Thesis-2020-Carman.pdf (3.22 MB)
Download file

The investigation of weak signals for safety and to aid task success in healthcare

Download (3.22 MB)
posted on 07.09.2021, 11:50 by Evi Carman
Due to the complexity of the healthcare system and the nature of healthcare work, the paradigms of resilience engineering, high reliability organisations and Safety-II can offer this setting novel means of improving safety by promoting proactive safety. These paradigms advocate and promote a system’s ability to succeed under varying conditions. One means of adjusting performance to ensure success of a task, is anticipating, identifying and responding to weak signals indicating changes in the system. These signals may provide an opportunity for proactiveness and resilience by identifying problems that threaten safety and allow for a more efficient and timely response. Despite accident reports, including those on healthcare disasters, increasingly stating weak signals were present prior to adverse events, research exploring weak signals and the role they may play in safety, especially in healthcare, is limited. This research aimed to address the following research question: What are the weak signals in healthcare that may identify that a task or process may not be completed successfully?
To address this question, this research aimed to provide the theoretical basis and approach to guide research into weak signals that are present in healthcare work systems. By adopting a qualitative explorative approach, the forms and sources of weak signals in healthcare were established, the underlying detection and interpretation mechanism were explored, and the potential applications examined. This was achieved through addressing several objectives that spanned four case studies. The objectives included developing a definition, framework and analysis approach to guide weak signal research in different healthcare contexts; depicting the work systems for the different healthcare contexts in which weak signals were explored and identifying the practical implications and different elements of weak signals that may support improvement work in the healthcare environment.
The definition and preliminary framework were developed through the collation of relevant literature on weak signals and information processing and macro-cognition theories. Weak signals in this research were defined as ‘information that provide an indication of upcoming or emerging events that may have significant implications, whereby the information provided is often imprecise, ambiguous and incomplete, which requires active identification and processing’. The purpose of the framework was to provide a structure for the identification and analysis of weak signals in the healthcare context. This framework was tested in a first case study on patient handling, that consisted of two focus groups with experts in patient handling. Based on these results and additional literature the framework was further refined and expanded. The second and third case studies focused on testing and applying the framework to two different healthcare contexts, one a specific process and the other a syndrome-specific context.
Case Study 2 explored the work system and weak signals for the hospital discharge process for patients that still require additional healthcare input from community healthcare staff. This case study applied a mixed methods approach utilising five different data sources. These consisted of five observation sessions, three focus groups held with hospital staff involved in the discharge process, six focus groups and ten interviews held with community healthcare staff involved in the discharge process, as well as the analysis of 348 reported incidents and one Ombudsman report associated with the discharge process. This case study yielded a large number of weak signal examples (a total of 435), that could be categorised into 46 category types based on the source. This case study assisted in expanding the weak signals in healthcare framework to include key behavioural elements associated with the identification, interpretation and prediction mechanisms underlying weak signals.
Case Study 3 explored the work system and weak signals for the identification and treatment of sepsis, using the third version of the weak signals in healthcare framework. In this case study, two Ombudsman reports and a total of 99 survivor and tribute stories were analysed, to identify the examples, structural and behavioural elements of weak signals. A key addition to the framework as a result of this case study was ‘rationalisations’, that are indicative of underlying behavioural elements of the framework.
Case Study 4 explored the application of the fourth and final version of the weak signals in healthcare framework and analysis approach to a different healthcare theme. In this case study, the presence of weak signals in the organisational failings of the Mid Staffordshire NHS Foundation Trust were explored. This application analysed the independent and public inquiry reports to identify the precursory weak signals and their rationalizations that occurred during this event.
The main contributions of this research were a conceptual structure for the knowledge on weak signals, the means of applying this theory, as well as knowledge on weak signals that present in specific healthcare contexts. The framework has several key applications, namely, to provide a structured approach to investigating weak signals in healthcare and to support use of weak signals on an individual level, through other system applications and for design of intervention or improvement projects.



  • Design and Creative Arts


  • Design


Loughborough University

Rights holder

© Eva-Maria Carman

Publication date



A thesis submitted in partial fulfilment of the requirements for the award of the degree of Doctor of Philosophy of Loughborough University.




Patrick Waterson ; Mike Fray

Qualification name


Qualification level


This submission includes a signed certificate in addition to the thesis file(s)

I have submitted a signed certificate

Usage metrics