Commencement of systemic corticosteroids in emergency Department versus in ward in management of acute adult asthma requiring hospitalization : a retrospective cohort study

Introduction: Systemic corticosteroids are commonly used in management of acute asthma, sometimes started before admission in emergency department, sometimes in ward after admission. This study is to determine whether commencing systemic corticosteroids in emergency department compared to in ward for managing acute adult asthma requiring hospitalization can improve the outcome: shorter length of hospital stay. Methods: A retrospective cohort study was conducted in an emergency department in Hong Kong. Adults aged 18 to 65 years-old who presented to the emergency department with acute asthma and subsequently hospitalized with use of systemic corticosteroids were recruited and divided into two groups: a group with commencement of systemic corticosteroids in emergency department (Group A, n=139) and the other group with commencement of systemic corticosteroids in ward (Group B, n=209). The outcome measurement was length of hospital stay. Results: A total of 348 subjects were recruited in final analyses. We used Mann-Whitney U test to test the difference in ranking of length of hospital stay (days) between these two groups. The mean rank of length of hospital stay in Group A was 159, and that in Group B was 185 (p=0.014). The difference was statistically significant with commencement of systemic corticosteroids in emergency department resulting in higher ranking-shorter length of hospital stay. Conclusion: It may be possible to result in earlier discharge in acute adult asthma requiring hospitalization when systemic corticosteroids is started before admission in emergency department, instead of in ward after admission.


Introduction
In the world, there are 300 million people with asthma 1 .In 2005, 255000 people died from asthma 2 .In Hong Kong, the prevalence of asthma among different age groups has different prevalence rates from 5% to 10%; and there were more than 330,000 people suffering from asthma [3][4][5] .Acute asthma is a common emergency condition in adult, especially healthy young adult; thus, its management is very important in emergency department [6][7][8][9] .The pathophysiology of asthma involves airway inflammation, airway obstruction and bronchial hyper-responsiveness; hence, there are increased airway secretions, inflammatory cellular infiltrate and smooth muscle hypertrophy [10][11][12][13][14] .Systemic corticosteroids (including oral, intravenous or intramuscular route) can be used to control and suppress the inflammatory process in the airway in acute setting, and their use in treatment of acute asthma were established 15,16 .Moreover, a Cochrane review showed that use of systemic corticosteroids within 1 hour of presentation to an emergency department can significantly reduce the need for hospital admission in patients with acute asthma 17 .Also, systemic corticosteroids can help to reduce the rate of relapse; and thus, it should be administered to patients with severe asthma [18][19][20][21] .However, in management of acute asthma requiring hospitalization, some emergency medicine (EM) physicians would like to commence systemic corticosteroids in emergency department because they think that this would benefit these hospitalized patients.On the other hand, some EM physicians would not like to do so because they think there would be no significant difference to commence systemic corticosteroids in emergency department or in ward.However, there is no study to address this difference of beliefs.Therefore, this study was to determine, for managing acute adult asthma requiring hospitalization, whether commencement of systemic corticosteroids in emergency department before admission, compared to in ward after admission, can improve the outcomeshorter length of hospital stay.

Study Design
It was a retrospective cohort study to test the difference in ranking of length of hospital stay (days) between 2 groups of acute adult asthma: a group with commencement of systemic corticosteroids in emergency department before admission (Group A, n=139) and the other group with commencement of systemic corticosteroids in ward (Group B, n=209).Ethical approval was obtained from the Research Ethics Committee of Kowloon West Cluster of Hospital Authority in Hong Kong.Clinical Data Analysis and Report System (CDARS) was used to search all patients, attending the Accident and Emergency Department of Yan Chai Hospital, between 18 to 65 years-old, with acute asthma requiring hospitalization and administration of systemic corticosteroids from 1 January 2013 to 31 December 2014.Exclusion criteria were chronic lung diseases, other than asthma, (including chronic obstructive pulmonary disease and bronchiectasis), major medical illness (including pneumonia, ischemic heart disease, congestive heart failure, stroke, renal failure and liver failure, etc.), acute asthma within 4 weeks, use of systemic corticosteroids within 4 weeks and life-threatening acute asthma, shown intable 1.

Inclusion criteria Exclusion criteria
Acute asthma Chronic lung disease other than asthma showed that these factors were not statistically significant different between these 2 groups because all the p-value were greater than 0.05.Also, the lists and proportion of co-morbidity were shown in table 3.  The mean, median and mode of the length of hospital stay between 2 groups were showed in table 4. The mean length of hospital stay for Group A was 1.83 days and for Group B was 2.12 days.The median lengths of hospital stay in two groups were same-2 days and the mode of lengths of hospital stay in two groups were also the same-1 day.The range of length of hospital stay in Group A was from 1 to 7 days and that in Group B was from 1 to 10 days.By Mann Whitney test, the mean rank of length of hospital stay in Group A was 159.2 and that in Group B was 184.6.The difference of their mean rank was statistically significant with p value of 0.014, shown in table 5. Therefore, it was statistically significant that Group A had higher mean rank of length of hospital stay; this implied that Group A with commencement of systemic corticosteroids in emergency department was likely to have shorter length of hospital stay, and thus, earlier discharge.

Discussion
This study was designed as retrospective cohort study to determine whether commencing systemic corticosteroids in emergency department before admission compared to in ward for managing acute asthma requiring hospitalization can improve the outcome: shorter length of hospital stay.Shorter length of hospital stay was chosen as the outcome measurement; it implies earlier discharge.Shorter length of hospital stay or earlier discharge is very important for patients, clinicians, departments and administrators because this save resources for all of them.We found the mean rank of length of hospital stay among the acute asthma adult with commencement of systemic corticosteroids in emergency department (159.2) was higher than the group with commencement of systemic corticosteroids in ward (184.6).This difference was statistically significant.The higher rank implied the shorter length of hospital stay.The higher mean rank implied overall shorter length of hospital stay on average.Therefore, the group with commencement of systemic corticosteroids in emergency department had overall shorter length of hospital stay in average compared with the group with commencement of systemic corticosteroids in ward.It is likely related to the fact that earlier commencement of systemic corticosteroids helps to control and reduce the inflammatory process of acute asthma earlier 15,16 , and thus, resulting in shorter length of hospital stay and earlier discharge.There were quite a lot of possible confounding factors in our study.These were age, sex, smoking status, comorbidity status, severity of acute asthma, admission to EMW or medical ward, during hospitalization with use of salbutamol, ipratropium and antibiotics and before hospitalization with use of inhaled steroid, long-acting inhaled beta-agonist and oral theophylline.These possible confounding factors were identified and their p-value of statistical difference between Group A and Group B were calculated to see whether these factors were statistically significant different or not.The results showed that these factors were not statistically significant different.
Co-morbidity was one of important confounding factor to affect the length of hospital stay.To minimize its effect, patients with significant comorbidity including chronic lung diseases and major medical illness were excluded from this study.As a result, there were more than 70% of patients had no co-morbidity and more than 90% of patients had no or only one co-morbidity, shown in table 3. Therefore, the effect of co-morbidity in this study was markedly reduced.
Another important confounding factor was the severity of acute asthma.For stratification of the severity of acute asthma, we tried to stratify the severity as accurate as possible.We used BTS/SIGN Asthma Guideline 2014, which was based on clinical findings of respiratory rate, heart rate and completion of sentence in one breath and peak expiratory flow (PEF) percentage of the best or predicted.However, the stratification in our study was not very accurate because most of our cases (n=336, 97%) did not measure PEF, even with PEF measured, there were no best PEF mentioned nor patients' height for calculating the predicted PEF, thus, no PEF percentage could be calculated.And, there were missing data (n=217, 62%) for the completion of a sentence in one breath, this also affected the accuracy in assessing the severity of acute asthma.Since this study was a retrospective study, the missing data could not be retrieved; this was an important drawback in this study.GINA guidelines 2014 were not used for stratification of the severity of acute asthma in our study because there were even more missing data regarding agitation status, sitting position and use of accessory muscles.Subgroup analysis of patients with oral or intravenous corticosteroids was not done because it was shown that corticosteroids administered orally and intravenously had similar efficacy in the treatment of adult with acute asthma 23,24 .Moreover, the subgroup analysis of different dosages was also not done because it was shown that higher doses do not appear to be more effective for acute asthma 21 .
It is important to note that commencement of systemic corticosteroids in emergency department before admission instead of in ward will not pose extra harm to patients because systemic corticosteroids are just given earlier.
For patients already with frequent use of systemic corticosteroids, to commence systemic corticosteroid in emergency department may increase the exposure to systemic corticosteroid and related cumulative adverse effects like osteoporosis 24 and adrenal insufficiency.Therefore, for this group of patients, it may be beneficial to them if we hesitate to commence systemic corticosteroids in emergency department.

Conclusion
The mean rank of length of hospital stay was higher among patients with commencement of systemic corticosteroids in emergency department before admission than those in ward in our study.This implies to commence systemic corticosteroids in acute adult asthma in emergency department before admission instead of in ward, may result in shorter length of hospital stay or earlier discharge.Further prospective and randomized controlled trial will be useful to confirm our results and guide clinical practice.

Table 2 .
Characteristics of patients (n=348) SD=standard deviation; ED=emergency department; Comp. of a sent.§= completion of a sentence in one breath; * Chi-Square test; ǂ t-test; ƚ based on BTS/SIGN asthma guideline 2014

Table 3 .
Co-morbidity of patients

Table 4 .
Length of hospital stay (days) in commencement of systemic corticosteroids in emergency department (Group A) and in ward (Group B)

Table 5 .
Mean rank of length of hospital stay with commencement of systemic corticosteroids in emergency department (Group A) versus in ward (Group B) * Mann-Whitney U test