SiMSK: Project aims (Page in development)
Project Aims
There are a wide range of data collected on healthcare systems, from patient registers and national databases to local hospital attendance and treatment data. Unfortunately these are often only used for a single purpose and the information in this data is not widely used for improving patient care. This modelling approach offers a platform to combine several data sources into one model improving the real world validity of the model. Data is simply uploaded into the model (in the specified format) and the model analyses and implements it.
In the Rheumatology model we use local hospital attendance data to model the frequency of new patients, their clinic attendances and discharge or longer term follow up. The most important aspect of this model is how each patient’s long term condition develops over time and in response to different treatments. This is modelled using a probabilistic method informed by several data sources, the BSR-BR, HQIP audit data and NOAR. It is also possible input trials data or hypothetical patient responses to explore their effects in a real world setting.
The agent based modelling approach means that each real world actor can be modelled individually as an agent. This means each patient is modelled individually and their each treatment journey is unique, based on their timings and responses, all informed by the data.
The model is initiated based on the input data. This creates a population of patients in various stages of follow up, a group of clinicians and nurses, a waiting list and annual clinic timetable. The model calculates default values for these, but different experimental values can be explored.
The model then cycles through a working day for the required time period. New patients are referred from GPS, patients try and book their appointments, patients attend clinic and their diseases diagnosed or treatments reviewed and updated, patient's diseases progress over time and in response to their treatment, socioeconomic status and history. Daily costs for staff, facilities, administration and drugs are calculated. The working day is then repeated. The standard settings give us a baseline model which was validated against the data.
The aim of the model is to improve the existing healthcare systems in terms of reducing costs and improving patient outcomes. To achieve this model can be varied in any number of ways and the outcomes compared to the baseline. These findings can then be used by the Department, consultants, CCGs and hospital planners to have an improved understanding of their options for improving the system and the effects of these decisions. It should be noted that the results are a computational prediction, informed by data, of how events are likely to occur at the patient population level. They should not be taken as accurate prediction of individual events.