Forecasting demand improves health delivery

WA Health needed accurate forecasts of hospital demand for planning and resource allocation, as well as to underpin strategic planning for the state.

The healthcare system in Western Australia, just like that of the other states and territories, is experiencing a growing number of patient admissions and increasing emergency department (ED) overcrowding. To help mitigate this, the Western Australia Department of Health undertakes modelling and forecasting of statewide hospital activities — inpatient, ED and outpatient — for the purposes of strategic planning and reform.

Overcrowding in EDs indicates that hospitals are unable to adequately cope with demand. Thus, the ability to accurately forecast future demand enables sustainable strategies to be implemented to allocate resources more effectively and efficiently. This ultimately leads to improved safety and quality of care for patients and staff, and to sustainability of the healthcare system.

The Department of Health’s Dr Bella Mai is part of the team responsible for modelling and forecasting hospital activities in Western Australia. The team’s analysis takes into account key influencing factors in the modelling, such as patient age, place of residence, admission type (emergency or elective), stay type (same-day or multiple days), major disease category, ED discharge disposition and ED triage category.

“It is important to determine future health services demand and influencing factors so we can develop strategies to ensure we operate most effectively and efficiently to avoid inefficiency-associated costs or strains on patient care,” said Dr Mai.

The department provides forecasts for all public hospitals in the state.

“Based on our modelling, clinical planners can plan for the demand by arranging the required infrastructure or ensuring they have the right staff, technology and facilities in place,” said Dr Mai.

The ED modelling found that, for example, if it is ‘business as usual’, ED demand would grow by around 5% per year over the next five years. However, Dr Mai says that while predicting ED demand is necessary for strategic planning, as with forecasting in general, the predictions are viewed as short-term forecasts.

“Our models and forecasts need to be updated regularly as more data becomes available and as the situation changes,” she said. “For example, there may be changes to health policy and practice or there may be external factors to consider, such as the effects of the mining boom ending. These changes may be considered for the scenario modelling.

“We can either include data translated from expert opinions or conduct a comparative scenario model comparing WA activity with other states,” Dr Mai said.

In order to improve its modelling and forecasting capabilities, the department has implemented an analytics solution from SAS called Forecasting for Desktop.

“Previously the department had to outsource the services of demand modelling, which was costly and time-consuming,” she said. “But now that we are able to use advanced analytics and forecasting tools ourselves, we can perform all kinds of modelling work for statewide strategic planning and specific health services strategic planning.

“Ultimately, health services planners can more accurately match the services they provide with the expected demand.”

The department also produces capacity modelling to optimally redistribute the demand to the supply. The results assist the state’s hospital services to reconfigure reflecting the reform direction — to provide the right care, at the right time, n the right place. They are also helping in the planning of new hospitals, in order to estimate potential future demand and resource allocation before the hospital is built.