Reviewing and refining the safe use of health information technology systems to improve patient safety in healthcare

Over the last three decades, health information technology (HIT) has been an inherent part of modern healthcare, improving workflow and services to be more effective and efficient. However, HIT system deployment, adoption, and implementation can create significant disruptions, such as interoperability and interfaces between various other systems. Despite many benefits and strengths, HIT has made healthcare a truly complex socio-technical system. This layer of complexity may pose substantial risks to patient safety and organisational efficiency that may compromise intended benefits. Thus, HIT can cause patient harm, patient inconvenience, or delays in care delivery, both individually and collectively, often in unexpected and unforeseen ways (1,2).

Therefore, there is a need to identify the types of patient safety incidents, their contributing factors, the consequences of those incidents, and actions taken to manage those incidents. This would help minimise the risk to patient safety and devise a set of strategies to prevent them. The project will also determine the role of Machine Learning Algorithms (MLA) and Artificial Intelligence (AI) in identifying HIT-related incidents, improving decision-making, and thus minimising the risk to patient safety (3).

Research Questions

1. What are the HIT-related (hardware and software) patient safety incidents that occur in routine clinical practice?
2. How do human, technical, and organisational factors contribute to those HIT-related patient safety incidents? What are the consequences of those incidents, and what actions are taken to minimise the risks of HIT-related patient safety incidents?
3. How can Machine Learning Algorithms (MLA) and/or Artificial Intelligence (AI) be used to identify HIT-related incidents, improve decision-making, and thus minimise the risk of patient safety?

Methods

The project will comprise three phases.

• Phase 1 (15 months): To address RQ1 and RQ2, we will use retrospectively collected HIT-related patient safety incident reports provided by the Medicines and Healthcare Products Regulatory Agency (MHRA) from the UK and Swedish Medical Products Agency (MPA). We will use both deductive and inductive approaches to analyse the incident reports’ free-text narratives.
• Phase 2 (12 months): To gain more contemporary and additional information about what goes wrong with HIT and particularly to discuss, in some detail, how incidents may be prevented and better managed, we will report the study findings of Phase 1 and then work with MHRA/MPA members and patient safety Leads and Directors. We will conduct two Focus Group Discussions (FGD) (2X8=16 participants) to glean additional information about the context and regularly used recovery strategies, particularly the role of MLA and/or AI, in minimising the risk of HIT incidents (framework analysis). This will be supplemented by a systematic review regarding the role of MLA and/or AI in preventing HIT incidents through the decision-making process to translate the findings into actionable guidance (RQ1-3).
• Phase 3 (9 months): The developed guidance on mitigating the risks of patient safety HIT incidents will be validated using an online Delphi survey of expert opinions (of the UK and Sweden) to test its relevance and acceptability. This will ensure the progressive refining of the guidance and the generation of general findings in collaboration with the above stakeholders. An agreement will be reached after the final round of questionnaires among the experts to view the validated guidance in alignment with the English and Swedish Healthcare (RQ3).

NB: Patient and Public Involvement (PPI) representatives will be integral to all phases 1-3, recruited by the UoB or the researchers, ensuring a collaborative and inclusive approach to the research.

Ethical considerations

Given the cross-national nature of the project, we will seek ethics approval/guidance from multiple ethics committees, i.e., the Ethics Committee of the UoB, as well as the Ethical Advisory Board of Southeast Sweden, to ensure the highest ethical standards are maintained throughout the research.

Key research output/ impact

• Generate valuable knowledge and insights into HIT-related patient safety incidents
• A clear, actionable, evidence-based guidance for identifying HIT incidents with refined existing incident reporting systems
• Better use of HIT by healthcare and more prospects of developed generalisable strategies to adopt elsewhere – a safer and more sustainable healthcare system
• A set of evidence-based policy recommendations for healthcare in the safe and effective use of HIT systems
• Dissemination of findings in articles for academic journals and conference presentations

How to apply

Formal applications can be made via the University of Bradford web site.