Tackling 'frequent flyers' won't solve the rising emergency hospital admissions problem

Patients who are regularly admitted to hospital as emergencies (known as 'frequent flyers') make up a large proportion of admissions, but focusing just on them won't solve the problem of rising admissions, say experts in the British Medical Journal today.

Martin Roland and Gary Abel from the Cambridge Centre for Health Services Research argue that this is one of several misconceptions about emergency admissions that must be tackled if we are to reduce the number of people being admitted as emergencies. Around the world, the pressure to reduce healthcare costs is huge. Emergency hospital admissions are an expensive aspect of care and rates have been rising for several years, particularly among the elderly and those with several conditions (co-morbidities).

In the UK, many initiatives have been set up to reduce emergency admissions, mainly in primary care and with a focus on high risk patients who are thought to use a disproportionate share of resources.

But Roland and Abel argue that there are "some fundamental flaws" in this approach. Exclusively focusing on high risk patients won't solve the problem, they say, as data show that most admissions come from low and medium risk groups. Instead they suggest interventions may need to be targeted on larger population groups, such as elderly patients.

They also challenge the widespread view that improving primary care could prevent many emergency admissions and suggest that some of the rise in admissions may be due to the introduction of four hour waiting targets in A&E. They say that primary and secondary care doctors need to work together to achieve a common set of goals.

They also point to the problem of "supply induced demand" for services which could explain the apparent increase in admissions found in some studies of intensive case management (e.g. community matrons).

Apart from a few exceptions, evaluations of interventions to reduce emergency admissions have been disappointing. Evaluations need to be evidence based, they need to allow for admission rates for individual patients falling of their own accord (a phenomenon known as regression to the mean) and to take account of variation due to chance. The authors suggest some guidelines for those needing to focus on this important and expensive aspect of healthcare