Objectives & Background The increase in emergency admissions over the past 15 years has come almost entirely from the EDs,for patients who spend two days or less in hospital. It has been suggested that these cases are of low severity and admitted because of a lack of alternative services, defensive medicine and the four hour target. We wished to develop means of classification as to whether the admission was jusitified or not using both explicit critiera and implicit expert approaches.

Methods Explicit approach: we wished to use the activities recorded in the patients medical and nursing notes covering their time in the AMU relating to observations (type, frequency and values), investigations (type and values), treatments (type and frequency), specialist review (specialty and grade) and post discharge care to generate a list of ‘candidate’ criteria. Candidate criteria were developed after review of 100 sets of notes.Three consultants in Acute Medicine were asked for each criterion –“would its presence justify admission?” and a majority verdict taken. The list of criteria was then that applied to 300 consecutie patients admitted to the Acute medical Unit. Unjustified cases would be those with the absence of any justifiable critieria.

Implicit approach: three different consultants in Acute Medicine were given copies of 120 randomly chosen patients anonymised notes and independently asked to judge if the patients admission was jusitified or not and a majority decision was taken.

Results Between January and August 2014, 24% of patients attending the ED ( 13766), were admitted to the hospital. 5802 of these were admitted to AMU of which 1063(18%) stayed less than 24 hours ( approximately 150 /month).

We studied 300 consecutive patients. Table 1 (screen shot) shows the criteria agreed by the expert panel,whose presence would justify admission together with their frequencies. Patients could meet more than one criteria. 275 met one or more of the AMU criteria (92%).Of the 120 cases reviewed by experts, 84 (70%) were considered justified and 36 unjustified.

Conclusion Using the explicit criteria 92% of cases were justified, and 70% by use of the expert panel note review. This does not suggest an excessive admission rate for the population that we have described, who can be both difficult to assess and to rapidly exclude serious pathology.