Original Contribution

Mortality-related factors after hospitalization for acute exacerbation of chronic obstructive pulmonary disease: the burden of clinical features

Ana Bustamante-Fermosel MD^a,*, Jose’ M. De Miguel-Yanes MD^a,

Mercedes Duffort-Falco’ MD^b, Javier Mun~oz MD^a

^aEmergency Department, Hospital General Universitario bGregorio Maran~o’n,Q Madrid, Spain

^bInternal Medicine Department, Hospital General Universitario bGregorio Maran~o’n,Q Madrid, Spain

Received 30 August 2006; revised 20 September 2006; accepted 25 September 2006

Abstract

Background: There is limited information about factors associated with mortality of patients with chronic obstructive pulmonary disease (COPD) admitted to hospital because of an acute exacerbation. Methods: A retrospective cohort study including all patients admitted to hospital through our emergency department (ED) was conducted. A total of 972 electronic discharge reports were reviewed. Patient baseline features, aspects concerning acute exacerbation, as well as demographic, cardiac ultrasound, and microbiological data were collected.

Results: in-hospital mortality rate was 6.4%. Of 315 patients with mild exacerbation according to Anthonisen criteria, only 1 died. In the univariate analysis, moderate to severe acute exacerbation of COPD, age older than 75 years, severe COPD, abnormal blood gas values, onset of complications during hospital stay, radiologic consolidation, a positive result in a microbiological respiratory sample, home oxygenotherapy, admission to the intensive care unit, left ventricular ejection fraction, and department of admission were statistically significant ( P b .05). The multivariate analysis showed that moderate to severe COPD acute exacerbation (odds ratio [OR] 7.3; 95% confidence interval [CI], 3.6-17.7), age older than 75 years (OR 4.9; 95% CI, 2.3-10.8), severe COPD (OR 4.6; 95% CI, 2.1-10), abnormal blood gas values (OR 4.7; 95% CI, 1.1-19.8), and complication during hospital stay (OR 2.8; 95% CI 1.4-5.4) were independently related to mortality.

Conclusion: We found that clinical aspect appears the most relevant of all potential determinants of in- hospital mortality for patients admitted for acute exacerbation of COPD. Thus, the clinical assessment and therapeutic decision taken in this first moment at the ED are the key that predict the prognosis of this patients. These data suggest that the risk of mortality after the admission to hospital of patients with COPD because of an acute exacerbation can be successfully predicted by making a clinical assessment at the ED. D 2007

* Corresponding author.

E-mail address: [email protected] (A. Bustamante-Fermosel).

Introduction

The prevalence of chronic obstructive pulmonary disease (COPD) follows an upward trend and involves high

0735-6757/$ - see front matter D 2007 doi:10.1016/j.ajem.2006.09.014

mortality and morbidity [1]. It is one of the few diseases whose incidence continues to rise [2]. Hospitalization for acute exacerbation represents the higher component of socioEconomic burden related to COPD. Yet, little is known about factors associated with mortality [3]. Previous studies have estimated in-hospital mortality for acute exacerbation of COPD to range from 4% to 30% [4-7]. Several studies have investigated predictive factors related to in-hospital mortality after acute exacerbation of COPD. Some charac- teristics reported as risk factors are Pao₂, oxygen saturation [3], older age [4,8], cardiac factors [4,8,9], other comorbid- ities [8], low body mass index, severity of illness, serum albumin level, and functional status [9].

The present study was designed to assess mortality rates

and potential determinants of in-hospital mortality for patients admitted for acute exacerbation of COPD. To do so, we conducted a retrospective cohort study by obtaining the information from the discharge reports.

Materials and methods

We included all patients with COPD according to Global Initiative for obstructive lung disease (GOLD) criteria [10] admitted to a large university hospital through our emergen- cy department (ED) from January 1, 2004, until December 31, 2004. Each patient was registered in the hospital patient administration system according to International Classifica- tion of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) and was classified considering diagnosis at discharge. Patients with a 491.21 diagnosis were included. Patients with a history of asthma specifically excluded from

491.21 and those wrongly classified were excluded.

We reviewed 972 electronic discharge reports, corres- ponding to 763 cases, because patients admitted more than once during 2004 were also included.

This study was approved by the Internal Review Board.

Outcomes

The aim of this study was to determine in-mortality rates and associated factors. Patient baseline features including severity of COPD, habits, comorbidities, previous admis- sions to hospital/intensive care unit (ICU), Long-term treatment, functional respiratory values where available, and relevant aspects concerning acute exacerbation were collected, as well as demographic, cardiac ultrasound, and microbiological data.

Baseline Severity of disease was defined by the GOLD 2005 criteria [10] in those cases where spirometry was available. In those cases for which spirometry was not available we defined clinical criteria that, similarly to GOLD 2005 criteria, identify 3 categories: mild, moderate, and severe. Three well-defined exacerbation causes were identified: infectious, noninfectious, and unknown. An infectious cause was proven if at least 2 of the following criteria were found: temperature greater than 388C,

leukocyte count greater than 12000/mm³ (12 d 10³/lL), or increase in sputum purulence. Comorbidity was quan- tified according to the Charlson index [11].

The month of admission was registered to assess its association with mortality and as a tool to know the level of exposure to the different air pollutants. Pollutant values were obtained through a public Web site: www.mamhienk. munimadrid.es). To assess the severity of the exacerbation, the classic clinical criteria of Anthonisen et al [12] were used, grouped into 2 categories (Table 1).

At our institution, patients with COPD exacerbation are admitted to either the pulmonary or the internal medicine department, or transferred to other units that depend on the hospital but are located in another building. Patients are thus classified into 2 groups: inpatients and derived patients. Socioeconomic level was estimated by taking the average annual income of the residential area in which each patient lived, according to statistical data from Consejer’ia de Economia e Innovacio’n Tecnolo’gica de la Comunidad de Madrid.

Cor pulmonale was defined as dilated right ventricle or when 2 of the 3 following criteria were met: edemas, ascites, or hepatomegaly. When echocardiogram was available, left ventricle dilation presence was recorded as well as pulmonary pressure. Systolic dysfunction was defined as an LVEF less than 45%. We considered a respiratory sample to be bpositiveQ when it yielded a positive result in Gram stain, acid-fast stain, or culture. Blood gas values were defined as abnormal when they fulfilled one of the following criteria during COPD acute exacerbation: Po₂ less than 50 mm Hg, pH less than 7.3, or Pco₂ greater than 70 mm Hg.

Statistical analysis

All categorical variables were dichotomized except for the cause of exacerbation, Charlson index, and the month

Table 1 Clinical criteria: acute exacerbation of COPD according to Anthonisen clinical criteria

Anthonisen criteria^a COPD acute exacerbation classification derived from Anthonisen criteria

Type 1: increased dyspnea, Moderate to severe sputum volume, and purulence COPD acute exacerbation

Type 2: two of the aforementioned symptoms were presented

Type 3: One of the aforementioned Mild COPD acute symptoms were presented and at exacerbation

least one of the following: upper respiratory tract infection; fever without other cause; increased wheezing; increased cough;

the past 5; or a 20% increase in either respiratory or heart rate relative to baseline

^a We used the same symptoms that Anthonisen used to graduate the severity of COPD acute exacerbation.

of admission, and presented as percentages. All continuous variables were expressed as mean (SD). In an initial univariate analysis, we tried to find an association between in-hospital mortality and all the independent variables. For continuous variables, Kolmogorov-Smirnov test served to identify those variables with a normal distribution, for which Student t test was used. For those variables that did not fit a normal distribution, Kruskal-Wallis test was considered suitable. For comparing categorical variables, v² and Fisher exact test were used. The relative risk is given as the risk factor associated with mortality as corresponding to a cohort study, except for the variables with 3 categories. A backward step logistic regression model was constructed to estimate the risk of in-hospital mortality associated with the independent variables. All variables associated with mortality with a P b .20 in the univariate analysis were included in the model. Signifi- cance at the P b .05 level was assumed.

Table 3 Patient characteristics: COPD acute exacerbation

features

Characteristics

Cause of exacerbation (n = 972): infectious Severity of acute exacerbation (n = 972):

moderate to severe complications (n = 972)

Department of admission in hospital (n = 972) Radiologic consolidation (n = 972)

Positive blood culture (n = 126) Positive respiratory sample (n = 162) ICU admission (n = 972)

Abnormal blood gas values (n = 947) Pollution levels^a (n = 932)

Values

76.4

67.8

18.2

64.2

20.4

8.8

38

5

19.6

Values are percentages or mean (SD).

PM₁₀ is particulate matter with an aerodynamic diameter of up to 10 lm.

^a Pollution concentrations are given in micrograms per cubic meter for all cases except CO, which is presented in milligrams per cubic meter.

SO2	10.68 (4.45)
CO	0.56 (0.23)
NO2	62.30 (16.37)
PM10	32.34 (13.93)
O3	28.13 (13.95)
Length of stay (n = 972)	10.59 (8.18)

Results

Patient characteristics

The study population was composed of 763 patients that generated 972 admissions (Table 2). Most patients were men (80.7%) and elderly (average age, 75.5 F 9.34 years). Annual income ranged from $8767 to $25,728. Fifty-five

Table 2 Patient characteristics: clinical, demographic,

echocardiography and baseline COPD features

Characteristics

Male

Age older than 75 years (n = 763) Cigarette smoker (n = 763) alcohol ingestion (n = 763) Income (n = 716), (euros/year)

No. of admissions per patient during 2004 (n = 963)

Home oxygenotherapy (n = 972) Corticosteroid use (n = 972)

ICU previous admissions (n = 763) Charlson index (n = 972)^a

I I

III

COPD severity (n = 972), mild to moderate Cor pulmonale (n = 972)

LVEF b45% (n = 270)

Left ventricular dilation (n = 243) Pulmonary pressure (n = 115) (mm Hg) FEV₁ (n = 153)

FEV₁/FVC (n = 150) b70%

Values are percentages or mean (SD).

Values

80.7

60

86.4

6.7

11926 (3400)

1.78 (2.38)

50.8

47.9

4.8

58

15.8

26.2

44.7

10.3

27.7

14

52.1 (15.4)

46.89 (21.30)

116 (77.3)

^a We identified 18 of the 19 categories, excluding the presence of COPD once it is a category that belongs to all patients considered for the study.

percent were classified as having severe COPD. Only 4.8% had been previously admitted to the ICU. A spirometry was performed in 153 patients during admission, showing a mean Forced expiratory volume in 1 second (FEV₁) of 46.9%. Two thirds (n = 658) had a moderate to severe acute exacerbation and, according to predefined criteria, in 743 admissions (76.4%) an infectious cause was the predispos- ing factor to the acute exacerbation. Fifty patients were admitted to the ICU and 48 (96%) needed ventilatory support. Characteristics of acute exacerbation of COPD are listed in Table 3.

In-hospital mortality

In-hospital mortality rate was 6.4%. One of the patients who died, after treatment with steroids and home oxygen- otherapy, was male, a smoker, older than 75 years, and whose microbiological respiratory samples tested positive (Table 4).

Patients admitted to hospital with acute exacerbation of mild COPD had a high Probability of survival. Only one death took place among 314 patients in this category. The Probability of death was higher in patients with moderate to severe acute exacerbation, as 61 of all 62 deaths belonged to this group.

More patients died in the general wards than in the ICU (53 vs 9). When we compared the 62 patients admitted to the ICU with the 53 patients who died in the general wards without having been admitted to the ICU, both populations showed a similar severity of acute exacerbation. The 3 main differences were that patients admitted to ICU were younger (28% vs 83% N75) ( P b.001) and that a lower percentage of

Table 4 Univariate analysis: in-mortality, clinical, demographic, echocardiography, and baseline COPD features

Characteristics	In-hospital deaths, n (%)a	in-hospital survival, n (%)a	RR (95% CI)b
Sex (n = 972)
Male	52 (6.5)	744 (99.5)	1.15 (0.60-2.21)
Female	10 (5.7)	166 (94.3)
Age (n = 972) (y)*
=75	48 (8.6)	392 (96.6)	2.49 (1.39-4.44)
b75	14 (3.4)	513 (91.4)
Cigarette smoker (n = 972)
No	5 (3.8)	127 (96.2)	1.79 (0.73-4.39)
Yes	57 (6.8)	783 (93.2)
Alcohol ingestion (n = 972)
No	61 (6.7)	846 (93.3)	0.23 (0.03-1.62)
Yes	1 (1.5)	64 (98.5)
Incomes: euros (n = 912)	58 (9400.89 [2199.24])	85 (49345.81 [2621.31])
No. of admissions during 2004 (n = 963)	61 (1.91 [2.38])	902 (1.77 [2.37])
Home oxygenotherapy (n = 972)*
No	10 (2.1)	468 (97.9)	5.03 (2.59-9.78)
Yes	52 (10.5)	442 (89.5)
Corticosteroid use (n = 972)
No	22 (4.3)	384 (95.7)	1.45 (0.88-2.40)
Yes	40 (8.6)	426 (91.4)
ICU previous admissions (n = 972)
No	60 (6.5)	865 (93.5)	0.66 (0.17-2.60)
Yes	2 (4.3)	45 (95.7)
Charlson index (n = 972)
I	31 (5.5)	533 (94.5)
II	8 (5.2)	146 (94.8)
III	23 (9.1)	231 (90.9)
COPD severity (n = 972)*
Mild to moderate	10 (2.3)	424 (97.7)	4.19 (2.16-8.16)
Severe	92 (9.7)	486 (90.3)
Cor pulmonale (n = 972)
No	52 (6)	82 (94)	1.61 (0.88-3.19)
Yes	10 (10)	90 (90)
Systolic dysfunction: FEVI b45% (n = 270)* No	6 (31)	189 (96.9)	3.47 (1.24-9.66)
Yes	8 (10.7)	67 (89.3)
Left ventricular dilation (n = 243)
No	9 (4.3)	200 (95.7)	2.05 (0.58-7.19)
Yes	3 (8.8)	31 (91.2)
Pulmonary pressure (n = 115) (mm Hg)	4 (58.50 [15.1])	111 (50.45 [15.15])
FEV1 (n = 153) FEV1/FVC (n = 150) b70%	2 (40 [9.89]) 0 (0)	151 (46.98 [21.41]) 116 (100)
z70%	2 (5.9)	32 (94.1)

^a In-hospital deaths and in-hospital survival are given as percentages for categorical variables and as mean (SD) for ordinal ones. For both variable types, the total number of cases for each item is also given.

^b Relative risk with 95% CI is given for those categorical variables in which it is possible to estimate.

* P b .05 (statistically significant).

ICU patients had severe COPD (66% vs 85%) ( P b.05) and were on home oxygenotherapy (64% vs 84%) ( P b .05).

In the univariate analysis, moderate to severe COPD acute exacerbation according to Anthonisen criteria, age older than 75 years, severe COPD, abnormal blood gas values, onset of complications during hospital stay, radio- logic consolidation, positive microbiological respiratory

samples, home oxygenotherapy, ICU admission, LVEF, and department of admission were statistically significant ( P b .05) (Tables 4 and 5).

The multivariate analysis showed that moderate to severe acute exacerbation of COPD (odds ratio [OR], 7.3; 95% confidence interval [CI], 3.6-17.7), age older than 75 years (OR 4.9; 95% CI, 2.3-10.8), severe COPD (OR, 4.6; 95%

Characteristics	In-hospital deaths, n (%)a	In-hospital survival, n (%)a	RR (95% CI)b
Cause of exacerbation (n = 972)
Infectious	49 (6.6)	694 (93.4)
Noninfectious	7 (7.3)	89 (92.7)
Unidentified	6 (4.5)	127 (95.5)
Severity of acute exacerbation (n = 972)*
Mild	1 (0.3)	314 (99.7)	29.20 (84.07-209.70)
Moderate to severe	61 (9.3)	597 (90.7)
Complication (n = 972)*
No	36 (4.5)	759 (95.5)	1.43 (0.89-2.29)
Yes	26 (14.7)	151 (85.3)
Department of admission (n = 972)*
Hospital patients	47 (7.5)	577 (92.5)	1.75 (0.99-3.08)
Derived patients	15 (4.3)	333 (95.7)
Radiologic consolidation* (n = 972)
No	36 (4.7)	738 (95.3)	2.82 (1.74-4.56)
Yes	26 (13.1)	172 (86.9)
Positive blood culture (n = 126)
No	15 (13)	100 (87)	1.39 (0.36-5.32)
Yes	2 (18.2)	9 (81.8)
Positive respiratory sample (n = 162)*
No	7 (6.9)	95 (93.1)	3.16 (1.33-7.47)
Yes	13 (21.7)	47 (78.3)
ICU admission (n = 972)*
No	53 (5.7)	864 (94.3)	3.13 (1.64-5.98)
Yes	9 (18)	41 (82)
Blood gas values (n = 947)*
No	24 (3.1)	738 (96.9)	5.83 (3.55-9.59)
Yes	34 (18.4)	151 (81.6)
Pollution levels (n = 932)
SO2	52 (10.28 [4.81])	880 (10.71 [4.43])
CO	52 (0.55 [0.22])	880 (0.56 [0.23])
NO2	52 (58.57 [14.46])	880 (62.52 [16.46])
PM10	52 (32.67 [12.75])	880 (32.82 [14.00])
O3	52 (30.23 [16.219])	880 (28.01 [13.80])
Length of stay (n = 972)	12.96 (12.09)	10.43 (7.82)

CI, 2.1-10), abnormal blood gas values (OR, 4.7; 95% CI, 1.1-19.8), and complication during hospital stay (OR, 2.8; 95% CI, 1.4-5.4) were independently related to mortality.

Table 5 Univariate analysis: in-mortality, COPD acute exacerbation features

^a In-hospital deaths and in-hospital survival are given as percentage for categorical variables and as mean value (SD) for ordinal ones. For both variable types, the total number of cases for each item of every variable is also given.

^b Relative risk (RR) with 95% CI is given for those categorical variables in which it is possible to estimate.

* P b .05 (statistically significant).

Discussion

This study highlights 3 relevant findings: first, in- hospital mortality of patients for acute exacerbation with COPD is relatively high; second, clinical features are the most reliable factors to predict this outcome; and third, when a patient has acute exacerbation of mild COPD according to classic Anthonisen criteria and is admitted to hospital, mortality tends to zero. These criteria show the most robust association with in-hospital mortality after comparing them with several demographic, physiologic,

therapeutic, or meteorologic categories traditionally associ- ated with an adverse prognosis.

Previous studies have focused on predicting in-hospital mortality, but in a different manner than we have. Patil et al

[5] found a mortality rate of 2.9% based on a national database, considering demographic characteristics and comorbidities; they did not include other variables that could have allowed a direct comparison with our population, such us severity of acute exacerbation or baseline status. Gunen et al [13] reported an 8.3% mortality rate, finding an association between in-hospital mortality and the following variables: lower Pao₂ and Arterial oxygen saturation, higher Paco₂, and longer hospital stay. Unlike us, they excluded patients with a specific cause of exacerbation, not consid- ering this variable as a determinant of prognosis. Besides, only patients with a severe acute exacerbation were admitted;

thus, the influence of severity of acute exacerbation was not analyzed. Finally, each patient was included only once in the study, so the number of admissions was not considered either. Groenewegen [3] reached an 8% mortality rate with a design comparable to ours, but their patients presented a higher number of comorbidities and a more severe baseline condition according to FEV₁.

In our study, the multivariate analysis showed that severity of acute exacerbation, age, abnormal blood gas values, COPD baseline severity, and onset of complications during hospital stay were independently related to mortality. Sixty-two patients died during admission, but only 9 of them died in the ICU, because 53 patients with a moderate to severe exacerbation died in the general wards without going through the ICU. We found that age, baseline severity, and home oxygenotherapy were key factors for admitting a patient to the ICU. These data imply that probably one of the main reasons that made patients less likely to survive was the decision not to admit them to the ICU, supported by those 3 factors. When evaluating the final outcome of the patients, it is worrisome to think that they might outweigh other clinical variables. To minimize this problem, ICU criteria admission must be defined and a specific admission policy about ICU usage, especially with patients who are more ill, must gain a foothold.

Several studies associate the severity of acute exacerbation with mortality [4,9,14-17]. During acute exacerbation, there is an increase in inflammatory cells and mediators [18,19]. Hogg et al [20] demonstrated that the progression of COPD from GOLD stage 0 to stage 4 was associated with a greater obstruction of the small airways mediated by a proinflamma- tory situation that leads to a thickening of the airway wall. This turns into a decline in lung function and eventually into higher mortality. In accordance with this, we find that severe COPD and moderate to severe acute exacerbation are independent predictors of in-hospital mortality.

Age has been a risk factor for in-hospital mortality, and our results point to this, too. One of the reasons given is that the older the patient, the higher the degree of obstruction found [3,17,21,22]. Blood gas values are predictive for in- hospital mortality as well. Most studies agree that a Po₂ of less than 55 mm Hg and a Pco₂ greater than 50 mm Hg are potential risk factors [3,8,13]. Because we used the values specified in GOLD 2005 for ICU admission [10], which are even more extreme than those described above, results from other studies can be extrapolated to ours.

Significance was found for the variable oxygen home therapy, yet no association with mortality was found for out- of-hospital use of steroids. This type of drugs does not change the decline of FEV₁ and, thus, long-term use of steroids cannot be further recommended [10,23]. Baseline severity of COPD seems to play an important role, although other authors found no association between Charlson Comorbidity Index and mortality [3].

Concerning the variables that are only significant in the univariate analysis, the association between the number and

the causes of exacerbation and a worse prognosis has been the target of several studies, in which mortality was not seen as the primary end point [24,25]. A higher number of acute exacerbations is associated with a faster decline in lung function; in a similar fashion, when lung function is poorer, exacerbations are more frequent [2]. Traditionally, the most common cause of acute exacerbation is the infectious one, with an incidence of 50% [25]; in our work, we found an incidence of 76.4%. Wilkinson et al [26] described the greatest reduction in FEV₁ in patients with a higher bacterial load, so that etiology seems to affect lung function too. This can explain why in our cohort mortality is significantly associated with positive respiratory samples.

Other studies have focused on the association between different pollutant concentrations and COPD mortality as well as severity of acute exacerbation [27-30]. We found no association with mortality. The difference between our results and those of previous studies may be that other studies considered the concentration a few days before the exacerbation as the level of exposure, whereas we took the average during the whole month before exacerbation.

Radiologic consolidation also appears significant in the univariate analysis. Chen et al [31] focused their inves- tigations on the association between respiratory comorbid- ities, such us pneumonia and influenza And in-hospital mortality, concluding that these patients pose a higher risk than those lacking respiratory comorbidities.

The association between in-hospital mortality and different cardiac factors has been studied in several ways, reaching heterogeneous conclusions [4,32,33] Concerning echocardiographic data, we just found statistical signifi- cance for having systolic dysfunction. Whereas Fuso et al reported an association between cor pulmonale and in- hospital mortality, our results point in the opposite direction; the disparity probably lies in the way criteria were settled. In our study, a misclassification bias cannot be ruled out when relying on some clinical data that are sometimes difficult to detect, such us liver enlargement or ascites.

In the present study, no influence of the length of stay on survival was detected. Most previous work managed length of stay as a primary end point instead of as a variable [31,33], probably because both of them are related to the outcome, missing a possible interaction between them.

One of the most relevant findings of our study is that when a patient with mild acute exacerbation is admitted to hospital, the risk of dying is nearly zero. This leads us to 2 facts: first, to explain the decision to admit patients having a mild exacerbation to hospital and the need for further observational prospective studies to clarify the suitability of managing this group on an outpatient basis. To answer the former, the admission to hospital was made based on clinical judgement and considering patient background. Because only 1 patient out of all 314 with a mild COPD acute exacerbation died, the statistical analysis of the different variables that may play a role in this decision is not possible. This is probably one of the major limitations of

the study, but 3 others must also be mentioned: first, the limitations due to the design of the study-a retrospective cohort study; some results are missing, which may hamper the analysis. This happened mainly with echocardiography and Respiratory function values and microbiological data; the analysis is weaker but its validity is not altered. Second, and derived from the same previously mentioned problem, clinical criteria had been created to classify the baseline COPD severity for those patients without measured data about functional respiratory status. The suitability of these criteria as well as a prospective study without missing values is warranted to confirm our results. And finally, the accuracy of diagnosis of the ICD-9 code is questionable. When this code is used, patients are commonly misidenti- fied and vice versa. Many patients are admitted initially with 1 ICD code and then switched the next month [6]. Probably patients in this study are underestimated. It will be necessary to determine retrospectively whether the patients with asthma and COPD were correctly categorized.

Conclusion

We conclude that the risk of mortality can be predicted through assessment of the clinical situation of the acute exacerbation of patients with COPD at the ED. This is the most powerful variable of the model, complemented by 4 other variables: age older than 75 years, severe chronic COPD, abnormal blood gas values, and onset of compli- cations during admission. These are all aspects easily and usually registered at the ED. Taking into account our main conclusion, the initial management of these patients may become the most crucial issue for altering their outcome: those patients with a severe to moderate COPD acute exacerbation could be managed in a more aggressive way and admitted to the ICU, and those with a mild COPD acute exacerbation might probably be managed as out- patients. Further studies should be conducted in this field regarding all aspects of patients with COPD to confirm our results and to assess the actual rates of mortality and the most suitable management according to the severity of their acute exacerbation.

Acknowledgment

The authors are grateful to bFundacio’n para la inves- tigacio’n Biome’dicaQ del Hospital General Universitario bGregorio Maran~o’nQ for the human support received.

References

1. Mannino DM, Homa DM, Akinbami LJ, et al. Chronic obstructive pulmonary disease surveillance. MMWR Surveill Summ 2002;51: 1 - 16.
2. Mannino DM, Watt G, Hole D, et al. The nature history of chronic obstructive pulmonary disease. Eur Respir J 2006;27:627 - 43.
3. Groenewegen KH, Schols MWJ, Wonters FM. Mortality and mortality-related factors after hospitalization for acute exacerbation of COPD. Chest 2003;124:459 - 67.
4. Fuso L, Incalzi RA, Pistelli R, et al. Predicting mortality of patients hospitalized for acutely exacerbation chronic obstructive pulmonary disease. Am J Med 1995;98:272 - 7.
5. Patil SP, Krishnan JA, Lechtzin N, et al. In-hospital mortality following acute exacerbations of chronic obstructive pulmonary disease. Arch Intern Med 2003;163:1180 - 6.
6. Cydulka R, McFadden ER, Emerman C, et al. Patterns of hospitalization in elderly patients with asthma and chronic obstruc- tive pulmonary disease. Am J Respir Crit Care Med 1997;156: 1807 - 12.
7. Seneff MG, Wagner DP, Wagner RP, et al. Hospital and 1-Year survival of patients admitted to intensive care exacerbation of chronic obstructive pulmonary disease. JAMA 1995;274:1852 - 7.
8. Incalzi RA, Fuso L, De Rosa M, et al. Co-morbidity contributes to predict mortality of patients with chronic obstructive pulmonary disease. Eur Respir J 1997;10:2794 - 800.
9. Connors Jr AF, Dawson NV, Thomas C, et al. Outcomes following acute exacerbation of severe chronic obstructive lung disease: the SUPPORT investigators [Study to Understand Prognoses and Prefer- ences for Outcomes and Risk of Treatment]. Am J Respir Crit Care Med 1996;154:959 - 67.
10. Global strategy for the diagnosis, management, and prevention of chronic obstructive pulmonary disease: NHLBI/WHO shop. National Heart, Lung, and Blood Institute: Claude Lefant. World Health Organization: Nikolai Khaltaer. Updated 2005.
11. Charlson ME, Pompei P, Ales KL, et al. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis 1987;40:373 - 83.
12. Anthonisen NR, Manfreda J, Warren CP, et al. Antibiotic therapy in exacerbations of chronic obstructive pulmonary disease. Ann Intern Med 1987;106:196 - 204.
13. Gunen H, Hacievliyagil SS, Kosar F, et al. Factors affecting survival of hospitalised patients with COPD. Eur Respir J 2005;26: 234 - 41.
14. Hurst JR, Wedzicha JA. Chronic obstructive pulmonary disease: the clinical management of acute exacerbation. Postgrad Med J 2004; 80:497 - 505.
15. Connors A, Dawson V, Thomas C, et al. Outcomes following acute exacerbation of severe chronic obstructive lung disease. Am J Respir Crit Care Med 1996;154:959 - 67.
16. Wedzicha JA. Exacerbations: etiology and pathophysiologic mecha- nisms. Chest 2002;121:136 - 41.
17. Anto’ JM, Vermeire P, Vestbo J, et al. Epidemiology of chronic obstructive pulmonary disease. Eur Respir J 2001;17:982 - 94.
18. Bhowmok A, Seemungal TA, Sapsford RJ. Relation of sputum inflammatory markers to symptoms and lung function changes in COPD exacerbation. Thorax 2000;55:114 - 20.
19. Barnes PJ. Chronic obstructive pulmonary disease. N Engl J Med 2000;343:269 - 80.
20. Hogg JC, Chu F, Utokaparch S, et al. The nature of small-airway obstruction in chronic obstructive pulmonary disease. N Engl J Med 2004;350:2645 - 53.
21. MacNee W. Acute Exacerbations of COPD. Swiss Med Wkly 2003; 133:247 - 57.
22. Wedzicha JA. Airway infection accelerates decline of lung function in chronic obstructive pulmonary disease. Am J Respir Crit Care Med 2001;164:1757 - 60.
23. Viejo JL. Inhaled glucocorticoids in chronic obstructive pulmonary disease. Arch Bronconeumol 2004;40:30 - 5.
24. Donaldson GC, Wedzicha JA. COPD exacerbations: 1. Epidemiology. Thorax 2006;164 - 8.
25. Sapey E, Stockley RA. COPD exacerbations: 2. Aetiology. Torax 2006;250 - 8.
26. Wilkinson TM, Patel IS, Wilks M, et al. Airway bacterial load and FEV1 decline in patients with chronic obstructive pulmonary disease. Am J respir Crit Care Med 2003;1090 - 5.
27. MacNee W, Donaldson K. Exacerbations of COPD: environmental mechanism. Chest 2000;117:390 - 7.
28. Hagen JA, Nafsted P, Skrondal A, et al. Association between outdoor air pollutants and hospitalization for respiratory disease. Epidemiol- ogy 2000;136 - 40.
29. Wong TW, Tam WS, Yu TS, et al. Associations between daily mortalities from respiratory and cardiovascular disease and air pollution in Hong Kong, China. Occup Environ Med 2002;30 - 5.
30. Wordley J, Walters S, Ayres JG, et al. Short term variations in hospital admissions and mortality and particulate air pollution. Occup Environ Med 1997;108 - 16.
31. Chen Y, Stewart P, Dales R, et al. In a retrospective study of chronic obstructive pulmonary disease inpatients, respiratory comorbidities were significantly associated with prognosis. J Clin Epidemiol 2005;58:1199 - 205.
32. Antonelli IR, Fuso L, De Rosa M, et al. Co-morbidity contributes to predict mortality of patients with chronic obstructive pulmonary disease. Eur Respir J 1997;2794 - 800.
33. Roberts CM, Lowe D, Bucknall CE, et al. Clinical audit indicators of outcome following admission to hospital with acute exacerbation of chronic obstructive pulmonary disease. Thorax 2002;57:137 - 41.