Skip to main content

Exploring the promise of big data for medical error elimination

Periodic Reporting for period 2 - BigERRORS (Exploring the promise of big data for medical error elimination)

Reporting period: 2017-08-01 to 2018-07-31

This project is about achieving a better understanding of the phenomena of errors. Errors in organizations are essentially unintended and potentially avoidable deviations from organizationally-specified goals and standards. Errors are a recurring fact of organizational life and thus merit study in their own right as an important phenomenon. In particular, medical errors in the health care industry pose serious threats to human life and affect total national costs. An estimated 8-12% of patients admitted to hospital in the EU suffer from adverse events upon receiving health care.
I studied how big data can assist in improving our understanding of the error phenomena. So far, the error literature suggests a limited understanding of the origin of errors and of ways to eliminate them. I explored how big data can be used to eliminate errors by referring to aspects such as clinical workflow, information processing, health care decision-making, and the delivery of care. The use of big data technologies can reduce health care costs while improving its quality by making care more preventive and personalized. However, there is very little published management literature that tackles the challenges of using big data and explores the promise and opportunities for new theories and practices that big data might bring about. As in other fields and industries, big data has only recently begun to become a factor in health care. Big data calls for the interplay of error theory building and testing. This project was based on the promise that the next revolution in the science of eliminating and managing errors in organizations may come about by incorporating big data at the core of error theory and research. Taking a multidisciplinary approach that synergistically integrated managerial and health care perspectives with big data analytics and other technological aspects, the work I carried out during this research applied the novel concepts, approaches, and methods of error theory and big data techniques that include excessive data collection and new analytical methods. As a result of this study, I have extensively explored fundamental aspects of errors and suggested a better understanding of them.
Work performed in this study has been summarized in papers in leading journals, specifically Naveh and colleagues in Safety Science (2018), IEEE Transactions on Engineering Management (2017), and American Journal of Medical Quality (2018). Also, in the many presentations, professional development workshops, seminars, conferences presentations, new additions to courses, work with research students, and the creation of a new research community that elaborates on the topic of errors.

In this project, I showed how big data and machine learning approaches can assist in eliminating errors in hospitals. I designed and executed exploratory research to understand errors using big data as an initial test and the starting point of a roadmap to error elimination. I carried out this part of the research project with the Dana-Farber Cancer Institute in Boston, USA, one of the world’s leading cancer research institutes. The Dana-Farber Cancer Institute collects relevant structured data from measuring devices (sensors and other machine data), and also has unstructured data collected in media files.

This study will influence the area of errors in organizations and big data for years to come through the work that I did with regard to:
• Improving the theory behind medical error origins and consequences and the ability to predict errors and to eliminate and manage them.
• Providing meaningful practical insights for making hospitals safer based on data science.
• Suggesting a research roadmap for the realization of big data potential for dealing with the complex issue of medical treatment errors; and establishing a network for future study, development, and realization of big data’s potential for eliminating medical treatment

This study's deliverables and milestones refer to and make it possible to understand the promise of big data, explore its fundamental aspects, and contribute specifically to the understanding of:
• Simultaneously and mutually enabling of error elimination and learning from errors in organizations, and specifically in hospitals. Changing the paradigm of preventing errors from occurring while simultaneously investing in activities that “welcome” them as valuable
feedback. Prevailing practices in health care organizations convincingly argue for a ‘both-and’ approach to this goal, meaning that, in a perfect world, health care professionals may simultaneously eliminate and welcome (and thus learn from) errors. Yet health care
organizations that do well on both accounts are extraordinarily rare. We suggest an 'either-or' approach as a way to reduce errors. I believe that this is a significant aspect for better dealing with errors in organizations.
• Understanding the major role of standardization in error prevention and learning. I have shown that standardization is expected to reduce errors but not always does so. Employee choices explain the link between standardization and error reduction. Standardization
both weakens and strengthens employee choices.
• Identifying the role of human systems integration that explains how the technical and human components of a system interact to influence errors.
• Discovering the influence of operations management measures and methodologies on error occurrence. In this respect, we explored the influence of OM practices on error occurrence and also integrated operations management methodologies with the origination
research approach.
This research makes it possible to acquire and develop knowledge about big data to the end of implementing it on errors in organizational studies. Specifically, I have shown state-of-the-art directions to prevent and learn from errors by emphasizing the 'either-or' as well as the ‘both-and’ approach, explored new approaches to the role of standardization on error occurrence, and shown the importance of human systems integration and of operations management methodologies for error research. This progress is beyond state of the art because it suggests a new, partly counter-intuitive, theory, identifies research lacunae, and integrates different and contradictory lines of studies. I have broadened the current thinking on errors and built a knowledge resource that is more visible, inspectable, systematic, and influential. I have developed the groundwork for more innovative theorizing and a more effective practice aimed at preventing errors and at saving resources and lives.
I used empirical data, based on big data, that specifies the processes, mechanisms, and boundary conditions required to better understand various forms of errors and their consequences in organizations. I improved knowledge, brought the knowledge to the people that need it, and made a difference in dealing with errors in health care. The many presentations, professional development workshops, seminars, and conferences presentations make it possible to achieve communication and results dissemination. I developed and initiated a multidisciplinary international network of academia, health care industry, information technology companies, and governments across the EU and USA to develop and realize the big data and error roadmap.