Managing variation in hospital patient flows

Abstract: Performing healthcare services is a complex practice. With the continuous developments in the medical field, healthcare organisations and professionals have become increasingly specialised, and healthcare services increasingly fragmented. When focusing on how to treat and cure diseases, the organisational and managerial aspects of care delivery can be overlooked. This increases the risk of sub-optimisation, when focusing on single activities, patient flows or healthcare units without considering the impact on the organisation as a whole. With every individual patient also having different combinations of diseases, severity levels and responses to medical interventions, much variation is introduced into patient flows. To address these challenges, the purpose of this thesis is: To explore the impact and management of variation in patient flows within hospitals. This thesis reports the results from three different studies, performed at three different hospitals. The first study analyses the impact of changes in patient inflow, available resources, treatment times and duration of activities on bed occupancy and length of stay. It does so by developing a simulation model of an emergency department (ED) and inpatient ward. The results show that the impact of variation in acute patient flows is smaller for the ED than the inpatient ward, which indicates that EDs might be more adaptable than inpatient wards in handling variation in acute patient flows. Bed occupancy on inpatient wards also seems to be strongly influenced by variation in patient flows. The second study is based on a single case study of the management of variation in acute patient flows at a university hospital. In accordance with the results of the first study, this second study shows that many actions are used to handle high bed occupancy, while few are used to handle ED overcrowding. More actions are also used to improve the situation in clinical departments than in the hospital as a whole, increasing the risk of sub-optimisation. The study highlights the need for top management to support department chiefs in managing variation in patient flows, e.g. by establishing a hospital-wide strategy for handling variation in acute patient flows. In the third study, the combined results from a case study and a quantitative before/after study are used to analyse whether the implementation of standardised cancer care pathways leads to crowding out effects, in terms of longer waiting times for other elective patient groups. The study shows that care pathway implementation is likely to lead to crowding out effects, if these effects are not addressed in the implementation process. Hence, a strategy to handle variation in one patient flow influences the options for handling variation in other patient flows and at the healthcare units as a whole. Crowding out effects were evaluated as most likely for patients from the patient group that is overtaken by the care pathway patients, and least likely for patients from the patient group with the lowest priority, due to a redistribution of waiting times between different patient groups.  All in all, the research reported in this thesis demonstrates the need to take a systems perspective when managing variation in patient flows in order to avoid sub-optimisation, which otherwise often follows when focusing on just one part of a system.