Exploring the use of routine healthcare data through process mining to inform the management of musculoskeletal diseases

Healthcare informatics can help address some of the challenges faced by both healthcare providers and patients. The medical domain is characterised by inherently complex and intricate issues, data can often be of poor quality and novel techniques are required. Process mining is a discipline that uses techniques to extract insights from event data, generated during the execution of processes. It has had good results in various branches of medical science but applications to musculoskeletal diseases remain largely unexplored.
This research commenced with a review of the healthcare and technical literature and applied a variety of process mining techniques in order to investigate approaches to the healthcare plans of patients with musculoskeletal conditions. The analysis involved three datasets from: 1) a private hospital in Boston, US, where data was used to create disease trajectory models. Results suggest the method may be of interest to healthcare researchers, as it enables a more rapid modelling and visualisation; 2) a mobile healthcare application for patients receiving physiotherapy in Sheffield, UK, where data was used to identify possible indicators for health outcomes. After evaluation of the results, it was found that the indicators identified may be down to chance; and 3) the population of Wales to explore knee pain surgery pathways. Results suggest that process mining is an effective technique.
This work demonstrates how routine healthcare data can be analysed using process mining techniques to provide insights that may benefit patients suffering with musculoskeletal conditions. This thesis explores how strict criteria for analysis can be performed. The work is intended to expand the breadth of process mining methods available to the data science community and has contributed by making recommendations for service utilisation within physiotherapy at Sheffield Hospital and helped to define a roadmap for a leading healthcare software company.