What Human Factors issues are contributing to unsafe primary care in England and Wales? A national analysis of 64,347 patient safety incident reports
Problem
Patient safety incident reports offer an important lens to evaluate how and why patients experience healthcare-associated harm. Learning captured from incident reporting systems can help inform strategies to mitigate future risks and improve safety. However, the quality of incident report content is highly variable. In response to this, the World Health Organization has highlighted a priority for the next decade to provide healthcare professionals and staff with training and support to identify the ‘human factors’ issues contributing to incidents. We aimed to establish a baseline assessment of reported human factors issues contributing to unsafe primary care outcomes. Using a human factors framework, we sought to identify which system-wide performance-influencing factors are least and most frequently reported. We then established priorities and recommendations for the international human factors research and training agendas.
Approach
We carried out a secondary analysis of coded data from the National Reporting and Learning System which receives incident reports from healthcare organisations in England and Wales. We included reports from the community that met the national definition of a ‘patient safety incident’ and have previously been coded using the ‘contributory factors framework’ in the Primary Care Patient Safety (PISA) Classification System. Two researchers independently mapped codes in the PISA contributory factors framework to a well-established human factors framework, the Systems Engineering Initiative for Patient Safety (SEIPS). SEIPS depicts a sociotechnical work system with six interacting components including person(s), tools and technology, tasks, physical environment, external and organisational factors. Exploratory descriptive statistical methods were used to summarise frequencies and cross-tabulate coded variables.
Findings
From 64,347 reports, less than one-third (31%, n=20,208) included one or more contributing factor. About half of these reports (54%, n=10,896) related to General Practice/Primary care; these were analysed, yielding 15,372 contributing factors. There was good inter-relater reliability between researchers mapping codes between the PISA and SEIPS frameworks (kappa coefficient of 0.82). Person-related factors are most frequently reported (n=11,398, 59.7%). These can relate to clinicians, patients or teams and explores their skills, communication or clinical condition, for example. The least mapped SEIPS components related to tools and technology (n=738, 3.9%, e.g. availability), physical environment (n=661, 3.5%, e.g. layout) and external influences (n=14, 0.1%, e.g. cultural or regulatory influences).
Consequences
Patient safety incidents occur for a complex range of interacting system-wide reasons. The over-emphasis on person-related factors in incident reports is concerning, indicating limited attention is being paid to the recommended ‘systems-approach’ to learning about safety occurrences and which potentially foments the blame culture in the NHS. Support is needed for reporters (clinicians and patients) to holistically consider the range of system-wide human factors issues likely contributing to patient safety incidents, including those relating to the usability of technology, and the influence of the physical, social-cultural and external environments.