The course of COPD involves periods of exacerbation of symptoms in between more stable periods [3, 6]. Such exacerbations are responsible for a temporarily increased need for treatment or even hospitalisation of the patients. Some patients experience frequent exacerbations, whilst others seem to avoid such exacerbations completely [56]. These acute exacerbations of COPD (AECOPD) often have an infectious cause and are associated with a more rapid decline in lung function, increased use of healthcare resources, and increased mortality compared to COPD patients without frequent exacerbations [7-10, 57-59]. Previous literature has shown a varying rate of exacerbations amongst COPD patients. Some studies have provided an estimate of less than one exacerbation per patient per year [60, 61], whilst other studies have shown the exacerbation rate to be between two and three per patient per year [7, 62].
These previous studies used different sample sources, and they differed in how they defined an AECOPD. In general, population-based studies found lower rates of exacerbations [60, 61, 63] than those targeting more selected populations, i.e.
outpatient clinics or hospital registers [7, 62, 64]. In addition, more permissive definitions of an exacerbation resulted in higher exacerbation rates than more strict definitions. E.g., in the Hokkaido sample studied by Suzuki et al., the rate of exacerbations was 0.78 per patient per year when defining an exacerbation as a subjective complaint of symptoms, whilst only 0.06 per patient per year when defining an exacerbation as the need of hospitalisation due to respiratory symptoms [65].
Predictors of exacerbations have been examined in various studies [56, 60, 64-71], and it has repeatedly been seen that both higher age [64, 66, 67], increasing airflow obstruction [56, 60, 68], a history of previous exacerbations [56, 64, 68],
inflammatory biomarkers [69, 70], gastroesophageal reflux disease [71], and reduced quality of life [65, 67, 68] all increase the risk of exacerbation. But again, the results from these studies are difficult to compare due to differing sampling sources and design.
27 All in all, previous studies vary substantially in their methodology, and are difficult to compare. Very few have samples from general populations [60, 68]. When
undertaking this PhD project, the effect of sampling source and exacerbation
definitions on the results had not been studied in adequate circumstances. We wanted to investigate how the exacerbation rate potentially could differ between a general and a selected population, and how the definition of an exacerbation could affect the results.
28 Table 1; selected previous studies on incidence of acute exacerbations of COPD. 1st author / journal / year published
Study aim(s) Design / population / follow-upDefinition of COPDSpirometryDefinition of exacerbationStatistical methods Main results Comments Seemungal / Am J of Respir and Crit Care Medicine/ 1998 AECOPD effect on HRQoL. Evaluate predictors of AECOPD.
Outpatient clinics. No AECOPD 4 weeks prior to study. N=70 (52 male, 18 female). 1 year follow-up. Diary cards (symptoms, fever, treatment). PEF. SGRQ measured HRQoL.
FEV1 < 70% of predicted. ß2- agonist reversibility < 15% or 200 mL.
Yes. Pre- and post- BD.
Anthonisen criteria. Unreported AECOPD: no revision by physician.
Categorical variables: Chi2 test and Mann- Whitney U test. Continuous variables: t test. Univariate and backward multiple regression. No. of exacerbations grouped: 0-2 and >2 / yr.
Mean 2.7 exacerbations per person per year (1.5 reported. Range 1 to 8). Past AECOPD and daily cough or wheeze predicted AECOPD. AECOPDs strongly correlated to SGRQ.
Low power Population selected fro outpatient clinics. The majorit participants were men. No control group. Donaldson / Thorax / 2002
1) Identify exacerbations. 2) Evaluate relationship between lung function decline and AECOPD.
Outpatient clinics. N=109 (median 74 men) fulfilled 365 days of diary information. Only 32 participants (29 men) with FEV1. 4 years of prospective follow- up.
FEV1<70% of predicted, and negative reversibility. Absence of asthma/other significant respiratory disease.
Yes.Anthonisen criteria. Hospital admission for AECOPD.
Cross sectional random effects models to evaluate the effect of exacerbation frequency on lung function decline.
100 participants (moderate to severe COPD). 757 AECOPD in 3 years. Median 2.53 AECOPD/yr. Significantly faster decline in lung function amongst frequent exacerbators.
Participants selected fro outpatient clinic. All ha an FEV1< 70 of predicted Majority of participants were men. No control group Montes de Oca / Chest / 2009.
1) Evaluate AECOPD frequency. 2) Explore predictors of Cluster-sampling from households of general populations in 5 Latin-American cities.
Post-BD FEV1/FVC ratio < 0.70 Yes.Self-reported and symptom- defined (deterioration of breathing Wald test for differences between GOLD stages. Multivariate 759 of 5314 subjects had COPD. 7.9% AECOPD the past year.
Sampling fr a general population. not compare
29 PLATINO study.exacerbation frequency. All adults≥ 40 years invited. 5314 subjects completed interviews and spirometry. 12 months of follow- up.
symptoms that affected daily activities or caused missed work).
logistic regression to examine predictors of AECOPD.
Increasing exacerbation frequency with increasing severity of COPD. Mean exacerbation rate per year was 0.58.
results to controls. Han / Radiology / 2011. COPDGene study
To investigate the relationship between AECOPD and quantitative measures of emphysema and airway disease.
21 centres, cross- sectional. Participants from local communities and outpatient clinics. Age 45 – 80 yrs. ≥10 packyears. Spirometry and CT thorax. Questionnaires 4 times a year for 5 years.
Post-BD FEV1/FVC-ratio < 0.7.
Yes.A flare-up of chest trouble last 12 months AND increased healthcare utilisation. A maximum of 6 episodes accepted for the last 12 months.
Multivariate analysis with a zero-inflated Poisson distribution. Adjustment for age, sex, smoking status, and FEV1 percent of predicted.
1002 participants. Mean AECOPD freq: 0.68/pers/year. ≥2 AECOPD/yr: younger, more women, less current smokers, lower FEV1, higher total lung emphysema percentage.
Participants both from general population and from selected outpatient clinics. Possible annual exacerbation rate restricted to 6/pers. Resource-based definition of exacerbations. Suzuki / ERJ / 2014 To investigate determinants of COPD exacerbations.
5 yrs cohort. Physician-diagnosed COPD recruited from 10 hospitals. Age≥ 40 yrs. ≥ 10 packyears. Asthma excluded. Median follow-up time 5 yrs. Monthly/bimonthly visits. Every 6 months, data on any change gathered.
Post-BD FEV1/FVC < 0.70. All GOLD stages included.
Yes.1) subjective 2) Anthonisen 3) symptom + prescription change, 4) symptom + AB or 5) symptom + hospital admission Poisson regression. Significant variables in univariate models included in multivariate regression.
Decreasing AECOPD from definition 1 to 5 (0.78, 0.24, 0.20, 0.13, and 0.06 AECOPD/pers/yr). Higher AECOPD frequency with severe airflow obstruction, with more dyspnoea, and with lower HRQoL.
Selected sample from hospital/outpati ent clinics, and no control group. Husebø / PlosOne / 2014
To examine predictors AECOPD Prospective cohort study. GOLD stage II-IV.
FEV1/FVC < 0.7 15 min post-BD, FEV1 < 80% of predicted Yes.Anthonisen criteria. Severity according to Random effects negative binomial 350 of 403 had ≥1 exacerbation. Median duration 14 days.
Selected participants from outpatient clinics and
30 frequency and duration.3 yr follow-up. Age 44-76 years. ≥10 packyears. Peripheral blood: biomarkers.
according to Norwegian reference values.
healthcare utilisation.regression for the incidence rate ratios (IRR) for each potential predictor.
Mean annual exacerbation rate/pers 1.40. Predictors: higher age, female sex, frequent AECOPD, higher GOLD stage, chronic cough, use of ICS.
specialist practices. No control group. Abbreviations: AB; antibiotics. AECOPD; acute exacerbation of COPD. BD; bronchodilator. COPD; chronic obstructive pulmonary disease. FEV1; forced expiratory volume in 1 second. F forced vital capacity. GOLD; Global initiative for chronic Obstructive Lung Disease. HRQoL; health-related quality of life. ICS; inhaled corticosteroids. PEF; peak expiratory flow. SGRQ; S George`s Respiratory Questionnaire.
31 1.1.8 COPD – treatment and prevention
As with love [72], “there ain`t no cure” for COPD. First and foremost, the best way to preserve lung health and avoid COPD is to avoid smoking. Restrictive tobacco policies are one of the most effective measures to maintain a good public (lung) health [73]. Both increased smoking cessation rates and reduced smoking initiation have been attributed to tobacco-control programmes [74, 75]. Increased taxes on tobacco have been identified as the most effective intervention against
non-communicable diseases and the expected millions of premature deaths attributed to tobacco-use in the decades to come [76, 77].
All COPD patients should be recommended to quit if still smoking [3, 5, 6]. In sustained quitters, the Lung Health Study showed a slower decline in lung function, less need of hospitalisation, and a lower all-cause mortality rate [78]. Secondly, pulmonary rehabilitation has been proven effective in improving health-related quality of life, tolerance to exercise, and reducing the need for health care services [79]. A study by Maddocks et al showed that even fragile COPD patients could have great benefit from pulmonary rehabilitation, improving both dyspnoea, physical activity, and overall health status [80]. The third most important intervention is vaccination, both against seasonal influenza, and pneumococcal disease. In two separate Cochrane reviews, it was found that influenza vaccination was significantly associated with a reduced exacerbation rate [81], and that pneumococcal vaccination protected against community-acquired pneumonia [82]. Further on, there has been seen an additive effect of receiving these two vaccines together [83].
The medication available consists mainly of inhalation drugs that can alleviate symptoms, and to some degree reduce the exacerbation rate and need for hospitalisation [84, 85]. It is recommended that all COPD patients that are symptomatic try out either rescue medication (if intermittent symptoms) or maintenance therapy (if persistent symptoms). The effect of rescue medication is rapid, but short-lasting. The most widely used drug for rescue medication is the ß-agonist salbutamol with nearly 300.000 users in Norway [86], but also the
short-acting anticholinergic drug ipratropiumbromid has quite a wide use. Maintenance therapy should include some form of long-acting medication, either a LABA (long-acting ß-agonist), or a LAMA (long-(long-acting muscarinic antagonist) [3, 5, 6], or a combination of the two. ICS (inhaled corticosteroids) have been used widely, but have an association with increased rates of pneumonia. Hence, ICS should be preserved for those experiencing frequent exacerbations where a somewhat reduced rate of serious events has been seen [84], or for patients with eosinophilia where recent evidence suggests a benefit [87]. Additionally, there can be a slight positive effect on inflammation adding the systematic PD4-inhibitor roflumilast in severe cases of COPD [88]. End-stage COPD may result in respiratory failure with hypoxaemia and/or hypercapnia, and some patients may profit from long-term oxygen treatment (LTOT). When exacerbating, patients often need systemic corticosteroids, and in many cases antibiotics. Patient education should enable patients to increase the rescue medication as appropriate when experiencing a worsening. If severe, exacerbations may lead to hospitalisation, and in some cases even intubation and treatment in intensive care units [3, 5, 6].