Patient safety, present and future

Healthcare is undergoing an unprecedented trust and value crisis. It is grasping for quick and easy solutions and trends to oversimplify system safety concepts. Patients still experience needless harm and often struggle to have their voices heard, processes are not as efficient as they could be, and costs continue to rise at alarming rates while quality issues remain. We tend to think about patient safety in a linear dimension that is only associated with the progressive reduction in the number of errors and accidents, with the simple notion that more is always better. We consider figures in isolation from the underlying context, prerequisites of risk models and the affordances of clinical domains. There is no ultimate reference model for patient safety, but many models that can be adapted to fit the clinical requirements and constraints. It is therefore not necessarily a bad result to observe a lower safety figure in a medical domain compared to the figures obtained in non-medical ultra safe models. The poor safety figures may represent the best safety optimization while coping with the special, local health care requirements such as a high frequency of unplanned and non-standardized challenges. The paper distinguishes three classes of safety models that fit different field demands: the resilient (adaptive), the high reliability (HRO), and the ultra safe models. The paper benchmarks the traits of each model while highlighting the specific needs for optimization. Since no model is ideal, there are numerous general lessons going well beyond the specifics of health care and addressing other industries, namely nuclear industry. (author)