1. Objectives

The objectives of this review article are to:

  1. 1.

    Discuss the definition, aetiology, and health impact of acute exacerbations of chronic obstructive pulmonary disease (COPD).

  2. 2.

    Review the published literature on effective treatment of acute exacerbations of COPD.

  3. 3.

    Highlight the available literature pertaining specifically to treatment of elderly persons with acute exacerbations of COPD.

  4. 4.

    Familiarise readers with strategies to prevent acute exacerbations of COPD.

2. Disease Definition and Aetiology

COPD has traditionally encompassed the clinical conditions of chronic bronchitis and emphysema. This classification scheme, however, has many weaknesses. Chronic bronchitis, or the presence of cough and sputum production for at least 3 months in 2 consecutive years, fails to highlight the clinical significance of airflow obstruction in the impact of this disease. Emphysema is a pathological diagnosis, and is one of many other pathological abnormalities found in the lungs of patients with COPD. The Global Initiative for Chronic Obstructive Lung Disease (GOLD) has thus defined COPD as “a disease state characterized by airflow limitation that is not fully reversible. The airflow limitation is usually both progressive and associated with an abnormal inflammatory response of the lungs to noxious particles or gasses.”[1]

This definition shifts the focus away from the labels of emphysema and chronic bronchitis, while highlighting the need to demonstrate airflow limitation in order to definitively diagnose COPD. The presence of COPD is thus confirmed by spirometric testing, which shows a reduction in the ratio of the forced expiratory volume in 1 second (FEV1) to forced vital capacity (FVC).

For the vast majority of patients afflicted with COPD, the “noxious particles or gasses” mentioned in the GOLD definition come from tobacco smoke. Efforts to reduce tobacco use have been successful in many countries, but worldwide tobacco consumption continues to increase.[2] Exposure to indoor pollution in the form of biomass fuels (for cooking and heating) also increases the risk of developing COPD, particularly in developing countries.[3–5] Other environmental exposures, such as occupational dusts and fumes, also increase the risk of developing COPD.[6]

3. Burden of Disease

COPD is a major global health problem, and despite the advances of modern medicine, the prevalence and mortality of COPD continue to rise. COPD is currently the fourth leading cause of death worldwide and is projected to rise to the third leading cause by 2020.[1] When disability is considered, COPD is projected to rise from the 12th leading cause of worldwide disability in 1990 to the fifth leading cause by 2020.[7] The economic burden of this disease is also substantial. In the US, COPD is responsible for 13.8 million office visits and more than 670 000 hospitalisations annually.[8] Among adults ≥25 years old, COPD was listed as a primary or secondary diagnosis in 8.5% of all US hospitalisations, but among patients ≥65 years old, COPD accounted for 11.3–15.1% of all hospital admissions.[9] The overall direct and indirect medical cost of COPD for the US in 2004 was estimated to be in excess of $US37 billion.[8]

4. Acute Exacerbations of Chronic Obstructive Pulmonary Disease (COPD): Definition and Impact

The morbidity and economic cost of COPD are driven in large part by episodes of acute worsening of the underlying disease — commonly referred to as acute exacerbations of COPD. The precise definition of an acute exacerbation varies. The most widely referenced criteria are those of Anthonisen and colleagues[10]– acutely worsened dyspnoea, sputum purulence, and/or an increase in sputum volume — though these criteria were designed as part of a clinical trial and not for implementation in patient care settings.

More recently, an international working group proposed defining COPD exacerbations as “a sustained worsening of the patient’s condition, from the stable state and beyond normal day-to-day variations, that is acute in onset and necessitates a change in regular medication in a patient with underlying COPD.”[11] This definition remains clinical in nature, and reflects the current lack of objective criteria for defining exacerbations. In addition to the symptoms included in the criteria of Anthonisen and colleagues,[10] other symptoms associated with acute exacerbations of COPD can include wheezing, cough, fatigue, malaise and worsening exercise tolerance.

Spirometry often demonstrates decreases in FEV1 and peak expiratory flow. While these changes may predict the outcome of an exacerbation,[12] they can lag behind the clinical symptoms[13] and vary in degree of change; thus spirometry is generally not used in the definition of an acute exacerbation of COPD. To date, there are no biomarkers with adequate sensitivity or specificity for identifying acute exacerbations of COPD, though this remains an area of active research.[14]

The severity of acute exacerbations of COPD can vary widely. At one extreme, patients may experience only a transient, mild increase in dyspnoea; at the other extreme, patients may experience overt respiratory failure requiring hospitalisation and mechanical ventilation. Mild exacerbations may frequently go unreported. Even among participants in longitudinal COPD research cohorts, ≈50% of exacerbations are never reported to the research team.[13,15] In one prospective cohort of 101 patients with moderate-to-severe COPD (mean FEV1 42% of predicted), the average decrease in lung function at onset of exacerbation was small (≈5% from baseline), reflecting the fact that most exacerbations were defined by symptom changes on diary cards and did not require medical intervention.[13]

As might be expected, changes in lung function are appreciably larger among patients who seek medical attention for exacerbations. Assuming that recovery of lung function approximates premorbid values, patients treated for exacerbations as outpatients appear to have lost ≈20% of their baseline FEV1 at initial presentation.[16] Patients who require hospitalisation for an exacerbation may lose as much as 30% of their baseline FEV1 in the short term.[17,18] Therefore, it is not surprising that severe exacerbations requiring medical intervention have a substantial and protracted impact on health status.

Severe exacerbations requiring hospitalisation carry with them a substantial risk of death during the hospitalisation, though estimates vary anywhere from 3% to 30% depending on the particular subgroup studied.[19] The lower mortality estimates largely arise from studies investigating overall in-hospital mortality of all patients admitted for COPD exacerbations,[19,20] while the upper estimates derive from studies investigating mortality of patients specifically admitted with respiratory failure or admitted to the intensive care unit.[21,22] COPD exacerbations requiring hospitalisation also carry a substantial risk of recurrence and death over the ensuing months. In a Canadian administrative database of 22 620 patients discharged for COPD exacerbations, 25% required readmission over the next 12 months and 11% died over that period.[23] A smaller, but prospective, single-hospital, Dutch study of 171 patients admitted for COPD exacerbations found that 55% required readmission and 23% died over the subsequent 12 months.[24] Among patients admitted with hypercapnia, the prognosis is even worse: a cohort study of 1016 patients admitted with hypercapnia found that 44% of patients were readmitted within 6 months, and 33% died over that short period.[22]

Less severe exacerbations not requiring hospitalisation are also associated with significant health risks. In particular, a subset of patients with COPD appears to experience frequent, recurrent exacerbations. These patients experience a more rapid decline in quality of life[15,25,26] and may additionally experience a more rapid rate of decline in lung function.[27]

5. Aetiology and Pathogenesis of Acute Exacerbations of COPD

Multiple factors can cause acute exacerbations of COPD, including bacterial infections, viral infections, atypical infections and environmental pollution; however, a substantial portion of exacerbations have no clear aetiology.

Pathogenic bacteria are found in ≈50% of patients with an acute exacerbation of COPD.[28–31] The most common organisms are Haemophilus influenzae, Streptococcus pneumoniae, Moraxella catarrhalis, and H. parainfluenzae. However, the issue of differentiating colonisation from infection has been problematic, with bacteria routinely recovered from sputum samples from patients with stable (non-exacerbated) COPD.[31] At the same time, recent studies have isolated new strains of bacteria[32] and demonstrated strain-specific immune responses during acute exacerbations,[33] findings which suggest that true infection is truly a trigger for many exacerbations.

Studies of viral identification in COPD exacerbations have faced similar issues. When advanced polymerase chain reaction techniques are employed, viruses can be demonstrated in ≈40% of acute exacerbations of COPD, with picornaviruses (of which rhinoviruses are a member) being the most commonly demonstrated.[34,35] However, potentially pathogenic viruses have also been identified in respiratory secretions during stable COPD, so a problem exists in terms of assigning causality.[35] Other viruses found during exacerbations include influenza, parainfluenza, respiratory syncytial virus, adenovirus and human metapneumovirus.

The atypical organisms Mycoplasma pneumoniae and Chlamydia spp. have been implicated in acute exacerbations of COPD, though their importance is unclear. The available evidence suggests that Chlamydia is rarely responsible, being found in only 4–5% of acute exacerbations,[36,37] though it may be more frequently found in the most severely ill. One study of patients mechanically ventilated for respiratory failure showed acute serological evidence of Chlamydia in seven of 38 patients (18%), though concomitant bacterial pathogens were present in two of those seven cases.[29] The significance of Mycoplasma infection is difficult to assess, with one study demonstrating serological conversion in 14% of acute exacerbations, although in 71% of these cases, there was serological conversion for another respiratory organism as well; when these cases were excluded, Mycoplasma as the sole infectious agent was present in only 4% of hospitalisations.[38]

In addition to infections, environmental pollution has been implicated as a cause of acute exacerbations of COPD. Industrialisation of societies leads to increased production of particulate matter <10µm in diameter, sulphur dioxide, nitrogen dioxide and ozone, all of which are pro-inflammatory to the lung mucosa. Particularly with ozone, increases in environmental concentrations lead to increased risk of hospitalisation for COPD.[39] Up to 9% of admissions for acute exacerbations of COPD may be related to atmospheric pollutants, particularly during the summer months.[40]

6. Aging and COPD Exacerbations

The prevalence of COPD, both by self-report and by spirometry, increases with age (figure 1).[41] This increase may be due to (i) the slowly progressive nature of the disease, which often does not manifest itself until later in life; and/or (ii) the fact that COPD is more likely to become symptomatic late in life because of COPD-moderated loss in lung function coupled with normal age-related loss of lung function. Because the prevalence of COPD increases with age, it comes as no surprise that exacerbations also become more prevalent with advancing age. In the US in 2000, of the estimated 726 000 hospitalisations for COPD, 478 000 (66%) occurred in individuals >65 years of age, in whom the hospitalisation rate far exceeded that in younger people (table I).[41]

Fig. 1
figure 1

Estimated prevalence of self-reported chronic obstructive pulmonary disease (COPD) by age group in the US: questionnaire data from the National Health Interview Survey (NHIS) 1988–94,[42] and questionnaire and pulmonary function data from the Third National Health and Nutrition Examination Survey (NHANES III) 1988–94[43] (reproduced from Mannino et al.,[41] with permission). FEV 1 = forced expiratory volume in 1 second; FVC = forced vital capacity.

Full size image
Table I
figure Tab1

Estimated number of annual chronic obstructive pulmonary disease (COPD) exacerbations and annual COPD hospitalisation rate by age (US, 2000)[41]

Full size image

Several studies have also identified age as an independent risk factor for the development of COPD exacerbations and hospitalisations,[44–46] though others have not.[47–49] In a prospective, large (n = 1829), multicentre clinical trial involving COPD patients, advancing age was found to be a strong independent risk factor for an exacerbation requiring healthcare intervention and an exacerbation requiring hospitalisation.[50] The relative hazard increased by 9% for every 5 years over the age of 40 years for exacerbation and by 36% for every 5 years over the age of 65 with respect to hospitalisation.[50] In patients who have been hospitalised, multiple studies have identified advancing age as an independent risk factor for in-hospital mortality,[21,22,51–53] highlighting the impact of exacerbations in the elderly.

The reasons behind the association between aging and COPD exacerbations are not entirely clear. Age-related decreases in respiratory secretion clearance from small airways may contribute to increased risk of bacterial colonisation and/or infection.[54] Worsened expiratory flow limitation related to both normal aging and COPD may additionally impair clearance of infectious organisms. The increased prevalence of aspiration in the elderly may represent yet another mechanism of increased COPD exacerbation risk.[55] Aging is also associated with declines in both adaptive and innate immunity, and these declines may predispose elderly patients to respiratory infections.[56]

7. Pharmacological Treatment of Acute Exacerbations of COPD

Drugs used to treat acute exacerbations of COPD currently consist of three main established classes of therapy: (i) bronchodilators; (ii) corticosteroids; and (iii) antibacterials. We will review the published evidence supporting the use of these medications (see sections 7.1–7.3). For readers interested in further information, we also recommend the GOLD guidelines[1] and the joint statement from the American Thoracic Society (ATS) and European Respiratory Society (ERS).[57]

7.1 Bronchodilators

Short-acting β2-adrenoceptor agonists (SABAs) and anticholinergics are the main bronchodilators used to treat acute exacerbations of COPD. β2-Adrenoceptors are expressed on airway smooth muscle cells, and SABAs (which include salbutamol [albuterol], bitolterol, fenoterol, isoetarine [isoetharine], isoprenaline [isoproterenol], levosalbutamol [levalbuterol], orciprenaline [metaproterenol], pirbuterol, procaterol and terbutaline) increase the concentration of cyclic adenosine monophosphate (cAMP) in these smooth muscle cells; this causes a decrease in intracellular calcium, which results in bronchodilation and improved lung function.[58] Anticholinergics such as ipratropium bromide and oxitropium bromide block muscarinic receptors, and in the airways, inhibition of the muscarinic M1 and M3 receptor subtypes reduces smooth muscle contraction.[59] Of note, β-adrenoceptor agonists are available in inhaled, oral and parenteral preparations. Anticholinergics such as atropine are also available in all three forms. However, the available evidence supports the safe and effective use of only the inhaled forms of these drugs, and neither the GOLD[1] nor ATS/ERS[57] guidelines support use of the oral or parenteral forms of these drugs. Our discussion is therefore limited to the inhaled forms of these bronchodilators.

Both inhaled SABAs and inhaled short-acting anticholinergics are thought to be effective in improving FEV1 during acute exacerbations of COPD, though no placebo-controlled studies have been conducted. Three relatively small randomised trials compared these two classes of bronchodilators. Two studies, one of 39 patients[60] and the other of 52 patients,[61] involving patients admitted for COPD exacerbations compared ipratropium bromide with fenoterol and found no significant differences in FEV1 between these two classes of bronchodilators. Another study, in 32 patients, compared ipratropium bromide with orciprenaline in patients presenting to emergency departments or pulmonary clinics with acute exacerbations.[62] This study found similar degrees of improvement of FEV1 and FVC with either bronchodilator. This study also showed a small, but statistically significant, decrease in mean partial pressure of oxygen in arterial blood (PaO2) [from 64.8mm Hg to 58.6mm Hg] 30 minutes after administration of orciprenaline, but the change was transient, with resolution by 90 minutes following treatment. Patients in the ipratropium bromide arm actually had a small, but statistically significant, increase in their mean PaO2 (from 68.5mm Hg to 74.5mm Hg) at 30 minutes after ipratropium bromide administration and, as with orciprenaline, this effect was no longer evident 90 minutes following treatment. There appeared to be no clinical consequences of these small, transient gas exchange abnormalities.

While SABAs and anticholinergics appear to have similar physiological effects on lung function improvement, they work through different mechanisms, and there has been long-standing interest in combining the two classes of drugs for additional benefit. In patients with stable (non-exacerbated) COPD, existing data suggest an additive effect from combination therapy,[63,64] but this effect is not seen during acute exacerbations. Five randomised studies comparing a SABA regimen with an anticholinergic plus SABA regimen during acute exacerbations were unable to demonstrate any statistically significant difference in lung function between treatment groups.[61,65–68]

The delivery method of inhaled bronchodilator therapy has also been investigated, as these medications can be delivered via either nebulisation or via metered-dose inhaler. Seven randomised, controlled trials have addressed this question in acute exacerbations of COPD.[69–75] Of these studies, only one small, non-blinded study involving seven COPD patients showed some apparent benefit for nebuliser therapy;[73] the other six studies that addressed the issue of nebuliser versus metered-dose inhaler therapy showed no significant differences between delivery methods in a pooled total of 158 patients.[69–72,74,75]

Of note, nearly all these trials used spacer devices with the metered-dose inhalers. Use of spacer devices may additionally be of benefit for elderly patients, who often have cognitive or physical impairments that give rise to difficulties with use of metered-dose inhaler devices that can often be overcome with the use of spacer devices.[76,77]

In one study of 17 young (age 20–36 years) and 17 elderly (age 60–76 years) healthy volunteers, bronchodilator responses to salbutamol were decreased in elderly volunteers compared with their younger counterparts.[78] This suggests that in elderly patients, inhaled β-adrenoceptor agonists may not be as effective as in the young, though it is not clear if these findings can be extrapolated to patients with COPD. Another study suggested that bronchodilator responses to β-adrenoceptor agonists, but not to anticholinergics, decrease with age.[79] However, because this study was largely composed of patients with an asthmatic phenotype, extrapolating these findings to patients with COPD may be inappropriate. There is a lack of studies on the independent effect of age on inhaled bronchodilator responses in the specific setting of COPD.

Adverse effects of inhaled bronchodilators do not appear to be significantly different between the young and the elderly, though comparison studies are largely lacking. One small study of 17 elderly and young patients (mean age 71 and 23 years, respectively) attempted to compare the adverse effects of inhaled bronchodilators in elderly versus young patients.[80] Inhaled β-adrenoceptor agonists increased heart rate, lowered serum potassium and prolonged the cardiac QT interval, but there were no significant differences in the magnitude of these changes between young and elderly patients. Because the study was small, however, it may have been underpowered to detect these differences.

Hypokalaemia, tachycardia and cardiac QT interval prolongation are all adverse effects of concern in the elderly. β-Adrenoceptor agonists frequently cause hypokalaemia and, in fact, high doses of inhaled bronchodilators are effective when used in the treatment of hyperkalaemia.[81]β-Adrenoceptor agonists have also been associated with arrhythmias, particularly atrial fibrillation,[82,83] and are known to commonly cause tachycardia, which may be problematic for the elderly, given their higher prevalence of coronary artery disease. Because elderly patients are often taking other medications that may cause hypokalaemia (such as potassium-wasting diuretics) and QT interval prolongation (such as amiodarone, antibacterials, antidepressants and antipsychotics), clinicians should be aware of these potential adverse effects. Ipratropium bromide does not typically cause electrolyte abnormalities but may increase the risk of supraventricular tachycardia.[84] The main adverse effect of ipratropium bromide is dry mouth,[85,86] but some cases of urinary retention in older men with prostatic hypertrophy have been reported[87] and there have been reports of acute angle-closure glaucoma when ipratropium bromide is nebulised.[88]

Of note, the most commonly used β-adrenoceptor agonist in the US is salbutamol, a 50 : 50 mix of the (R)- and (S)-isomers of the drug. Because the (S)-isomer has weak β2-adrenoceptor activity and was felt to have potentially detrimental pro-inflammatory effects[89] and more potential for adverse effects, a pure (R)-isomer was developed and is now available as levosalbutamol. Levosalbutamol appears to provide a similar degree of bronchodilation as salbutamol when used in stable COPD patients and in patients with acute exacerbations of asthma.[90,91] Two observational studies found no advantage for levosalbutamol over salbutamol in terms of adverse effects among hospitalised patients with obstructive lung disease.[92,93] A third retrospective, observational study compared levosalbutamol use with salbutamol use 1 year previously and concluded that levosalbutamol shortened hospital length of stay.[94] All studies included patients with respiratory diseases other than COPD. These limited data indicate that a large prospective, randomised trial would be required to confirm any potential clinical benefit of levosalbutamol in the treatment of acute exacerbations of COPD.

Methylxanthines are another class of bronchodilators, of which the most commonly available forms are oral theophylline and intravenous aminophylline. These agents inhibit phosphodiesterase activity and increase intracellular cAMP levels, though there may be additional pathways by which methylxanthines act in COPD, such as restoration of corticosteroid-responsiveness through histone deacetylase[95] and improved diaphragm function.[96,97] Four published, randomised, placebo-controlled trials of theophylline for the treatment of acute exacerbations of COPD (sample sizes between 30 to 133 patients), showed no improvement in FEV1 when aminophylline was added to standard therapy.[98–101] There were no significant differences in clinical endpoints such as relapse rates,[98,100] hospital length of stay[101] or hospitalisation from the emergency department.[98]

Methylxanthines also carry with them a significant risk of serious adverse effects. These are generally related to high serum levels,[102] which are of particular concern in the elderly, whose metabolism of methylxanthines is generally decreased and can vary widely.[103,104] A meta-analysis of four randomised studies of use of methylxanthine for acute exacerbations of COPD found a significantly increased risk of nausea and vomiting in the methylxanthine-treated group.[105] There were also trends toward an increase in tremor and palpitations/arrhythmias, though the studies were not powered to look for adverse effects.

Key conclusions regarding the use of bronchodilators for the treatment of acute exacerbations of COPD are listed in table II.

Table II
figure Tab2

Use of bronchodilators for the treatment of acute exacerbations of chronic obstructive pulmonary disease (COPD): key points

Full size image

7.2 Corticosteroids

Corticosteroids represent another cornerstone in the treatment of acute exacerbations of COPD. This therapy has been established on the basis of multiple randomised, controlled trials. Four randomised trials have investigated the use of intravenous[18,106] and oral[17,18,107] corticosteroids in the treatment of acute exacerbations of COPD requiring hospitalisation. All four trials demonstrated improvement in FEV1 in the corticosteroid group compared with placebo. Three of these studies measured hospital length of stay, with two studies showing a statistically significant decrease in length of stay of 1.2 days[18] and 2 days[17] in the corticosteroid-treated group. The other study showed a statistically nonsignificant trend towards shorter length of stay in the corticosteroid-treated group (by 2 days).[107] Two studies examined relapse rates or treatment failures following hospital discharge.[17,18] One study showed significantly fewer relapses at 30 and 90 days after admission, but these differences were no longer evident at 6 months.[18] The other study measured relapses within 6 weeks following admission and found no differences between the corticosteroid-treated group and the placebo group.[17]

Two randomised studies evaluated oral corticosteroid use in outpatients with acute exacerbations of COPD who did not require hospitalisation.[108,109] One study of 27 patients showed a more rapid improvement in peak flow and dyspnoea scale scores in those treated with corticosteroids.[108] Perhaps more importantly, significantly fewer treatment failures (defined as the need for open-label prednisone or hospitalisation) occurred in the corticosteroid-treated group. The other study enrolled 147 patients with acute exacerbations of COPD who were treated and discharged from the emergency department.[109] This study confirmed the results of the earlier study, with the prednisone-treated group experiencing improved FEV1, improved dyspnoea scale scores 10 days after treatment and fewer relapses (defined as an unscheduled visit to a physician’s office or emergency department within 30 days).

Thus, systemic corticosteroids have been shown to improve the course of acute exacerbations of COPD (in terms of spirometric improvement, decreased length of hospital stay and decreased risk of relapse), but questions remain about the dosage and duration of therapy. Dosages have ranged from as low as oral prednisone 30mg once daily[17] to as high as intravenous methylprednisolone 125mg every 6 hours.[18] Data from the largest trial of patients admitted for COPD exacerbations showed no difference between a 2-week course of corticosteroids and an 8-week course, thus suggesting that 2 weeks of therapy is adequate.[18] A study of 36 patients hospitalised for COPD exacerbations compared a 10-day course of corticosteroids with a 3-day course and demonstrated better improvements at 10 days in the 10-day group compared with the 3-day group.[110] Compared with results in the 3-day treatment group, the 10-day corticosteroid treatment group exhibited statistically significant improvements at 10 days in mean FEV1 (236 vs 68mL, respectively), mean FVC (319 vs 17mL, respectively), and mean PaO2 (21.2 vs 11.3mm Hg, respectively). The study was underpowered to detect differences in clinical outcomes between the groups.

The adverse effects of systemic corticosteroids are numerous. While the long-term, detrimental adverse effects of chronic systemic corticosteroid therapy (such as osteoporosis, adrenal insufficiency, cataracts and skin thinning) are well-established, the risks of short-term use of systemic corticosteroids for acute exacerbations of COPD are less clear. The most common adverse effect noted in clinical trials of systemic corticosteroids for COPD exacerbations has been hyperglycaemia, with all four inpatient studies of corticosteroids showing more episodes of hyperglycaemia,[18,107] glucosuria[17] or higher mean glucose levels[106] in the corticosteroid-treated groups compared with the placebo groups.

Psychiatric adverse effects from short-term systemic corticosteroids may include insomnia, anxiety and depression. These adverse effects may be dose-related,[111] though in the study using the largest doses of corticosteroids (methylprednisolone 125mg intravenously every 6 hours), there were no significant differences between the corticosteroid-treated and placebo groups in acute psychiatric illnesses requiring psychiatric consultation.[18] The study was not designed or powered to detect less severe psychiatric adverse effects, so these potentially important clinical events may have been missed. An outpatient trial (in which patients received prednisone 40mg once daily) reported significantly more insomnia in the 74 corticosteroid-treated patients compared with the 73 placebo-treated patients (48% vs 21%, respectively; p = 0.001) and non-significant trends towards more depression (19% vs 10%, respectively; p = 0.14) and anxiety (27% vs 19%, respectively; p = 0.28) in corticosteroid-treated patients.[109]

Secondary infections are also of concern when using systemic corticosteroids to treat acute exacerbations of COPD. When systemic corticosteroids were administered for 8 weeks (using a tapering schedule) following admission for a COPD exacerbation, there was a trend towards more re-hospitalisations for serious infections in that group compared with groups receiving either only 2 weeks of corticosteroids or placebo.[18] Systematic reviews of clinical trials not involving COPD patients suggest that brief courses of systemic corticosteroids do confer a small risk of both lethal and nonlethal secondary infections.[112,113] The same may be true for COPD patients who receive short-term systemic corticosteroids, though the available information is inadequate to allow a firm conclusion.

Acute, corticosteroid-induced myopathy has also been associated with use of systemic corticosteroids for acute exacerbations of COPD. Typically, the risk of myopathy increases with increases in corticosteroid dosage[114] and is more commonly found in intubated, critically ill patients,[115] though isolated cases have been reported even following a single oral dose.[116] It is important to be aware of this complication because the myopathy can progress to a severe and chronic myopathy if corticosteroids are not promptly withdrawn.[117]

Gastrointestinal bleeding is still frequently cited as a complication of systemic corticosteroid use. A systematic review on this subject, however, failed to find any such risk;[113] earlier studies suggesting such a relationship had not adjusted for concomitant use of NSAIDs. The only large study that assessed this risk in patients being treated for acute exacerbations of COPD failed to find any increase in the incidence of corticosteroid-related gastrointestinal bleeding.[18]

Key conclusions regarding the use of corticosteroids for the treatment of acute exacerbations of COPD are listed in table III.

Table III
figure Tab3

Use of systemic corticosteroids for the treatment of acute exacerbations of chronic obstructive pulmonary disease (COPD): key points

Full size image

7.3 Antibacterials

As discussed in section 5, the most common aetiology for acute exacerbations of COPD is infection, with recovery of bacterial pathogens (and new strains of pathogens) from the lower airways at times of exacerbations occurring frequently. Antibacterials, therefore, might be expected to hasten recovery in these patients. However, antibacterials would not be expected to help exacerbations triggered by viral infections or environmental pollution. Unfortunately, differentiating bacterial exacerbations from non-bacterial exacerbations is clinically quite difficult, short of performing bronchoscopy for lower airway culture collection, a procedure that risks precipitating respiratory failure in those with marginal lung function. Currently, there is no non-invasive test with established ability to discriminate bacterial from non-bacterial COPD exacerbations. Thus, most patients experiencing more severe exacerbations of COPD are empirically treated with antibacterials.

Eleven randomised, placebo-controlled trials have been conducted to determine whether or not antibacterials are effective therapy in the treatment of acute exacerbations of COPD.[10,118–127] A recent meta-analysis concluded that, compared with placebo, antibacterials reduce the relative risk (RR) of death by 77% in patients with moderate to severe exacerbations.[128] This meta-analysis concluded that eight patients would require treatment with antibacterials to prevent one death (95% CI 6, 17). Of note, data from only four of these 11 trials were included in the mortality analysis, which was largely influenced by one particular study conducted in the intensive care unit.[124] In this study of 93 patients with the most severe exacerbations requiring mechanical ventilation, 9% of the antibacterial-treated patients died in the hospital, compared with 39% of the placebo-treated group. When this study was excluded from the mortality analysis, the authors stated that the point estimate of the reduction in risk did not change, but the 95% CIs for the risk reduction and number needed to treat widened and the risk reduction estimate became statistically insignificant. The meta-analysis[128] also suffers the same pitfalls of any meta-analysis: differing definitions of exacerbations, differing inclusion/exclusion criteria, differing treatment regimens (particularly problematic with the wide variety of antibacterials available internationally), differing outcome measures, and the possibility of publication bias against ‘negative studies’ that were unable to show differences.

The meta-analysis[128] was also unable to draw conclusions on the relationship between severity of illness and effectiveness of antibacterial therapy because of the small number of available studies and incomplete data on severity of illness in the studies. This question has arisen as a result of findings from one of the landmark studies on antibacterials in COPD exacerbations.[10] In that study, exacerbation severity was graded a priori according to the presence of three cardinal symptoms of COPD exacerbations: worsening dyspnoea, increasing sputum purulence and/or increasing sputum volume. The presence of all three symptoms was graded as the most severe exacerbation, while the presence of only one of the three symptoms was graded as the least severe. When all three symptoms were present, antibacterials provided the greatest benefit, with 63% of those receiving antibacterials improving compared with 43% of those receiving placebo. In patients with less than three cardinal symptoms, antibacterials were less effective (70% improving vs 60% receiving placebo in those with two cardinal symptoms and 74% vs 70%, respectively, in those with only one cardinal symptom). A separate study also demonstrated that in the absence of purulent sputum (as assessed by comparing sputum sample colour to a colour chart), 32 of 34 patients (94%) spontaneously recovered without antibacterial therapy; the two patients that eventually required antibacterial therapy actually developed purulent sputum prior to receiving the antibacterials.[129] Lastly, the previously discussed study of patients mechanically ventilated for exacerbations of COPD showed significant reductions in mortality when antibacterials were compared with placebo.[124] Thus, the available evidence supports the notion that patients with more severe exacerbations derive the most benefit from antibacterials.

The optimal choice of antibacterial and duration of therapy remain unresolved, as different trials have used differing antibacterial regimens and different durations of therapy. When antibacterials are used, selection of therapy should generally be directed to cover suspected organisms, including H. influenzae, S. pneumoniae and M. catarrhalis, which are the most frequently recovered organisms (see section 5). In the era of antibacterial-resistant organisms, the choice of a specific antimicrobial should also be guided by regional antimicrobial susceptibility patterns. Lastly, for the most severe exacerbations requiring mechanical ventilation, consideration should also be given to coverage of Gram-negative enteric bacteria, such as Pseudomonas spp. and Stenotrophomonas spp., on the basis of two studies which demonstrated recovery of these organisms in 16%[28] to 28%[29] of patients mechanically ventilated for acute exacerbations of COPD.

Key conclusions regarding the use of antibacterials for the treatment of acute exacerbations of COPD are listed in table IV.

Table IV
figure Tab4

Use of antibacterials for the treatment of acute exacerbations of chronic obstructive pulmonary disease (COPD): key points

Full size image

8. Other Therapies for Acute Exacerbations of COPD

In addition to pharmacological therapy, two other therapies are widely used in the treatment of acute exacerbations of COPD: oxygen therapy and assisted ventilation (see sections 8.1–8.2).

8.1 Oxygen Therapy

Acute exacerbations of COPD are frequently associated with gas exchange abnormalities and patients often present with hypoxaemia with or without hypercarbia. The aetiology of these gas exchange abnormalities is likely multifactorial, with demonstrated elements of both ventilation-perfusion mismatching (possibly from bronchoconstriction and mucus plugging of airways) and increased oxygen consumption (presumably from increased work of the respiratory muscles).[130] Because significant hypoxaemia can lead to metabolic acidosis and end-organ damage, oxygen therapy is routinely administered during acute exacerbations of COPD with associated hypoxaemia. While oxygen therapy reduces mortality in stable (non-exacerbated) COPD patients with resting hypoxaemia,[131,132] the mortality benefits of oxygen therapy in acute exacerbations have not been studied, for reasonable ethical concerns about not treating hypoxaemia in the acutely ill patient.

The main concern of clinicians administering oxygen therapy has been the risk of inducing hypercapnia which could precipitate the need for mechanical ventilation. The mechanism of this hypercarbia is not clear, with proposed mechanisms including hyperoxia-induced changes in dead space and hyperoxia-induced release of hypoxic ventilatory drive. Regardless of the mechanism, hypercapnia is a well described phenomenon, particularly when COPD patients inhale very high oxygen concentrations.[133–136] Clinicians have thus traditionally avoided high oxygen concentrations and aimed for the lowest oxygen concentration that provides oxygen saturations of 90% (thus approximating a PaO2 of 60mm Hg, which is generally adequate for tissue oxygenation). This strategy is often referred to as ‘controlled oxygen therapy’, implying close monitoring to determine the minimum amount of oxygen required. When a strategy of controlled oxygen delivery is adopted, the risk of hypercapnia appears to be quite low,[137–140] although large, randomised trials in this area are lacking.

Oxygen may also be delivered by a myriad of different systems, though systems that can potentially deliver unwanted high fractions of inspired oxygen (FiO2), such as face masks with oxygen reservoirs or nasal cannulas with oxygen reservoirs, are generally avoided. The most commonly employed oxygen delivery systems for exacerbations of COPD are nasal cannulas and Venturi masks. Nasal cannulas are less bulky than face masks (and therefore well tolerated by patients), but FiO2 cannot be adjusted accurately and unintended delivery of high FiO2 may cause hypercarbia during acute exacerbations of COPD. Venturi masks, in contrast, can more accurately control FiO2 by using the Bernoulli effect to control the amount of room air entrained into the system, thereby preventing unintended delivery of high FiO2. However, one small, crossover study of 18 patients demonstrated no difference in the risk of hypercarbia when nasal cannula oxygen was compared with Venturi mask oxygen.[138] The Venturi mask did provide significantly less time with saturations <90% over a 24-hour period compared with nasal cannulas (3.7 vs 5.4 hours over a 24-hour period). Larger studies in this area have not been performed.

The question of how to titrate oxygen therapy following an acute exacerbation of COPD remains largely unstudied. Though some patients with advanced COPD require continuous chronic home oxygen therapy, many patients experience hypoxaemia only during acute exacerbations of COPD and do not require long-term oxygen therapy. This was exemplified in one of the seminal studies of oxygen therapy for stable COPD, the Nocturnal Oxygen Therapy Trial.[131] Although this trial was not specific to patients following an acute exacerbation, of 1043 hypoxaemic COPD patients screened for study entry, between 170 and 201 (16–19%, specific data not reported) resolved their hypoxaemia within a 3-week observation period, such that they became ineligible for the trial.

Key conclusions regarding the use of oxygen for the treatment of acute exacerbations of COPD are listed in table V.

Table V
figure Tab5

Use of oxygen for the treatment of acute exacerbations of chronic obstructive pulmonary disease (COPD): key points

Full size image

8.2 Non-Invasive Ventilation

In some cases of acute exacerbations of COPD, despite the use of bronchodilators, corticosteroids, antibacterials and controlled oxygen therapy, patients can fail to improve and may progress to overt hypercapnic respiratory failure (generally defined as an arterial partial pressure of carbon dioxide [PaCO2] >45mm Hg). One mechanism of hypercapnic respiratory failure is use of excessive amounts of oxygen, as described in section 8.1. More commonly, however, patients are believed to develop a mechanical disadvantage of their diaphragm from the acutely worsened air trapping and hyperinflation. This is thought to lead to worsening hypercapnia and acidosis, thus further impairing optimal muscle metabolism, and thence to a downward spiral culminating in overt ventilatory failure. Historically, patients with hypercapnic respiratory failure from an acute exacerbation of COPD were treated with tracheal intubation and mechanical ventilation in the intensive care unit. However, intubation and mechanical ventilation can lead to multiple complications such as ventilator-associated pneumonia, barotrauma and delirium, which is especially common in the elderly.[141,142] Thus, use of non-invasive ventilation (NIV) for acute exacerbations of COPD complicated by hypercapnic respiratory failure became of interest.

NIV for acute exacerbations of COPD generally entails use of a device capable of delivering inspiratory-phase positive pressure (with or without expiratory-phase positive pressure) via a tight-fitting full-face mask or nasal mask. By pressurising the airway during an inspiratory manoeuvre, NIV reduces the work of breathing and may be beneficial in hypercapnic acute exacerbations of COPD by reducing the energy expenditure of the diaphragm (thus reducing its carbon dioxide production) and/or increasing ventilation to underventilated alveolar units. NIV also allows patients to use the mask intermittently, and thus communicate, swallow medications, eat and drink, which are distinct advantages compared with tracheal intubation.

Multiple studies have now investigated the use of NIV in acute exacerbations of COPD. Indeed, 14 randomised controlled trials investigating the use of NIV in hypercapnic respiratory failure (defined as a PaCO2 >45mm Hg) associated with acute exacerbations of COPD were analysed in a recent systematic review.[143] The review included six studies conducted in intensive care units,[144–149] seven studies conducted in general medical wards[150–156] and one study with an unspecified setting.[157] Two of the included studies were published only in abstract form.[148,157] This systematic review demonstrated reductions in multiple clinically important endpoints when NIV was added to usual medical care. These included reductions in mortality (RR 0.52; 95% CI 0.35, 0.76), risk of treatment failure (RR 0.48; 95% CI 0.37, 0.63), risk of intubation (RR 0.41; 95% CI 0.33, 0.53) and hospital length of stay (weighted mean difference of 3.2 days; 95% CI 2.1, 4.4). The analysis concluded that ten patients would have to be treated with NIV to prevent one death, five patients would have to be treated to prevent one treatment failure (defined as a death, need for intubation or intolerance to treatment) and four patients would have to be treated to prevent one intubation.

Despite the robust body of evidence supporting the use of NIV for hypercapnic acute exacerbations of COPD, some patients ultimately fail NIV and go on to require tracheal intubation and mechanical ventilation. The risk factors for predicting NIV failure have varied in multiple studies, and include high severity-of-illness scores,[158,159] inability to form a good seal with the device,[159] concomitant pneumonia[160] and even low serum albumin levels.[158] There is, however, no widely adapted model for predicting the success or failure of NIV. Thus, when using NIV to treat hypercapnic respiratory failure caused by acute exacerbations of COPD, patients require close clinical monitoring to ensure that they are not amongst the subset of patients for whom NIV will ultimately fail to improve their gas exchange abnormalities. Suggested protocols detailing methods for initiating NIV and monitoring response to therapy have been published,[161] and are beyond the scope of this review. Additionally, there are patients who present with acute indications for urgent intubation (such as haemodynamic instability, respiratory arrest, life-threatening hypoxaemia, or inability to maintain airway protection) for whom NIV should not be used in place of tracheal intubation.

Some clinicians may hesitate to offer tracheal intubation and mechanical ventilation to elderly patients with severe acute exacerbations of COPD because of concerns that such action will carry with it a high risk of death and/or prolonged mechanical ventilation. The mortality of patients requiring invasive mechanical ventilation for acute exacerbations of COPD, as mentioned in section 4, can be as high as 30%. However, when studies have examined mortality in acute respiratory failure, diagnosis of COPD has not emerged as an independent risk factor for mortality.[162] In fact, when compared with other diseases for which mechanical ventilation is commonly instituted (such as acute respiratory distress syndrome), mechanical ventilation for COPD is actually associated with a lower risk of mortality, shorter duration of mechanical ventilation and shorter intensive care unit length of stay.[163] However, of relevance to this review, advancing age has been identified as an independent predictor of mortality from any cause of acute respiratory failure.[162,163] The data thus suggest that while many factors may influence the decision of whether or not to institute mechanical ventilation (such as other co-morbidities, previous quality of life and perhaps advancing age), mechanical ventilation for acute exacerbations of COPD carries similar (and perhaps better) prognosis than mechanical ventilation for many other diseases resulting in acute respiratory failure.

Key conclusions regarding the use of NIV for the treatment of acute exacerbations of COPD are listed in table VI.

Table VI
figure Tab6

Use of non-invasive ventilation (NIV) for the treatment of acute exacerbations of chronic obstructive pulmonary disease (COPD): key points

Full size image

9. Prevention of Acute Exacerbations of COPD

A discussion of the treatment of acute exacerbations of COPD is incomplete without some discussion of strategies to prevent exacerbations. A full discussion of strategies to prevent exacerbations, however, requires reviewing a rather extensive body of literature and also requires discussion of statistical controversies related to analysis of exacerbation rates,[164] thus placing this type of discussion beyond the scope of this review. Readers interested in a more in-depth analysis of strategies to prevent COPD exacerbations are referred to recent reviews of this topic.[165–167]

Cessation of smoking is of paramount importance in the management of all persons with COPD. Smoking cessation slows the rate of decline in FEV1,[168–170] reduces the risk of COPD hospitalisation[171] and, most importantly, reduces mortality in patients with COPD.[172] Elderly smokers may be less likely to receive smoking cessation advice,[173] but this is not justified by the literature — smoking cessation interventions shown to be effective in the general population (such as counselling, self-help programmes and nicotine replacement) are equally effective in helping elderly persons to quit smoking.[174] Detailed reviews of effective smoking cessation interventions are available elsewhere.[174–176]

In addition to the SABAs discussed in section 7.1, long-acting β2-adrenoceptor agonists (such as salmeterol and formoterol) have become available, and this class of bronchodilators appears to have some effect in preventing exacerbations of COPD. A recent systematic review[177] concluded that, on the basis of results from four studies comparing salmeterol with placebo,[178–181] salmeterol prevents COPD exacerbations, with 24 patients requiring treatment with salmeterol to prevent one exacerbation (95% CI 14, 98).

Long-acting anticholinergics are also available, of which the only currently approved agent is tiotropium bromide. Tiotropium bromide has also been evaluated in a recent systematic review,[182] and based on the results of eight randomised controlled trials,[183–190] tiotropium bromide also appears to prevent COPD exacerbations, with 14 patients requiring treatment to prevent one COPD exacerbation (95% CI 11, 22).

While use of systemic corticosteroids for acute exacerbations of COPD has been rooted in solid evidence, the role of inhaled corticosteroids is much more contentious. Even among leading authorities in the field of COPD, there is considerable disagreement of opinion on the use of inhaled corticosteroids for COPD.[191,192] Nevertheless, three systematic reviews examining the effect of inhaled corticosteroids on COPD exacerbations have been completed.[166,193,194] Each review used slightly different selection criteria and therefore included a variable number of studies (six[166,193] to ten[194]) when assessing exacerbations, with four studies[195–198] being common to all three reviews. Despite including different studies, each analysis concluded that inhaled corticosteroids reduce the risk of exacerbations, with similar estimates for risk reduction, that is, 24%,[166] 30%[193] and 33%.[194] Only one of these analyses reported a number needed to treat,[194] and this was derived from an analysis of patients with moderate to severe COPD only. The authors concluded that 12 patients would need to be treated with inhaled corticosteroids for 17.7 months to prevent one exacerbation (95% CI 9, 18).

Because of the largely infectious nature of exacerbations, vaccination against respiratory pathogens might be expected to help prevent COPD exacerbations. The two main respiratory pathogens for which adult vaccination is widely available are S. pneumoniae (pneumococcus) and influenza virus. Pneumococcal vaccination has not been well studied in patients with COPD; only a few small, underpowered studies[199,200] and retrospective studies that included other lung diseases[201] comprise the bulk of the available data. One recent prospective study in COPD patients interestingly demonstrated a reduction in community-acquired pneumonia (data on exacerbations were not collected) only in those COPD patients <65 years of age, with no reduction in patients ≥65 years of age.[202] The US Advisory Committee on Immunization Practices recommends pneumococcal vaccination for all adults ≥65 years of age, whether or not they have COPD or other co-morbidities.[203]

Influenza vaccination has likewise been understudied in COPD, with the available database again consisting of small, randomised, controlled trials[204,205] and large, retrospective studies that included other lung diseases.[206] However, a systematic review on the subject concluded that influenza vaccination was effective in prevention of exacerbations.[207] Importantly, influenza vaccination of patients with COPD also appears to be safe, with no worsening of dyspnoea, exercise capacity or lung function following vaccination.[208,209]

Mucolytic agents (such as acetylcysteine [N-acetylcysteine], carboxymethylcysteine [S-carboxymethylcysteine] and bromhexine) have been evaluated for prevention of COPD exacerbations, the hypothesis being that improved mucus clearance will reduce the risk of exacerbations. A systematic review of the literature that analysed 26 trials involving 7335 patients concluded that use of mucolytic agents results in a very small reduction in the risk of exacerbations.[210] However, there was significant heterogeneity between included studies and the most recent and most carefully designed randomised trial found no reduction in exacerbation risk.[211] In that trial, patients not taking an inhaled corticosteroid did experience a reduction in risk with acetylcysteine, but as this was a subgroup analysis, this result must be interpreted with caution.

Pulmonary rehabilitation has been extensively studied recently, but its effect on modulating exacerbations has not been investigated in depth, with most studies focusing on health status scales and exercise performance. One observational study of 26 patients with COPD suggested that pulmonary rehabilitation reduces exacerbations, when comparing exacerbation frequency before and after pulmonary rehabilitation.[212] A prospective, randomised trial in 200 patients showed that participation in a 6-week pulmonary rehabilitation programme did not reduce the risk of hospital admission, but those in the pulmonary rehabilitation arm, when admitted, had shorter hospital stays.[213] Likewise, a prospective, randomised study of 60 patients showed no reduction in hospitalisations after pulmonary rehabilitation, but the study did find a reduction in the number of exacerbations per patient.[214] The frequency with which pulmonary rehabilitation programmes should be completed to maintain these potential benefits is unclear. One study demonstrated that reductions in exacerbations over 2 years were maintained only in the group of patients who performed pulmonary rehabilitation 1 year after their first pulmonary rehabilitation programme.[215] The mechanisms by which pulmonary rehabilitation might reduce COPD exacerbations are unclear.

Lastly, there has been increasing interest in the use of self-management or case-management programmes to reduce the risk of COPD exacerbations requiring hospitalisation. These programmes may involve many different interventions such as disease education, medication education, ‘action plans’ such as those used for asthma patients, or regular phone calls from a nurse or respiratory therapist. A recent systematic review was unable to discern any beneficial effect of these interventions, but the authors found the data “too sparse to discern any clinically relevant benefit or harm arising from such interventions”.[216] Two studies in particular have demonstrated efficacy in reducing hospitalisations,[217,218] but larger studies will be required before these programmes can be widely and definitively endorsed.

Key conclusions regarding strategies to prevent acute exacerbations of COPD are listed in table VII.

Table VII
figure Tab7

Strategies to prevent acute exacerbations of chronic obstructive pulmonary disease (COPD): key points

Full size image

10. Conclusion

COPD increases in prevalence with advancing age and is associated with significant morbidity, mortality and healthcare costs. Acute exacerbations of COPD are particularly problematic for patients with this disease and can be very severe, even culminating in death. Studies investigating the appropriate management of acute exacerbations of COPD have provided good evidence for the beneficial use of bronchodilators, corticosteroids, antibacterials, oxygen and assisted ventilation in the acute phase of the disease. Established strategies to prevent exacerbations of COPD include smoking cessation, pharmacotherapy and immunisations. We found limited data on the independent effects of aging on COPD exacerbation treatment and prevention. Given predictions of worldwide increases in COPD prevalence, this is an area of research demanding further investigation.