Obstructive sleep apnoea: Improving healthcare services by combining process modelling and population analysis

Context: Disease management broke through in the early 1990s to counterbalance hyper-specialization with a more comprehensive approach. Its role became immediately relevant in chronic conditions and, consequently, in Obstructive Sleep Apnoea (OSA). This is a common chronic condition for which is important to organise services at the local level, taking into account organisational factors and the characteristics of the assisted population. Objectives: The aim of this work is to propose and apply, coherently with a disease management approach, a combination of healthcare process modelling and population analysis as a way to identify critical issues and explore shared solutions. Methods: A multidisciplinary working group was created with scholars who are skilled in process analysis, statistics and medicine. Through semi-structured interviews and on-site meetings, healthcare processes were represented with a standard graphical language: Unified Modeling Language (TM) . Population analysis was based on statistical analysis performed on a 5-year retrospective cohort assisted by a Community Pulmonary Service. Results: A shared graphic presentation of the current healthcare process and the results of the statistical analyses constituted the knowledge base to identify critical issues and recommend corresponding solutions, which include: a) refine the local patient database with additional details on comorbidities and risk factors; b) support a greater involvement of "gate-keepers" in the screening phase; c) provide practical tools for the definition of strategies to increment the adherence to therapy; d) include recommendations for physical exercise and interdisciplinary cooperation; and e) define process indicators for measuring the quality of the screening and therapeutic phases. Conclusion: The concomitant analyses of formalised processes and critical risk factors represent a useful approach for systematically identifying areas of improvement in healthcare processes and allow us to discuss solutions. Moreover, the specific adoption of UML? for graphical modelling and representation of patient care processes allows us to formalise them by adopting a standard language that can be taken as the basis for implementing web services to support the execution of the modelled processes.