rss
Emerg Med J 30:320-323 doi:10.1136/emermed-2011-200788
  • Original article

Association between admission delay and adverse outcome of emergency medical patients

  1. Sumalee Kiatboonsri2
  1. 1Department of Medicine, Ramathibodi Hospital, Mahidol University, Bangkok, Thailand
  2. 2Division of Pulmonary and Critical Care Medicine, Department of Medicine, Ramathibodi Hospital, Mahidol University, Bangkok, Thailand
  1. Correspondence to Dr Detajin Junhasavasdikul, Department of Medicine, Faculty of Medicine, Ramathibodi Hospital, 270 Rama 6 rd, Rajathevi, Bangkok 10400, Thailand; detajin{at}yahoo.com
  1. Contributors DJ developed study concept, designed data collection form, collected, cleaned and analysed data, and drafted and revised the paper. He is guarantor. PT developed study design, wrote the statistical analysis plan, analysed data, and revised the paper. He is guarantor. SK developed study concept, analysed data, and revised the paper.

  • Accepted 10 April 2012
  • Published Online First 5 May 2012

Abstract

Aim To determine whether admission delay (lead-time) and other factors are associated with hospital mortality rates of emergency medical patients.

Methods Patients presenting with emergency conditions during August to November 2009, and admitted to medical wards, including intensive care units, were enrolled. The time each patient spent in the ED, and other parameters were recorded. The primary outcome was the association between lead-time and hospital mortality. The secondary outcome was the association between lead-time and delta Modified Early Warning Score (MEWS) (MEWS at ward − MEWS at ED).

Results 381 cases were analysed. The overall mortality rate was 8.9%. By univariate analysis, the significant factors associated with mortality outcome were lead-time, ECOG (Eastern Cooperative Oncology Group) score, MEWS at ED, delta MEWS and sepsis. By multivariate analysis, the remaining significant factors were MEWS at ED, delta MEWS and sepsis. There was no significant relationship between delta MEWS and lead-time. In a sub-group of patients admitted to intensive care units, however, there was a positive correlation between lead-time and delta MEWS.

Conclusion MEWS, delta MEWS and sepsis were predictors of hospital mortality in emergency medical patients. Lead-time was not associated with mortality, which could be due to benefits of various treatments initiated in the ED. In patients requiring intensive care, however, the longer lead-time probably led to higher MEWS and mortality.

Background

It has been shown that the longer the duration some patients spend on arrival at the emergency department (ED) until their admission is associated with mortality or other unfavourable outcomes. These types of patients included, for example, those with emergency surgical conditions,1 major trauma2 ,3 and acute coronary syndrome.4 ,5 Outcomes of critically ill medical patients, that is, those needing intensive care unit (ICU) admission, were also associated with lead-time (duration from first arrival at the ED to ICU admission).6 Delayed ICU admission resulted in higher in-hospital mortality.7 However, there is only limited data in patients with less critical medical conditions. Duke et al have tried but failed to identify the significance of lead-time in these less severe patients.6

In the study of Duke et al, the longest lead-time in the series was 24 h. However, the shortage of available hospital beds can result in much longer admission delay. We hypothesised that, even in less severe medical patients, a particular increase of lead-time may eventually increase the rate of adverse outcome. Thus the main objective of this study was to determine whether admission delay (lead-time) was associated with mortality outcome among emergency medical patients admitted at Ramathibodi Hospital.

Materials and methods

Patients

This study was carried out at Ramathibodi Hospital, a medical school and tertiary care centre located in Bangkok, Thailand. All medical patients who arrived at the ED with emergency medical conditions and were admitted to medical wards between August and November 2009 were enrolled in the study.

All emergency patients were triaged and initially managed by an emergency team including residents of emergency medicine and staff (ED team). Patients with medical conditions, after initial management by the ED team, were subsequently allocated to the internal medicine residents who attended the emergency room for further appropriate treatment until admission. The admission depended on the number of available beds while the priority of admission and the site of care (ICU or general wards) were considered from the severity of disease. The remaining cases awaiting admission would stay in the observation area of the ED and would be subsequently managed by internal medicine residents. Essential investigations, treatments, and basic monitoring, such as vital signs measurement, were provided until the admission was available. A senior internal medicine resident was assigned to manage the waiting list, that is, which patient to be admitted next.

Patients aged ≥15 years, who presented with emergency medical illness or exacerbation/deterioration of underlying chronic disease were included in the study. We excluded some groups of patients who visited the emergency department for other reasons, such as outpatient investigation and chemotherapy administration. All cases without indication for admission were also excluded.

Study design and data collection

A prospective, observational study was performed. The study was approved by the Ethics Committee of Ramathibodi Hospital, Mahidol University. All patients gave informed consent.

Data recording was done by medical residents who attended the ward. The doctors collected data by using the record form specifically designed for this project. A scoring system was needed to assess the severity of ED patients, but none had previously been routinely used in the ED. After some literature review, we decided to use the Modified Early Warning Score (MEWS). The MEWS was approved since it was shown to have good potential to predict mortality rate before admission.8 Furthermore, Subbe et al have proven that the MEWS was associated with mortality in a group of medical emergency admissions.9 The MEWS comprises five physiological parameters which can be evaluated at the bedside without requiring complicated calculations. The five parameters are: systolic blood pressure, pulse rate, temperature, respiratory rate and level of consciousness recorded as AVPU scale (A=alert, V=responding to voice, P=responding to painful stimuli, U=unresponsive). All parameters are calculated by using the scoring table.

The MEWS can be used immediately with the patients at the time of presentation at the ED,10 ,11 on arrival at the ward,12 or recorded subsequently to evaluate changes. An increase in the MEWS is associated with an increase in mortality,9 while improvement of this score suggests better outcome.11

We recorded the MEWS at time of arrival at the ED and ward. Sex, age, time of arrival at ED, time of arrival at ward and Eastern Cooperative Oncology Group (ECOG) score were also recorded. The ECOG has been used to describe performance status or severity of patients with chronic disease, especially those with malignancy.13 The ECOG score ranges from ‘0’ in fully active patients, to ‘4’ in completely disabled ones. ‘Time of arrival at ED’ was defined as the time the patient first arrived at the ED. ‘Time of arrival at ward’ was defined as the time the patient reached the ward. Finally, the ‘lead-time’ was defined as total duration spent from ‘time of arrival at ED’ to ‘time of arrival at ward’.

Completed record forms were removed from the chart and sent to the researcher for data verification. The completion of missing data was performed by the research team by carefully reviewing the medical records. If the data was not complete after review, that particular form was discarded from the study.

The data related to mortality outcome as well as principal diagnosis were derived from the database of the medical informatics department of Ramathibodi Hospital, recorded according to the WHO ICD-10 system.

Primary and secondary outcomes

The primary outcome was the association between the lead-time and in-hospital mortality. The secondary outcome was the association between the lead-time at ED and delta MEWS (MEWS at ward − MEWS at ED).

Statistical analysis

From mortality rates in the past 14 months, we calculated sample size in a single group with target α error=0.05 and minimal size of difference that would be detected at ±2.5%. The sample size required was at least 499 patients.

The results were analysed using SPSS V.17.0.

The continuous parameters between survivors and non-survivors were presented by means and SD or median and range in cases where the data were not normally distributed. The categorical variables were presented with percentages. Comparison was made by using the t test or Mann–Whitney U test for continuous variables and the χ2 or Fisher's exact test for categorical variables. Logistic regression was used to finally determine the factors that affected mortality outcome. Correlation between delta MEWS and lead-time was calculated by the Pearson correlation. A p value <0.05 indicated statistical significance.

Results

Baseline characteristics

During the research period, 391 forms were submitted to our team. After reviewing medical records and exclusion of incomplete forms, 381 patients were finally eligible for analysis (table 1). With this sample size, univariate analysis of data had shown significant association between lead-time and mortality, so we decided to stop collecting data before the calculated target was reached.

Table 1

Baseline characteristics

Primary outcome

Overall mortality rate was 8.9%. The median length of hospital stay was 7 days (range 1–111 days). The factors associated with survival outcome were lead-time, ECOG score, MEWS at ED, delta MEWS and sepsis, as shown in table 2.

Table 2

Univariate analysis of factors associated with mortality

By logistic regression, the factors still associated with mortality were MEWS at ED, delta MEWS and sepsis. Our main interest concerning association between lead-time and mortality was not significant (table 3).

Table 3

Multivariate analysis of primary outcome

Secondary outcome

No significant correlation was observed between lead-time and delta MEWS in the overall group of patients. However, in post-hoc subgroup analysis of patients admitted to intensive care units (n=92), we found a significant correlation between lead-time and delta MEWS (R2=0.119, p=0.001; figure 1).

Figure 1

Pearson correlation of lead-time and delta Modified Early Warning Score (MEWS) in patients admitted to intensive care units. A longer lead-time was associated with an increased MEWS.

Discussion

Lead-time and adverse outcomes of all patients

In univariate analysis, it was noticeable that lead-time was associated with mortality but in an unexpected direction. Instead of admission delay making patients more at risk, the non-survivors had less lead-time than survivors. This may have resulted from medical residents tending to admit more-severe patients and leaving less-severe group in the ED. Thus the lead-time in the latter group was longer.

In addition, the patients waiting in the ED received treatment while waiting for an available bed. This may lead to improvement of the MEWS in the survivors group, although a longer lead-time was observed.

However, the association was not found in multivariate analysis. It is possible that the association between lead-time and mortality found in univariate analysis was just an effect of other confounders.

Lead-time in critical patients

By sub-group analysis, we found a correlation between lead-time and delta MEWS in critical patients requiring admission to the intensive care unit as in the studies of Subbe et al 9 and Duke et al.6 However, this correlation was not perfect, perhaps because of too small a number of cases in this subgroup.

Sepsis with mortality

Patients presenting with sepsis had more risk of death. This alerted us to give more attention to ED patients presenting with sepsis.

Earlier reports indicated that timing of appropriate antibiotic initiation was associated with mortality in sepsis or septic shock patients.14 ,15 There was no demonstrable association between lead-time and mortality in our post-hoc analysis of sepsis patients. This may be due to the small number of patients in the group (n=25), or partial treatment by antibiotics given to the patients while waiting for admission.

MEWS and its potential benefit

In our study, MEWS was easy to record, needed no complicated calculations and, as indicated in previous studies,9–11 can predict death. It can either be applied as an absolute value recorded once at the ED, or subsequently recorded to observe its change from baseline. We can also use delta MEWS to imply the effectiveness of treatment at the ED, as good management should result in a stable or decrease of subsequently recorded MEWS (delta MEWS ≤0).

Limitation of this study

Most patients in our study had received initial investigations and treatments by ED residents and staff before consulting medical residents. These included basic procedures, for example blood chemistry and antibiotics, and more advanced procedures, such as echocardiogram and thoracentesis. According to our ED guidelines, patients are managed by the ED team within a few minutes in the case of critical patients, or up to 2 h after arrival in less severe ones, but the actual time in each case was not recorded in our study. This may actually be the factor that determined mortality rather than the whole lead-time. From this point of view, fast and appropriate management by ED physicians may be the key concept to improving patients' outcome.

Moreover, we cannot show the association between lead-time and mortality in sepsis patients. A further study using a larger sample size may be able to demonstrate this. Time to first antibiotics should also be recorded.

Conclusion

The MEWS, delta MEWS and sepsis were predictors of hospital mortality in emergency medical patients. In univariate analysis, shorter lead-time was unexpectedly related to increased mortality of medical patients, but such an association was not found in multivariate analysis. This might be explained by case selection and early treatment provided in the ED. Prompt management by ED physicians may be the key concept that affects patients' outcome. In severe patients who need intensive care, however, the longer lead-time probably led to the higher MEWS and higher mortality. Admission delay should be avoided and establishment of admission fast-track should be useful in this group of patients.

In addition, the MEWS should be recorded in all medical patients arriving at the ED. This could be a helpful tool for triaging ED patients and monitoring the effectiveness of treatment in the ED before admission. Finally, this score can also be used for further analysis to improve ED management.

Acknowledgments

The authors would like to express their gratitude to all involved medical residents and officers, the medical informatics department and the medical records department of Ramathibodi Hospital for their contributions to this project.

Footnotes

  • Competing interests None to declare.

  • Ethics approval Provided by the Ethics Committee of Ramathibodi Hospital, Mahidol University.

  • Provenance and peer review Not commissioned; externally peer reviewed.

References


Free sample
This recent issue is free to all users to allow everyone the opportunity to see the full scope and typical content of EMJ.
View free sample issue >>

Don't forget to sign up for content alerts so you keep up to date with all the articles as they are published.

Poll

Among patients with minor TBI (GCS 13-15) getting CT scans ≥ 24 hours after injury, what proportion have a traumatic finding?

Results

0.5% - 43% response rate
3% - 41% response rate
10% - 16% response rate

Related original article: PCT head imaging in patients with head injury who present after 24 h of injury: a retrospective cohort study