We used data for actual A&E attendances and admissions over the past five years across the country and tried to find a statistical model that would explain what was driving this activity on any given day. This analysis showed consistent and reliable patterns in the impact of different days of the week and bank holidays throughout the year.
Many of these patterns, such as high admissions in the winter, will not surprise anyone. But our analysis adds value in revealing how these patterns interact with each other. Where the various drivers of A&E activity push in opposing directions, our findings indicate which effect is likely to prevail.
Our analysis is based on a regression model and uses daily data on attendances and admissions for the past five years. The model allows us to predict A&E activity – attendances and admissions – on a daily basis. The calendar we prepared, shows our 2015-16 results for England. Local forecasts at CCG level can also be produced in Frontier’s model, but with a greater degree of uncertainty. This is something that we plan to work on in the future.
Just a perfect day
We present the data against an illustrative “average day”. This is no day in particular, though there may be such perfectly average days. We calculate these benchmark days as all the activity that happened in a year divided by all days on which it happens (i.e. the 365 days of the year or 366 in the case of a leap year like 2016). As a result, around 50% of days are expected to be above this benchmark and 50% below.