Learning analytics: Decreasing preventable harm in healthcare
The development of a national learning analytics (NLA) capability represents a transformative opportunity for the future of healthcare in Ireland, write Martin McCormack, Chief Executive Officer, and George Shorten, Immediate Past President, College of Anaesthesiologists of Ireland.
The landmark report To Err is Human (2000) by the Institute of Medicine exposed a critical truth: as many as 98,000 people in the United States die annually from medical errors and preventable injuries in hospitals. By 2022, this figure had risen to 227,000 deaths per annum, making preventable injury the third leading cause of death in the US, surpassing fatalities from stroke, diabetes, and liver disease. These statistics underscore the urgent need for systemic reform to address the persistent challenge of medical errors.
Achievement of the “quadruple aim” in healthcare (improved individual experience of healthcare; improved population health; decreased cost per capita of healthcare; and improved experience of providing healthcare) requires a systems-based approach which relies heavily on its management of information.
In Ireland, as in the US, similar concerns have been highlighted. The Irish National Adverse Events Study (INAES), which reviewed incidents in eight hospitals in 2009, reported an incidence of 10.3 adverse events per 100 admissions. These findings further emphasise the systemic flaws within healthcare systems that require a data-driven approach to patient safety.
Learning analytics (LA) is a relatively new discipline that has developed extensively within higher education over the past decade; it focuses both on the learning process of students and on monitoring their trajectories in training programmes.
Recognising this, the Forum of Postgraduate Medical Training Bodies (Forum of PGTBs), the Health Service Executive (HSE), the Department of Health, and the State Claims Agency agreed to collaborate to develop an NLA capability.
“The joint development of a national analytics capability would empower healthcare leaders in the HSE and policymakers in the Department of Health with actionable insights.”
Martin McCormack, Chief Executive Officer, and
George Shorten, Immediate Past President, College of Anaesthesiologists of Ireland
Grounded in the recommendations of the 2021 Forum of Postgraduate Medical Training Bodies Strategy, such a capability would apply analytics to deliver timely personalised feedback on clinical performance to healthcare professionals, better match training to specific risks and to design ‘just in time’ targeted education to where patterns of preventable harm are identified as recurrent. This collective effort marks a crucial step toward healthcare which is more consistently safe.
A unified approach to learning from errors
Both To Err is Human and the Forum of PGTBs’ strategy stress the importance of system-wide learning to prevent errors and improve outcomes. However, as Ireland builds new regional executive organisations (REOs) and implements new electronic health records (EHRs), healthcare data remains fragmented, limiting the ability to identify patterns and address systemic issues effectively.
In particular, the information systems around clinical care and those around training of healthcare professionals have each evolved dramatically in recent years, but separately. Ideally, a systems-based approach would access relevant data in an ethical manner and present the results of analysis in a timely and comprehensible manner to the person or persons who is both responsible and capable of responding to them. The response itself is then captured and forms an important data stream as the system matures and adapts.
An electronic healthcare record is a fundamental piece of digital infrastructure for a modern healthcare organisation. In many countries healthcare professionals have become familiar with and adept at interacting with an EHR as they provide clinical care. Frameworks have been proposed to integrate the EHR into education and training; such potential is currently underexploited.
A national learning analytics capability would harmonise the analysis of data available from clinical sources (such as an EHR), from training sources (such as a learning management system), and from registries of adverse events (such as INAES). This approach would leverage the insights provided into the transfer of learning to practice using newly developed data analytics tools This unified approach would in turn enable the implementation of preventative measures and support targeted interventions, addressing the call of the Forum of PGTBs’ strategy to reduce unintended harm.
Improving feedback on clinical performance
Accurate, actionable feedback is essential for continuous professional development and improving patient outcomes. However, feedback mechanisms in many healthcare settings are often inconsistent and reactive, limiting their effectiveness.
A national analytics capability would transform these processes, providing clinicians with real-time and ‘highly informative’ insights, while enhancing the learning efficacy of professionals. For example, detailed performance metrics for specific skills could be benchmarked against proficiency standards, enabling clinicians to identify strengths and areas for improvement.
Reducing unintended harm through education
Unintended harm often arises from systemic flaws or knowledge gaps that traditional training methods cannot adequately address. By analysing patterns in clinical errors and adverse events, the national learning analytics capability could inform the development of specialist training programmes.
For example, learning analytics might reveal the team characteristics, events, and circumstances which, when coincident, tend to predispose to greater risk. This in turn provides granular targets for training scenarios including using simulation.
This approach ensures that certain training efforts are focused where they are needed most, reducing harm and enhancing patient safety.
Enhancing system design for patient safety
The collaboration between the Forum of PGTBs, the HSE, the Department of Health, Science Foundation Ireland research centres, and the State Claims Agency reflects a shared commitment to addressing the systemic issues underlying medical errors.
This aligns closely with To Err is Human which emphasises that “good people are working in bad systems” and calls for structural changes to improve safety. A national learning analytics capability would help REOs to adopt a proactive approach to system design.
Core LA functions would include aggregation, interpretation, prediction, and visualisation using data of quite different types and sources. Predictive learning analytics could identify risk factor aggregates for such events as potential drug errors or communication failures, enabling highly targeted training interventions. This proactive stance is essential for building safer and more resilient healthcare systems.
Strengthening leadership and policy development
Leadership is critical to driving the patient safety agenda. However, effective leadership requires access to reliable, comprehensive data to guide decision-making and evaluate the impact of initiatives.
The joint development of a national analytics capability would empower healthcare leaders in the HSE and policymakers in the Department of Health with actionable insights. Aggregated data on clinical outcomes and patient safety could inform resource allocation, policy adjustments, and strategic planning. The State Claims Agency’s involvement further highlights the potential for analytics to mitigate risks and reduce litigation costs, reinforcing the value of this initiative.
Breaking down silos for system-wide learning
Healthcare in Ireland, as in many countries, often operates in silos, with limited communication and data sharing between organisations dealing with the same patient.
This fragmentation hampers efforts to learn from errors and disseminate best practices. Given its data science expertise, and the partnership approach being adopted (across state agencies, research centres, the HSE, and training bodies), Ireland may be particularly well positioned to deliver a nationally coordinated facility which achieves benefits for patients in years rather than decades.
This aligns well with the LA Action List proposed by the Science for Policy Reports Joint Research Centre “to create organisational structures to support the use of learning analytics and help educational leaders to implement these changes”.
The collaborative approach to developing a national analytics capability aims to break down silos, creating a shared platform for data collection, analysis, and learning. For instance, information on near-misses and adverse events could be pooled across REOs to identify national trends and inform consistent safety guidelines. Such integration would foster a culture of collaboration and continuous improvement, benefiting patients and providers alike.
Empowering healthcare professionals and patients
The national learning analytics initiative has the potential to empower both clinicians and patients. For clinicians, access to real-time analytics would support their professional development, improving confidence and competence.
For patients, increased transparency around safety performance would build trust and encourage active participation in their own care. Moreover, patient-reported outcomes and experiences could be integrated into the analytics framework, providing a more holistic view of healthcare quality.
This reflects the priorities of the national Patient Safety Office and National Patient Safety Council: advancing patient safety while enhancing the overall care experience.
Conclusion
The development of a national learning analytics capability represents a transformative opportunity for healthcare in Ireland. By uniting the efforts of the Forum of PGTBs, the HSE, the Department of Health, and the State Claims Agency, this initiative addresses the unavoidable conclusion that recurrent patterns of preventable harm occur much too often in Ireland and elsewhere and provides a means by which to delivered targeted interventions to lessen them.
Through centralised data collection, real-time personalised feedback, targeted education, and proactive system design, a national analytics infrastructure would not only serve to decrease the incidence of unintended harm, but also drive a culture of accountability and continuous improvement.
As the healthcare system in Ireland faces growing complexity and pressure, the ability to learn from data will be essential for delivering safer, more effective, and more equitable care. This collaborative effort is not just a step forward – it is a vital investment in the future of Irish healthcare.
