Increasing hospital admission rates and problems with bed availability are of public, economic and healthcare interest.1 2 The rising rate of emergency admissions to hospital is partly the result of an increasing population with an increasing number of visits (fig 1) and of advances in health care with improved disease recognition and management. The growth in the acute admission rate is a medical rather than a surgical issue.2 3 Increased hospital occupancy results in an increased length of stay in the emergency department (ED) for admitted patients, which is a major cause of overcrowding in the ED.4

• Download figure
• Open in new tab
• Download powerpoint

Figure 1 Numbers of emergency department medical visits since 1990.

Adan Hospital is a governmental general hospital in Kuwait established in 1981 and now serves a population of 700 000. The ED receives about 200 000 cases each year, a number that does not include maternity and paediatric cases which are managed in separate casualty departments. The primary healthcare service is provided through a number of polyclinics of which only 20% are working round the clock, so the hospital ED offers this service for 80% of the area overnight. With a progressive increase in the number of visits to the ED and inpatient hospital occupancy of 100% on most days, we have an almost daily bed crisis. Other general hospitals in Kuwait are facing the same problems due to similar circumstances. Attempts to overcome the excess load and overcrowding by reducing hospital referrals and improving the quality of the primary healthcare service are ongoing through educating general practitioners and family medicine staff.

In our ED we have introduced a new policy aimed at reducing hospital medical admissions. We assigned an admission avoidance team of ED doctors to be responsible for admission/discharge decisions. Doctors in the team implemented validated disease management guidelines and protocols. We also maximised the use of the emergency department observation unit (EDOU) by: (1) using it as a short stay observation ward for common medical emergencies (eg, chest pain observation unit); (2) extending the length of stay; and (3) allowing its use for certain procedures such as blood transfusions and outpatient antibiotic therapy. We observed the change in the total medical admission rate and admission rates for selected common medical conditions for a 3-year period after institution of this policy.

METHODS

A change in ED working policy was started on 1 January 2004 which consisted of (1) assigning an admission avoidance team; (2) implementing disease management guidelines; and (3) maximising the use of the EDOU.

Assigning an admission avoidance team

The working hours in the ED are divided into three daily shifts; each shift is covered by 5 doctors, 20 nurses and 1 social worker. The most senior 2 doctors in each shift were assigned to the admission avoidance team, which was therefore composed of 6 doctors per day. The chosen doctors were senior registrars or experienced registrars, most of whom had MRCP or FRCS A&E degrees. The admission/discharge decisions for medical EDOU cases were the responsibility of the doctors in the team. No special training or instructions were given to the nurses or to the social worker in this policy.

Implementing disease management guidelines

Validated disease management guidelines were chosen and approved by a panel of senior casualty and internal medicine doctors. To guarantee good quality disease management with safe decision-making for patients and doctors, international validated guidelines and protocols which have been tested and shown to be reliable and cost-effective were chosen.5^–11 Selected guidelines for the medical conditions used in this study are shown in table 1. The conditions studied were chest pain, bronchial asthma, heart failure, pneumonia and pyelonephritis. A guideline can be updated or changed whenever necessary or applicable. In our weekly regular scientific meetings, these guidelines were addressed to ED doctors to follow in their practice.

Maximising the use of the EDOU

The EDOU is composed of 14 male and 14 female beds. EDOU medical cases are not considered hospital admissions unless the patient is admitted to the ward under the care of the medical department. Maximising the use of the EDOU was undertaken by: (1) extending the length of patient stay up to a maximum of 48 h; (2) allowing its use as a specialised short-stay observation unit—for example, for chest pain (4 beds) or heart failure (2 beds)—with the necessary monitoring and access to investigations; (3) allowing certain procedures such as red cell transfusion for chronic anaemia12 and outpatient parenteral antibiotic therapy;5 13 (4) increasing access to immediate diagnostic testing such as Doppler ultrasonography, ventilation/perfusion lung scanning, CT scanning, ultrasonography and exercise ECG; and (5) using it for follow-up of non-admitted patients if needed.

The numbers of medical visits, EDOU medical cases and medical admissions (total and diagnosis-specific) were collected for 3 years starting from the date of application of the new policy (1 January 2004). The numbers of ED visits and hospital admissions were obtained from the hospital recording electronic system and the EDOU data—including numbers of patients, diagnoses and admission/discharge numbers—were obtained from hand-filled ED registration books. The numbers for the same categories for the preceding 3 years were collected for comparison.

The work was based entirely on personal efforts and was not supported by any financial or administrative motivations.

Statistical methods

The change in the annual admission rate was calculated by comparing the means of absolute numbers of total and disease-specific medical admissions over the study and comparison periods. The p value was calculated using the two-sample unpaired t test. The proportions (%) of hospital medical admissions to the EDOU for each disease category were calculated and comparisons were made between the two periods using means, standard deviations and p values (two-sample unpaired t test). The percentage reductions in admission rates relative to the comparison period were calculated.

RESULTS

The absolute numbers of medical admissions and admissions to the EDOU during the two periods of the study are shown in table 2. The proportion (%) of medical hospital admissions to EDOU cases was significantly reduced for all categories, with a reduction from 12.8% (95% confidence interval (CI) 11.4% to 14.3%) to 8.2% (95% CI 6.9% to 9.4%), p = 0.02 for total medical admissions; from 44.4% (95% CI 35.3% to 53.4%) to 21% (95% CI 5% to 37%), p = 0.006 for chest pain; from 32.3% (95% CI 22.9% to 41.7%) to 16.4% (95% CI 12.4% to 20.5%), p = 0.003 for bronchial asthma; from 91.6% (95% CI 87.8% to 95.4%) to 59.8% (95% CI 52.6% to 67.1%), p<0.001 for heart failure; from 93.6% (95% CI 89.5% to 97.7%) to 38.2% (95% CI 31.1% to 45.4%), p = 0.002 for pneumonia; and from 77.9% (95% CI 63.6% to 91.9%) to 44.2% (95% CI 38.5% to 50.0%), p<0.001 for pyelonephritis (table 3).

To demonstrate the effect of such a reduction in admissions on cost, we calculated the cost reduction for pneumonia. The results show that a mean of 491 cases of pneumonia were discharged yearly from the EDOU under the new policy compared with 47 cases before implementation of the policy, so 444 admissions were avoided. The mean daily cost for inpatient stay (without adding the treatment cost) in our hospital is 60 Kuwaiti Dinars (KD) (US$210), a value obtained from the statistics section of the Ministry of Health. The mean inpatient length of stay for community-acquired non-ventilator associated pneumonia was calculated as 5.5 days. The average yearly cost reduction for pneumonia therefore equals 444×5.5×60  =  146 520 KD (US$512 845).

To study the effect of the reduction in admissions on patient outcome, 98 patients with pneumonia who were discharged from the EDOU and treated with outpatient antibiotic therapy were followed; 80 patients (78.4%) preferred outpatient treatment to hospital admission and were satisfied, 3 patients (3%) were readmitted to the hospital with poor response, and no mortality was reported among the 98 cases.

DISCUSSION

Emergency admissions are rising and bed crises are occurring almost daily in many hospitals.14 Increased waiting time for transfer to an inpatient bed has become the most important cause of ED overcrowding.4 15 The main aim of our policy was to reduce medical admissions with consequent reductions in cost and overcrowding.

Although education and training to improve management and decision-making are important for all the working staff, assigning an admission avoidance team guarantees responsibility and reliability in some critical circumstances. Only a few reports have studied the effects of an admission avoidance team in limited situations.10 11

Providing ED doctors with chosen guidelines would lead to a more unified action, better decision-making and probably fewer admissions. In a study by Atlas et al, providing physicians with the pneumonia severity index score increased the proportion of patients with pneumonia treated as outpatients from 42% to 52%.16 Reports show that the application of disease management protocols or guidelines was effective in reducing admissions for bronchial asthma,17 heart failure,6 18 pneumonia19 and chest pain.20

We modified our EDOU to allow application of the guidelines and perform certain procedures such as blood transfusions and parenteral antibiotic therapy. Before implementation of the policy, patients had to be admitted to hospital for parenteral antibiotic therapy as parenteral antibiotics are not given on a community basis. A review of the literature showed that EDOUs for specific medical emergencies such as chest pain,21^–23 heart failure24 25 and bronchial asthma26 are effective in reducing admissions and cost and in improving patient outcome and satisfaction. In the case of heart failure, Peacock6 reviewed the most recent data on ED management of heart failure and found that 75% of heart failure presentations are anticipated to be discharged with management in the observation unit using an intensive therapy protocol compared with 20% under standard care. For chest pain, Goodcare et al22 observed a 32% reduction in admissions with a chest pain observation unit compared with routine care.

In a study of the effect of an EDOU on the total admission rate, Lateef et al27 reported that observation in an EDOU resulted in a 6.4% reduction in admissions compared with direct inpatient admission. In a systematic review by Daly et al,28 short-stay observation units were found to increase patient satisfaction, reduce admissions and improve cost effectiveness. However, modifications to the observation unit were implemented alongside new clinical protocols and it was not possible to distinguish the relative benefits of each. A study of the effects of the combined changes (ie, observation units and protocols) is needed.

No studies were found of the combined effects of guideline and observation units in Arabic or Middle Eastern countries. The only study of the effect of several approaches on the reduction in total medical admissions was conducted by Rossi et al.29 This study showed that improving quality in emergency services by organisational, professional and economic changes resulted in a 11.2% reduction in medical hospital admissions.

In our study, the effect of the reduction in admissions on patient outcome in terms of morbidity (readmissions), mortality and patient satisfaction was not fully evaluated because of the large number of patients and the diversity of the clinical conditions studied. Moreover, using validated international protocols and guidelines is expected to produce a favourable outcome. However, we have been able to demonstrate a favourable outcome among 98 non-admitted EDOU patients with pneumonia. The average yearly cost saving for pneumonia was US$512 845. The total effect of the reduction in admissions on cost is expected to be huge owing to the large number of avoided admissions compared with the period before implementation of the new policy. The success of our policy in significantly reducing the admission rate gives ample proof that the multidisciplinary evidence-based approach is essential for ED management. In spite of the variations in resources between countries and hospitals, we believe that any ED in a general hospital can establish or modify an EDOU and doctors can choose suitable applicable protocols to avoid unnecessary admissions.