Thankfully, the long-term trend in delayed transfers of care has been reducing. Official statistics show that in 2008 on average just under 5,330 patients per day faced a delay in any one month. By 2012, this had dropped to just under 3,990 – a decrease of 5 per cent compared to the year before.
However, many of the NHS finance directors we surveyed last December for our quarterly monitoring report said that delayed transfers of care were a key concern for their organisation. Asked directly if delays had got worse over the last year, 63 per cent said they had.
So what might explain this contrast between the official statistics on patient delays and finance directors' views?
One answer could be that our survey is not representative of finance directors as a whole. Yet complementing our survey is one from the NHS Confederation which found that, of the chief executives and chairs who reported an impact on their services over the last year due to a shortfall in local authority spending, 92 per cent reported an increase in delayed discharges from hospital.
It would appear that, representative of the whole service or not, these surveys are picking up some concern on this particular issue.
Another explanation for the fact that the service-level view of delayed transfers doesn’t correspond to the official statistics is that the official statistics don’t take into account regional variation. National figures can mask wide variation at local levels. For example, a Health Service Journal analysis of Department of Health data on delayed transfers found that half of all delays were recorded in a quarter of local authority areas.
Furthermore, 12 local authority areas (out of 154) had rates of delayed days more than double the national median.
More interpretation of the data is required, but we are not helped by the way in which the data is collected. There are two main metrics in the official delayed transfers data: the number of patients delayed and the total number of days delayed. But these two metrics are measured using different methodologies.
The number of patients delayed is counted using a census on the last Thursday of every month – meaning it only counts patients delayed at that one point in time – whereas the number of delayed days is a total of all days that patients were delayed in the month, including patients that were not counted in the census.
It would be more useful to have a record of the number of delayed days for each patient across the whole month. This data is likely to be collected locally and producing these numbers nationally would allow us to see where bottlenecks are (in which regions and where in the system). The numbers could also be used for regional profiling, benchmarking and highlighting best practice.
We also don't know which patients are facing delays in their care but, again, this would be useful information to have. More information on the patients' characteristics, such as their age and details of their conditions and intended management of care would help us identify what needs are not being met.
With hospitals facing increasing pressure on beds and a fresh focus on quality of services – while purse strings are tightened even further – it seems certain that delayed discharges will remain an important issue facing hospitals.
Our quarterly monitoring report calls for further investigation into the dissonance between the official statistics and the views of finance directors. Until we can add more insight and evidence, it's likely the disparity – not to mention pressure on the system – will remain.