Patient safety in primary and emergency care

Abstract

This thesis explores aspects of risk and safety in health care. Patient safety is about preventing harm to patients. A perspective of resilience is used, which is a proactive approach to making care safer. Resilient organisations recognise the fact that work is done in a complex and adaptive system that can be improved, not only by studying what goes wrong, but even more so by studying what works well. The thesis adds to previous research by studying patient safety in first-contact care, primary health care and the emergency department, resulting in new knowledge that could be used to make care safer.

We investigated reported preventable harm and serious safety incidents in primary health care and in emergency departments (Study I). In these contexts, diagnostic error was the most common type. A diagnostic error is when a patient does not receive the correct diagnosis within a reasonable timeframe when there were clear opportunities to establish the correct diagnosis. The consequence is delayed adequate treatment. There is limited knowledge of the panorama of diagnoses that are involved in diagnostic errors. We explored the diagnoses that were most frequent in this material. Cancer was the most common missed diagnosis in primary health care, in particular colorectal cancer, and fractures were the most common missed diagnoses in the emergency departments. Furthermore, little is known about patient-related risk factors for preventable harm. We explored factors associated with an increased risk of reported preventable harm, or serious safety incidents, in primary health care and in the emergency departments (Study II). The most prominent risk was psychiatric disease. This was, to our knowledge, the first study in an out-patient setting, with all types of psychiatric diseases.

This thesis also examines what patients and health care professionals perceived as the major risks in primary health care and what solutions they would prefer (Study III and IV). Solutions generated from the people in the system are likely to be more accurate and easier to implement than top-down solutions. In Study III, qualitative analysis of free-text answers to structured questionnaires was performed. The results were used to build a survey for Study IV, where specified risks and solutions were rated according to importance. Regarding risks, the areas that were thought to need most improvement were continuity of care, communication and knowledge. Solutions included: information about what to do when tests were fine, but symptoms remained, so called safety-netting; the use of a nationwide medication record online; and a personal doctor with a restricted number of patients per doctor, to facilitate continuity of care.

These studies support future work for safer and more resilient health care. There were suggestions from the level of the general practice up to the national level. Further research should test interventions that proactively support systems in improved accuracy in diagnosis and correct medication, for example, an intervention to improve continuity of care or to practice safety-netting. Proactive interventions like these could probably improve the resilience of the system in question.