems, and how these dynamics contribute to patterns of over- and under-treatment.

Methods

A systematic review of the benefits and limitations of threshold-based scoring systems in healthcare will be completed. A qualitative, multi-site ethnographic study will be conducted. Approximately 150 hours of non-participant observation in emergency and acute medical settings across two NHS hospitals will capture real-time decision-making and interactions around scoring tools. Semi-structured interviews with ~40 clinicians will probe reasoning processes, risk perception, uncertainty, and experiences of threshold use or override. Document and artefact analysis of national and local sepsis policies and escalation protocols will map the cognitive and organisational cues shaping decisions. Data will be analysed thematically and interpreted using social science theory and frameworks.

Expected Contribution

The project will generate a critical account of how threshold-based scoring systems interact with clinicians’ cognitive processes and organisational influences. It will provide actionable insights to improve sepsis decision making and reduce over- and under-treatment, and more broadly, contribute to understanding the consequences of threshold-based decision-support tools across the NHS.

Outputs

The project will provide an empirically grounded critical account of how threshold-based tools such as NEWS2 influence clinicians’ decision-making, risk perception, and treatment behaviour in sepsis care. Outputs will include peer-reviewed publications, conference presentations, and policy-relevant recommendations for improving threshold-based decision support and mitigating unintended consequences, including antibiotic overuse. Practitioner-focused summaries will provide guidance for decision-support tools across the NHS and other clinical settings.

Training

The PhD student will receive advanced training in qualitative methods, including ethnography, observation, and thematic analysis, supported by supervision and social science theory training through the University of Leicester. They will have opportunities for peer learning from qualitative researchers within the SAPPHIRE research group. Training on patient safety and implementation science can be provided via the Greater Manchester NIHR Patient Safety Research Collaboration. Additional opportunities will include training in presenting research to diverse academic and policy audiences, and developing skills in writing for peer-reviewed publication, enhancing both research expertise and professional impact.

Apply at:

https://le.ac.uk/study/research-degrees/research-subjects/health-sciences

PhD entry requirements:https://le.ac.uk/study/research-degrees/entry-reqs

Supervisor contact details:

Prof Carolyn Tarrant, ccp3@leicester.ac.uk

Dr Eva Krockow, emk12@leicester.ac.uk