The pandemic has wreaked havoc on your care requirement forecasts, but how can you fix it? Do you throw them all away and wait for calmer waters or can they still be put to good use?
Our Demand Modelling tool helps to predict social care activities over time, and due to the huge variety of things that influence activity, we decided against trying to build a hugely complex model. Complex models often attempt to cater to all factors like population changes, hospital visits, weather, and other things that are harder to quantify like commissioning choices.
Instead, we decided to see how far we could get with simple time series forecasting (TSF), which is a statistical approach to projection based on patterns that are picked up from past activity.
The Demand Model automatically detects the most appropriate model for each care category which optimises forecast accuracy and prior to the pandemic we were getting some great results, predicting the activity of major care categories to less than 2% variance from the actual one year into the future.
Here is a typical forecast from a stable care category (Long-term nursing care for people 65+ across all primary support reasons). The black line is the historic activity which shows a slight seasonal pressure that has been picked up by the forecast which was made at the start of 19/20 and half-way through the year is tracking exactly as expected.
Throughout 2020, most care categories were severely affected by COVID and trends were impossible to predict, but by using forecasts taken immediately prior to the onset of the pandemic the Demand Model was able to measure the impact of COVID on individual care categories.
COVID impact in Long-Term Nursing (red line indicates March 2020 when activity was first affected)
Once the COVID impact had been measured, it was crucial that the Demand Model could return to providing accurate forecasts to help users monitor and manage the “new normal”. This is not a straightforward task using time series forecasting, as the historic activity for most care categories has been seriously disrupted throughout 2020. By analysing the data across multiple LAs, we were able to isolate the period of “COVID shock” and essentially erase it from the view of the models so that they didn’t disrupt future forecasts.
Here you can see an example of Home Care Hours which increased significantly throughout the COVID shock period (marked with red bars). The category has settled at 10% above pre-COVID levels but the post-COVID forecast is tracking on a near-identical trajectory (even displaying similar seasonality) with a high level of accuracy.
Pre and post-pandemic forecasts for Home Care Hours (red lines highlight COVID-shock period)
COVID has radically altered the landscape of social care, but through intelligent use of data and continuous measuring and monitoring of activity at a granular level, you can get a deep understanding of changes that supports effective planning and management strategies.
For more information about Demand Model download our 'Guide to utilising health and social care data to predict demand'