Public concern about mass media pressure related to surgical waiting lists usually leads hospital managers to give priority to the admission of programmed surgical procedures over both medical and surgical urgent admissions.1 2 Cancellation of elective admissions is only considered in situations of serious overcrowding in the emergency department (ED). Access block is defined as the inability of patients in the ED to be admitted to hospital beds.3 A report from the US Government Accounting Office cites holding admitted patients as the single greatest factor contributing to ED overcrowding; they take up beds in the ED, so reducing the resources available to assist other patients needing emergency care.4 As an isolated event, access block is not worrying but, in hospitals with a sustained high attendance pressure in the ED, it may lead workers to burn out. Access block has been reported as the most stressful aspect of work by emergency physicians.5 Reducing access block should be the highest priority in allocating resources to reduce ED overcrowding.6 The accumulation of patients in the ED waiting for a bed worsens the delay in emergency attention and increases the length of stay in the ED and the number of patients waiting to be visited. The patients and their relatives fill the corridors and waiting rooms leading to a perception of chaos.

Our hospital is a 900-bed teaching institution with a reference population of more than 1 million. A mean of 300 patients visit the ED daily and a mean of 30 (about 10% of ED visits) are admitted to in-hospital units, in addition to the mean of 40 elective admissions per day.

In recent years the measures directed to decrease the hospital length of stay, although effective, have been clearly insufficient to improve this situation. In October 2002 we opened a holding unit (HU) containing 16 beds. These 16 beds came from the less efficient services in such a way that the total number of beds in the hospital did not increase. Only patients coming from the ED who are admitted to hospital by a staff physician are placed in the HU. Three conditions are strictly observed: (1) no patients with an elective admission are placed in the HU; (2) every morning all the patients are moved from the HU to a conventional unit; and (3) if there are not enough unoccupied beds to move patients on from the HU, elective admissions are cancelled. In this way the hospital ensures that the ED has a daily capability of at least 16 beds for urgent patients who require hospital admission. Here we present the results of the assessment of the efficacy of our HU.

METHODS

All ED admissions are recorded in real time in the database of the hospital as well as the final destination of every patient (discharge or hospital admission). Monthly data validations are routinely carried out. Data for the study were extracted from this database, which remained unchanged during the period of interest.

Some key medical management indicators of the first 12 months after the opening of the HU have been compared with those of the same period in the previous year (without the HU). We analysed (1) the number of patients visiting the ED; (2) the number and percentage of urgent admissions; (3) the mean length of stay in the ED; (4) the number and percentage of patients waiting for an in-hospital bed in the ED at 08.00 h; (5) the number of elective admissions; and (6) the number and percentage of cancelled elective admissions. Statistical analysis included a descriptive comparison, the Student t test and the Pearson correlation test. The level of significance was set at p<0.05.

RESULTS

The first year following the opening of the HU, even though 3.1% more patients visited the ED and the number and percentage of urgent admissions was maintained, the mean number of patients waiting for a bed at 08.00 h decreased significantly by 55.6% (from 9.1 to 4.0; mean difference −5.1 (95% confidence interval (CI) −5.9 to −4.3)). There were 176 days during the year following the opening of the HU in which there were <3 patients waiting in the ED compared with only 60 days in the previous year. Table 1 shows the comparative data of the 1-year period just before and after the opening of the HU.

Although the length of stay in the ED correlated significantly with the number of patients attending, the number and percentage of urgent admissions and the number of patients waiting for a bed at 08.00 h in the ED (p<0.01), the improvement achieved during the HU period was not reflected in the mean length of stay in the ED which increased by 6.9% (from 3.89 h to 4.16 h).

DISCUSSION

There is no one single cause of access block, but it appears to correlate with major decreases in hospital bed capacity and community geriatric care facilities along with changes in workforce and community attitudes.7 It represents not only a problem in terms of time; when no other answer is found, patients waiting in the ED for >24 h may be reassessed as “borderline” admissions and are at risk of being discharged in a condition that would not be considered suitable for discharge from a ward bed.8 Furthermore, access block is no longer a problem restricted to the ED, since length of stay in the ED has been shown to be associated with inpatient length of stay after adjusting for case mix, time of presentation and patient age.9 10 In our experience, the opening of an HU has been related to an improvement in access block.

The opening of the HU has also resulted in adverse effects. Patients are moved to another bed twice in the first 24 h of admission, HU beds are not generally ready to use again until late and patients still wait for hours to be moved to them.

Against the current evidence,11 after improving access block with the opening of the HU, in our hospital we have failed to decrease the length of stay in the ED but this may be the effect of the increasing number of patients visiting the ED together with the other factors mentioned above. After this first year, we consider the creation of an HU as one of the interventions that may be undertaken to improve ED outcomes.