Myths versus facts in emergency department overcrowding and hospital access block

The single most important factor affecting ED overcrowding is the availability of sufficient inpatient beds.1,8 The number of acute-care public hospital beds in Australia per head of population decreased by 18% between the financial years 1995–96 and 2005–06,9 after even larger reductions in the preceding two decades. Although these large decreases were initially ameliorated by more efficient hospital function and shorter admissions, it is clear that in most major hospitals, occupancy levels have been driven up routinely to over 90%. These levels are known to be associated with poor hospital and system function, inefficient care and increased episodes of hospital and ED overcrowding.8 In fact, it is a universal finding that operating at full capacity for prolonged periods is inefficient and unsustainable. ED overcrowding is best seen as a marker of whole-of-hospital dysfunction that requires a whole-of-hospital or whole-of-system response.10,11 Bed availability depends not only on the physical number of beds, but also on the way the bedstock is managed (appropriate use, good flow practices),11 competing uses for beds (eg, elective versus acute care), the availability of step-down units, and appropriate community care.

Importantly, there is no evidence for the often-proposed myth of “general-practice-type” or “inappropriate” patients leading to ED overcrowding.12-14 Discretionary presentations by patients with low-complexity conditions, who might reasonably be managed elsewhere, constitute an insignificant workload in most EDs. These patients are uncommon in major EDs, and the most frequent reason for them to attend an ED is because they were referred by a GP. They rarely require admission or even use of trolleys, they use minimal ED resources (less than 3% of all costs or resources in most EDs), are easy to deal with, and do not impose on the key functions of the ED (assessment of sick patients, complex treatments and resuscitation).15,16 They may attend an ED because no other options are available and, importantly, they often feel their medical needs are urgent. They may spend a lot of time in waiting rooms, but this does not affect overall ED function. This myth is particularly problematic in that, if allowed to continue to be given credence, it continually diverts attention to solutions that cannot deal with the key issue causing dysfunction in the ED — that is, excessive numbers of admitted patients.

At 10:00 on 2 June and 1 September 2008, two national studies funded by the ACEM examined the point prevalence of overcrowding and access block in accredited Australian EDs. As shown in Box 1, over half the patients under treatment in non-New South Wales tertiary hospitals were waiting for beds, and 75% of those had been in the ED for more than 8 hours. Despite variation between states and hospital types (partly reflecting the size of different EDs), on average, the number of people waiting to be seen was less than the number who were under treatment and experiencing access block. If the patients experiencing access block were moved to appropriate inpatient beds, there would be more than enough staff and physical resources to treat all those waiting to be seen. The 2008 point prevalence study confirms data from previous similar studies and hospital time series17,18 showing that the problem is ubiquitous, sustained, and getting worse.

Development of new diagnostic approaches and therapies has contributed to increases in the “assessment” time in some groups of patients. Chest pain “rule-out” protocols using delayed marker measurements, and increased use of computed tomography (CT) scans for conditions such as abdominal pain are two examples. These are partly balanced by shorter, protocol-driven care in other conditions; for example, routine CT for minor head injury with immediate discharge if the result is normal, rather than observation. However, both effects are small compared with access block, as highlighted by data showing that during the time ED overcrowding has become persistent and prolonged with dramatic increases in ED stays, assessment times have increased only minimally (Box 2).

Telephone advice services are popular with patients and may decrease the demand for after-hours GP consultations in some settings,19,20 but have not been shown to reduce ED workload in Australasia.21,22 Establishment of additional GP services, even bulk-billing services near hospitals, does not significantly decrease ED workload either in theory12 or in practice.23

Hospitals that provide a local service in areas in which the population is undergoing significant demographic change, such as a large ageing cohort or rapid population growth, may experience ED overcrowding simply because the numbers of patients presenting exceed appropriate ED changes and resources. However, this is rarely the major issue and is rarely reported as the main cause of ED overcrowding. Indeed, studies24 and local experiences have shown that even with very large increases in ED size (eg, doubling cubicle numbers) or function, there is more access block and overcrowding. The most plausible explanation for this is that the extra capacity in the ED allows more patients to be accommodated there awaiting inpatient beds while bed capacity in the rest of the hospital (which has not been addressed) remains unchanged.

EDs have an obligation to minimise the effects of overcrowding, but any reduction will be largely achieved through whole-of-hospital changes. Long time series show that, in the absence of hospital-wide changes, access block tends to continue to increase even after mitigation efforts within EDs.17,18

Increases in the number and seniority of ED staff are associated with improvements in process measures and analysis of flow, and system redesign can allow better use of existing resources — for example, through dedicated “fast-track” areas for patients with low-complexity conditions. Such areas improve flow, but make only a small contribution to reducing occupancy by patients under treatment, and hence their effect on overcrowding is also small. Observation medicine within the ED is a useful adjunct or alternative to formal inpatient admission, and multidisciplinary assessment and discharge is effective at reducing representation, at least among older people. However, none of these responses can be used indefinitely if access block keeps increasing.

Adverse effects of hospital overcrowding have been described since the birth of modern medicine. In Australasia, access block was first shown to be associated with decreased ED function in 2000,25 and was defined by the ACEM in 2002.3 Worldwide, rigorously conducted studies in different centres have found an association between overcrowding and reduced access to care, decreased quality measures, and poor outcomes.26

Most important are studies, including a number from Australia, that show an increased mortality rate with increasing overcrowding.27-29 Two independent studies from different centres found about a 30% increase in overall early mortality if patients are admitted through overcrowded EDs to overcrowded hospitals.27,28 Annual excess deaths amounted to 13 in one of two hospitals in the Australian Capital Territory29 (13/(0.5 × 320 000 [population of ACT]) = ~ 8/100 000) and 120 among the three major hospitals in Perth28 (120/1.4 million [population of greater Perth] = ~ 8/100 000). Thus, a rate of 8/100 000 is conservatively estimated to correspond to 1500 additional deaths a year in Australia (similar to the road toll29), based on 2003 data. Access block has worsened since.10

Other serious problems associated with ED overcrowding include: delayed time-critical interventions and increased complications (eg, more recurrent myocardial infarctions or episodes of congestive cardiac failure); increased medical error; less frequent and less adequate pain control; prolonged hospital length of stay, further worsening the underlying cause and increased costs per stay; more legal actions and complaints; and severe problems with staff turnover and burnout. ED overcrowding causes ambulance diversion and delays to offloading ambulance patients,30,31 with both financial and service impact on prehospital care. The opportunity cost of lost ambulance resources has not been quantified but is also likely to be high.

Associations between overcrowding and outcomes demonstrated in peer-reviewed studies are shown in Box 3. While there may be some publication bias, there remain no published studies associating overcrowding with improvements in care and very few where it has no effect on care. The association between overcrowding and poor outcomes is accepted to be causative by most medical authorities and all emergency medicine colleges.

It is important to realise that ED overcrowding has serious adverse effects, because there is little incentive to fix a problem if it is not shown to have serious consequences. Routine denials in the media by politicians and bureaucrats that overcrowding has major adverse effects on patient care or outcomes shows that they do not see overcrowding as having serious consequences.