Page 1 of 31

ONLINE FIRST
This is an Online First, unedited version of this article. The final, edited version will appear in a numbered issue of CHEST and may contain substantive changes. We encourage readers to check back for the final article. Online First papers are indexed in PubMed and by search engines, but the information, including the final title and author list, may be updated on final publication. http://journal.publications.chestnet.org/

Online First articles are not copyedited prior to posting.

American College of Chest Physicians. Reproduction of this article is prohibited without written permission from the American College of Chest Physicians. See online for more details.

Downloaded From: http://journal.publications.chestnet.org/ by a Instituto Nacional de Enfermedades Respirat User on 07/02/2013

Page 2 of 31

Is it possible to identify exacerbations of mild to moderate COPD that do not require antibiotic treatment?

Short title: Antibiotics in exacerbations of COPD Authors: Marc Miravitlles (1), Ana Moragas (2), Silvia Hernndez (2), Carolina Bayona (3), Carl Llor (4).

Center:1. Pneumology Department, Hospital Universitari Vall dHebron, Barcelona. Spain. CIBER de Enfermedades Respiratorias (CIBERES), Barcelona,

Spain.(marcm@separ.es) 2. Primary Care Centre Jaume I, Tarragona, Spain. (amoragasm@meditex.es) (silviaha@comt.es) 3. Primary Care Centre Valls,

Spain.(cbayona.tarte.ics@gencat.cat). 4. Primary Care Centre Jaume I, University Rovira i Virgili, Tarragona, Spain.(carles.llor@urv.cat)

Conflicts of interest: MM reports receiving: honoraria for lectures from BayerSchering, Boehringer-Ingelheim, Pfizer, Nycomed, AstraZeneca, and Novartis, payment for development of educational presentations from Bayer-Schering, serving on the advisory boards of Bayer-Schering, Boehringer-Ingelheim, Pfizer, Nycomed,

GlaxoSmithKline, Almirall, AstraZeneca, and Novartis; and receiving consulting fees from Bayer-Schering, Boehringer-Ingelheim, Pfizer, Nycomed, GlaxoSmithKline, Almirall, AstraZeneca, and Novartis. CL reports receiving research grants from the European Commission (Sixth and Seventh Programme Frameworks), Catalan Society of

Downloaded From: http://journal.publications.chestnet.org/ by a Instituto Nacional de Enfermedades Respirat User on 07/02/2013

Page 3 of 31

Family Medicine, and Instituto de Salud Carlos III (Spanish Ministry of Health). AM reports receiving research grants from the Spanish Society of Family Medicine, Fundaci Jordi Gol i Gurina, and Instituto de Salud Carlos III (Spanish Ministry of Health). The other authors do not have conflicts of interest to disclose.

Correspondence:Marc Miravitlles Servei de Pneumologia. HospitalVall dHebron P. Vall dHebron 119-129, 08035 Barcelona, Spain e-mail: marcm@separ.es

Downloaded From: http://journal.publications.chestnet.org/ by a Instituto Nacional de Enfermedades Respirat User on 07/02/2013

Page 4 of 31

ABSTRACT Background: Anthonisens criteria are widely used to guide the need for antibiotics in exacerbations of chronic obstructive pulmonary disease (COPD). We evaluated the best predictors of outcomes in exacerbations of mild to moderate COPD not treated with antibiotics. Methods:We used data from 152 patients of the placebo arm of a randomised trial of amoxicillin/clavulanate for exacerbations of mild to moderate COPD. Clinical response in relation to Anthonisens criteria and point-of-test serum C-reactive protein (CRP) levels (cut off 40 mg/L) was assessed with multivariate logistic regression analysis. Results: Clinical failure without antibiotics was 19.9% compared to 9.5% with amoxicillin/clavulanate (p=0.022). The only factors significantly associated with an increased risk of failure without antibiotics were the increase in purulence of sputum (OR=6.1, 95% confidence interval: 1.5 to 25.0; p=0.005) and a CRP concentration >40 mg/L (OR=13.4, 95%CI: 4.6 to 38.8; p<0.001). When both factors were present, the probability of failure without antibiotics was 63.7%. The Anthonisen criteria showed an area under the curve (AUC) of 0.708 (95% CI: 0.616 - 0.801) for predicting clinical outcome. With the addition of CRP, the AUC rose significantly to 0.842 (95% CI: 0.76 0.924); p<0.001. Conclusion: Among the Anthonisen criteria, only an increase in sputum purulence is a significant predictor of failure without antibiotics. The use of a point-of-test CRP significantly increases the predictive accuracy of failure. Both of these easy to obtain factors may help clinicians to identify exacerbated mild to moderate COPD patients that can be safely treated without antibiotics in an ambulatory setting. Clinical Trial Registration number: clinicaltrials.gov; No.: NCT00495586

Downloaded From: http://journal.publications.chestnet.org/ by a Instituto Nacional de Enfermedades Respirat User on 07/02/2013

Page 5 of 31

Keywords: COPD; exacerbations; antibiotics; risk factors; clinical trial; C-reactive protein

ABBREVIATIONS AUC: Area under the curve CI: Confidence interval COPD: Chronic obstructive pulmonary disease CRP: C-reactive protein EOT: End of therapy FEV1: Forced expiratory volume in the first second FVC: Forced vital capacity L: Liter Mg: Milligrams OR: Odds ratio t.i.d: Three times a day (ter in die)

Downloaded From: http://journal.publications.chestnet.org/ by a Instituto Nacional de Enfermedades Respirat User on 07/02/2013

Page 6 of 31

INTRODUCTION The decision to prescribe an antibiotic for a patient with an exacerbation of COPD is one of the most frequent dilemmas encountered by physicians. This decision is usually taken with the aid of the combination of the three cardinal symptoms: increase in dyspnoea, in sputum volume and sputum purulence, described 25 years ago by Anthonisen et al (1) in a randomised placebo controlled trial in severe patients with exacerbated COPD (mean FEV1=33% predicted). These criteria have been extrapolated to all patients with COPD irrespective of the severity of airflow obstruction, but information about their predictive validity in mild to moderate patients (FEV1>50% predicted) is lacking (2). This is particularly important because most of the patients with exacerbations of COPD in the community are mild or moderate (3,4) and observational studies consistently show that most of these patients are treated with antibiotics irrespective of guidelines (5,6). Adequate antibiotic prescription is necessary to prevent the development of bacterial resistance and unnecessary side effects (7). The identification of clear symptoms or signs, or the description of objective tests that could reliably predict a good (or poor) outcome without antibiotics in exacerbations of mild to moderate COPD would have major clinical implications. We have recently completed a randomised, double-blind, placebo-controlled trial of 8 days of amoxicillin/clavulanate 500/125 mg t.i.din patients with acute exacerbations of mild to moderate COPD attended in primary care (8). The results indicate that treatment with amoxicillin/clavulanate is associated with a higher clinical success rate and a significantly longer period to the next exacerbation. However, up to 80% of patients were successfully treated with placebo. We used data from this trial to assess the best

Downloaded From: http://journal.publications.chestnet.org/ by a Instituto Nacional de Enfermedades Respirat User on 07/02/2013

Page 7 of 31

combination among the Anthonisen criteria and the point-of-care capillary C-reactive protein (CRP) test to predict clinical success with placebo.

METHOD The methods, design, and outcomes of the above-mentioned trial have been described previously in detail (8). In short, 310 patients were enrolled between October 2007 and July 2010 into a multicentre, double-blind, randomised placebo-controlled trial investigating the efficacy of amoxicillin/clavulanate 500/125 mg three times daily for 8 days in patients with exacerbations of mild to moderate COPD recruited in primary care. To investigate the predictors of failure without antibiotics, we analyzed data derived from the 152 patients included in the placebo arm. The study population consisted of patients that were at least 40 years of age, had a diagnosis of mild to moderate COPD (defined as having a smoking history of at least 10 pack-years, a ratio of post-bronchodilator FEV1 to FVC of <70% and a postbronchodilator FEV1>50% of the predicted value), and with an exacerbation defined as the occurrence of at least one of the following criteria: increase in dyspnoea, increase in sputum volume and/or sputum purulence. When the patients presented with the three symptoms, the exacerbations was classified as type I, with two symptoms as type II and with only one symptom as type III (1).The most relevant exclusion criteria were antibiotic use in the previous two weeks, bronchial asthma, active neoplasm, hospital admission, immunosuppression and hypersensitivity to beta-lactams, clavulanate or lactose. The study protocol was approved by the Research and Ethics Committee of Primary Care Fundaci Jordi Gol i Gurina (Barcelona, Spain; number: P6/031). Written informed consent was obtained from all the participants.

Downloaded From: http://journal.publications.chestnet.org/ by a Instituto Nacional de Enfermedades Respirat User on 07/02/2013

Page 8 of 31

Definition of clinical outcome Treatment success was defined as cure (a complete resolution of signs and symptoms associated with the exacerbation) or improvement (a resolution or reduction of the symptoms and signs without new symptoms and signs associated with the exacerbation). Clinical success was considered when either cure or improvement was observed (9). Failure was defined as incomplete resolution, persistence or worsening of symptoms that required a new course of antibiotics and/or oral corticosteroids or hospitalization (8). Evaluation was performed at the end of therapy visit (EOT) at days 9 to11. Serum C-reactive protein On the inclusion visit, a C-reactive protein (CRP) rapid test in capillary blood was performed using Quik Read CRP analyzers (Orion Diagnostica, Espoo, Finland) (10). Studies comparing this rapid test with the routine CRP laboratory test have shown a very good correlation thereby demonstrating its reliability (11). The best cut off for CRP serum concentrations to predict clinical success was set at 40 mg/L based on previous analysis (8). Statistical Analyses To identify the factors significantly associated with clinical failure without antibiotics, univariate and multivariate logistic regression analysis were performed. The first analysis was focused on investigating signs and symptoms of the exacerbation that could potentially identify episodes more likely to fail and included the three Anthonisen criteria: Increasing dyspnoea, increasing production of sputum and increasing purulence of sputum, together with fever >38, baseline peak flow and the presence of elevated CRP blood concentrations (>40 mg/L). The second model was focused on the risk

Downloaded From: http://journal.publications.chestnet.org/ by a Instituto Nacional de Enfermedades Respirat User on 07/02/2013

Page 9 of 31

factors of the patient that could be associated with an increased risk of clinical failure and included: age older than 65 years, sex, pack-years, ischaemic heart disease, cardiac insufficiency, diabetes and FEV1<65% predicted. Variables with a p value <0.1 in the univariate analysis were selected to be included in the model, and final selection of variables was performed by the backward stepwise selection analysis. The predictive value for clinical failure of the Anthonisen criteria alone and with the addition of CRP >40 mg/L was calculated by means of receiver operating characteristic (ROC) curves. Comparisons between the areas under ROC curves were assessed according to the method of Hanley and McNeil (12). All analyses were performed with the SPSS v.15 software (Chicago, USA).

RESULTS Baseline characteristics In the original randomised trial, 310 patients were enrolled. The placebo group of this trial, consisting of 152 patients, was used for the current study. Table 1 shows the baseline characteristics of the population treated with placebo according to the type of exacerbation (1). The mean age of the patients in the placebo group was 67.8 (SD=11) years, and 78.3% were male. The mean FEV1(% predicted) was 65.9% (SD=12.1%) and most patients had either a type II (47.4%) or a type I (29.6%) exacerbation. The median CRP level was 17 mg/L (Interquartile range 23 mg/L). Anthonisens criteria and prediction of failure without antibiotics. At the EOT visit, the overall clinical success rates for the antibiotic and placebo arm were 80.9% and 90.5% (p=0.022) and the clinical cure rates were 59.9% and 74.1%

Downloaded From: http://journal.publications.chestnet.org/ by a Instituto Nacional de Enfermedades Respirat User on 07/02/2013

Page 10 of 31

respectively (p=0.016). Considering the 9.5% failure rate with antibiotic, the aim was to identify factors that could predict a failure rate <10% in the placebo arm. Among patients not receiving antibiotics, those with type III exacerbations presented a failure rate of 2.8% (1/35), none of the 19 patients presenting only with increased dyspnoea failed. In general, patients with types II or III exacerbations presented a failure rate of 5.6% (3 out of 53) when no increase in purulence was one of the criteria; in contrast, all other patients had a failure rate >10%. When increased purulence was one of the criteria, types II and III exacerbations had a failure rate of 20.3% (11 out of 54), and type I exacerbations a failure rate of 33.3% (15 out of 45) (Figure 1). Table 2 shows the results of the univariate and multivariate analysis of the association of Anthonisens criteria with failure. Only an increase in sputum purulence was significantly associated with failure in both analyses. CRP and Anthonisens criteria for prediction of failure without antibiotics Previous analysis of our data based on ROC curves indicated that a cut off of 40 mg/L provided the best prediction for clinical failure in the whole population (8). In the placebo arm, a total of 77.3% of CRP determinations presented concentrations lower than 40 mg/L. In these cases, the failure rate was 12.4% (15 out of 121), being significantly lower than the 65.5% of failure rate observed among the 34 patients with CRP40 mg/L (P< 0.001). In the multivariate analysis including the Anthonisen criteria together with the CRP test results, only an increase in purulence and CRP concentrations >40mg/L were significant predictors of failure. The ROC curve analysis for Anthonisens criteria provided an area under the curve (AUC) of 0.708 (95% confidence interval (CI): 0.616-0.801) that was significantly improved when the CRP measurement was included in the model [AUC=

Downloaded From: http://journal.publications.chestnet.org/ by a Instituto Nacional de Enfermedades Respirat User on 07/02/2013

Page 11 of 31

10

0.842 (95%CI: 0.760-0.924), p=0.033] (figure 2). In logistic regression analysis the probability of failure without antibiotics was only 2.7% when CRP was <40 mg/L and no increase in purulence was present. This probability rose to 12.4% when only purulence was present, to 25.6% when only CRP was >40 mg/L and finally to 63.7% when both factors were present. Other risk factors of failure without antibiotics On analyzing the influence of other risk factors of the patients in the evolution of the exacerbation in patients not treated with antibiotics, we observed that only FEV1(%) < 65% was independently associated with an increased risk of failure in the multivariate analysis (OR= 3.0, 95%CI: 1.3 to 7.2; p=0.012) (table 3).

DISCUSSION The analysis of the results of the placebo arm of a randomised, double-blind antibiotic trial in exacerbated mild to moderate COPD provided a unique opportunity to investigate the factors associated with clinical outcomes and identify the patients that can safely be treated without antibiotics in an ambulatory setting (8). We revisited the classical Anthonisens criteria for antibiotic treatment (1) and demonstrated that only an increase in purulence of sputum was associated with a significant increase in the risk of failure without antibiotics, and the use of point-of-care testing of CRP significantly increased the predictive accuracy for the need of antibiotics in exacerbations of mild to moderate ambulatory COPD patients. The previously published results of our clinical trial showed the superiority of amoxicillin/clavulanate compared with placebo in terms of significantly higher clinical cure and success rates compared with placebo. However, it was of note that up to 80%

Downloaded From: http://journal.publications.chestnet.org/ by a Instituto Nacional de Enfermedades Respirat User on 07/02/2013

Page 12 of 31

11

of patients were successfully treated with placebo (8). It has been widely recognized that not all exacerbations of COPD require antibiotic treatment (1-3,13), and that excessive and unnecessary antibiotic use is causing an increase in bacterial resistance (7). Moreover, the microbiological analysis of respiratory samples to direct antibiotic therapy in ambulatory setting is usually not available (3). In this context, for the last 25 years, the clinical criteria described by Anthonisen et al (1) have been incorporated in guidelines to help in selecting patients that require empiric antibiotic therapy (14,15). However, these criteria were identified in a population of severe and very severe COPD with a mean FEV1 of only 33% predicted. Therefore, validation of these criteria or identification of new criteria for using (or not) antibiotics in mild to moderate ambulatory patients is still required (2). More recent studies have identified a change in colour or an increase in purulence as a good surrogate marker for the presence of bacteria in sputum (16,17) or in bronchoalveolar lavage (18) of patients with exacerbations of COPD. Our results have shown that an increase in purulence of sputum was the only sign or symptom associated with the exacerbation that was significantly and independently related to a worse outcome in patients not receiving antibiotics, whereas an increase in dyspnoea and in sputum volume were not. This clear difference between the three cardinal symptoms also confirms the reliable identification of the increase in purulence by primary care physicians. In this study no colour chart or specific instructions were provided to the investigators, they had to fill in the case report form according to their understanding of the meaning of an increase in purulence of sputum, as in real clinical practice. Interestingly, when no increase in purulence was documented, the rate of probability of failure without antibiotics was always below the 10% observed in the group treated with antibiotics.

Downloaded From: http://journal.publications.chestnet.org/ by a Instituto Nacional de Enfermedades Respirat User on 07/02/2013

Page 13 of 31

12

Despite the good predictive value of the increase in purulence, it would be of interest to investigate if any objective and easy to obtain measurement could improve the discriminative power for identifying subjects that can be safely managed without antibiotics. We used the point-of-care testing of CRP that has demonstrated to be a useful strategy for improving the prescription of antibiotics in lower respiratory tract infections (19,20) and a good tool to improve the adherence of patients to antibiotic therapy(21). The CRP has demonstrated to be the best biomarker to differentiate between stable and exacerbated COPD, although its accuracy is poor (22), and CRP concentrations decrease significantly with recovery from exacerbations (23). Moreover, persistently elevated CRP levels are associated with recurrence of exacerbations in COPD (24). More interestingly, previous studies observed that CRP levels were significantly increased in purulent compared to mucoid exacerbations (25), and finally, increased CRP concentrations have been linked to the presence of bacteria in sputum in exacerbations of COPD (25-29). Increased CRP has also been observed in association with the acquisition of a new bacterial strain in patients with COPD (30), a mechanism associated with the development of a new exacerbation (31). Up to four different phenotypes of exacerbations in COPD have been described, and the best serum biomarker for the phenotype of bacterial exacerbations was a high CRP level (29). In contrast, in the same study, increased CRP levels were not observed in the phenotype of viral exacerbations (29). The same observation of the lack of increase (or minor increase) in CRP levels in viral exacerbations has been reported previously (32,33). In accordance with our results, Daniels et al (28) demonstrated that CRP levels (but not procalcitonin levels) predicted the response to antibiotics in exacerbations of severe COPD, with increasing effects of antibiotics compared to placebo in patients with

Downloaded From: http://journal.publications.chestnet.org/ by a Instituto Nacional de Enfermedades Respirat User on 07/02/2013

Page 14 of 31

13

higher CRP levels. We have extended their findings by analyzing the predictive value of CRP levels in patients not treated with antibiotics and observed that low CRP levels were associated with success without antibiotics. Another important difference between these studies is the severity of the population studied, while all our patients had an FEV1(%)>50%, only 30% of the patients included in the former study had an FEV1(%)>50% (28). This difference is relevant because severity of airflow obstruction is a risk factor of poor outcome in ambulatory exacerbations in COPD (34); furthermore, bacterial aetiology of exacerbations is related to the degree of airflow obstruction (35). Nevertheless, most guidelines have extrapolated the results of the previous studies in severe or very severe patients to establish their recommendations of antibiotic treatment including mild to moderate COPD patients (14,15). Our study has some limitations. We could not investigate the relationship between symptoms and CRP levels and the presence of bacteria in sputum, because microbiological analysis of sputum was not readily available in the participating primary care centers. However, our results are consistent with previous studies that analyzed the presence of bacteria in sputum related to purulence and CRP concentrations (16-18,27-29).We cannot rule out some placebo effect because the criteria for success are based on clinical evaluation. However, the significant relationship between failure and increasing purulence, higher CRP and worse FEV1(%) suggest that most of the clinical assessments were correct. Finally, there was not a protocol for the administration of comedications, but the low percentage of use of oral corticosteroids, similar to that observed in other series in our country (5,6) most likely did not influence our results. In summary, ambulatory patients with exacerbations of mild to moderate COPD (FEV1>50% predicted) can be safely treated without antibiotics when no increase in

Downloaded From: http://journal.publications.chestnet.org/ by a Instituto Nacional de Enfermedades Respirat User on 07/02/2013

Page 15 of 31

14

purulence is present, and the point of test CRP is <40mg/L (probability of failure 2.7%). In cases where point of test CRP is not available, the presence of the cardinal symptoms of an increase in dyspnoea, an increase in sputum volume or both (without an increase in the purulence of sputum) is associated with a probability of failure <8% without antibiotics, which is inferior to the 10% failure rate observed for the whole population treated with antibiotics. These simple rules may help physicians in their empirical choice of therapy in this frequent group of patients.

ACKNOWLEDGEMENTS We thank all the study coordinators, the study investigators in the different primary healthcare centres and the nurses who performed the spirometric studies. We also thank Albert Gabarrs (IDIBAPS, Hospital Clnic, Barcelona) for his assistance in the analysis of data. Contributors: MM, CL and AM designed the study. CL and AM coordinated the trial. AM, CL and SH were responsible for the recruitment and follow-up of the participants. CB supervised the packaging and labelling of the study medication and contributed to the data analysis. MM and CL wrote the first draft of the article. All the authors had access to the data and contributed substantially to the submitted report. MM is the guarantor of the content of the manuscript, including the data and the analysis. Funding: This clinical trial was supported by a grant from the Instituto de Salud Carlos III (Spanish Ministry of Health) (EC07/90333).They had no role in the design, conduct, or analysis of this study. All authors had full access to all the data in the study and had final responsibility for the decision to submit the publication.

Downloaded From: http://journal.publications.chestnet.org/ by a Instituto Nacional de Enfermedades Respirat User on 07/02/2013

Page 16 of 31

15

REFERENCES 1. Anthonisen NR, Manfreda J, Warren CPW, Hershfield ES, Harding GKM, Nelson NA. Antibiotic therapy in exacerbations of chronic obstructive pulmonary disease. Ann Intern Med 1987;106:196204.

2. Puhan MA, Vollenweider D, Steurer J, Bossuyt P, terRiet G. Where is the supporting evidence for treating mild to moderate chronic obstructive pulmonary disease patients exacerbations with antibiotics?: a systematic review. BMC Medicine 2008,6:28.

3. Miravitlles M, Anzueto A, Ewig S, Legnani D, Stauch K. Characterization of exacerbations of chronic bronchitis and COPD in Europe. The GIANT study. Ther Adv Respir Dis 2009; 3: 267-277.

4. Llor C, Molina J, Naberan K, Cots JM, Ros F, Miravitlles M, on behalf of the EVOCA study. Exacerbations worsen the quality of life of chronic obstructive pulmonary disease patients in primary healthcare. Int J Clin Pract 2008; 62: 585-592.

5. Miravitlles M, Mayordomo C, Arts M, Snchez-Agudo L, Nicolau F, Seg JL onBehalf of the EOLO Group. Treatment of chronic obstructive pulmonary disease and its exacerbations in General Practice. Respir Med 1999; 93: 173-179.

Downloaded From: http://journal.publications.chestnet.org/ by a Instituto Nacional de Enfermedades Respirat User on 07/02/2013

Page 17 of 31

16

6. Murio C, Soler X, Prez M, Calero G, Ruiz-Manzano J. Acute exacerbations of chronic

obstructive pulmonary disease in primary care in Spain: the EPOCAP study. Ther Adv Respir Dis 2010; 4: 215-223.

7. Goosens H, Ferech M, Stichele RV, Elseviers M, for the ESAC Project Group. Outpatient antibiotic use in Europe and association with resistance: a cross-national database study. Lancet 2005; 365: 579-587.

8. Llor C, Moragas A, Hernndez S, Bayona C, Miravitlles M. Efficacy of antibiotic therapy for acute exacerbations of mild to moderate COPD. Am J Respir Crit Care Med 2012; 186: 716-723.

9. Chow AW, Hall CB, Klein JO, Kammer RB, Meyer RD, Remington JS. Evaluation of new anti-infective drugs for the treatment of respiratory tract infections.Infectious Diseases Society of America and the Food and Drug Administration. Clin Infect Dis 1992;15:S6288.

10. Seamark DA, Backhouse SN, Powell R. Field-testing and validation in a primary care setting of a point-of-care test for C-reactive protein. Ann Clin Biochem 2003;40(Pt 2):178180.

Downloaded From: http://journal.publications.chestnet.org/ by a Instituto Nacional de Enfermedades Respirat User on 07/02/2013

Page 18 of 31

17

11. Esposito S, Tremolati E, Begliatti E, Bosis S, Gualtieri L, Principi N. Evaluation of a rapid bedside test for the quantitative determination of C-reactive protein. Clin Chem Lab Med 2005;43:438440.

12. Hanley JA, McNeil BJ. A method of comparing the areas under receiver operating characteristic curves derived from the same cases. Radiology 1983; 148: 839-43.

13. Puhan MA, Vollenweider D, Latshang T, Steurer J, Steurer-Stey C. Exacerbations of chronic obstructive pulmonary disease: when are antibiotics indicated? A systematic review. Respir Res 2007;8:30.

14. Woodhead M, Blasi F, Ewig S, et al. Joint Taskforce of the European Respiratory Society and European Society for Clinical Microbiology and Infectious Diseases. Guidelines for the management of adult lower respiratory tract infections full version. Clin Microbiol Infect 2011;17(Suppl 6):E159.

15. Miravitlles M, Mons E, Mensa J, et al. Antimicrobial treatment of exacerbation in chronic obstructive pulmonary disease: 2007 consensus statement. Arch Bronconeumol 2008;44:100108.

16. Stockley RA, O'Brien C, Pye A, Hill SL. Relationship of sputum color to nature and outpatient management of acute exacerbations of COPD. Chest 2000;117:1638-1645.

Downloaded From: http://journal.publications.chestnet.org/ by a Instituto Nacional de Enfermedades Respirat User on 07/02/2013

Page 19 of 31

18

17. Miravitlles M, Kruesmann F, Haverstock D, Perroncel R, Choudhri S, Arvis P. Sputum colour and bacteria in chronic bronchitis exacerbations: a pooled analysis. Eur Respir J 2012; 39: 1354-1360.

18. Soler N, Agust C, Angrill J, Puig de la Bellacasa J, Torres A. Bronchoscopic validation of the significance of sputum purulence in severe exacerbations of chronic obstructive pulmonary disease. Thorax 2007;62:29-35.

19. Llor C, Cots JM, Lpez-Valcarcel BG, et al. Interventions to reduce antibiotic prescription for lower respiratory tract infections: Happy Audit Study. Eur Respir J 2012; 40: 436-441.

20. Cals JWL, Butler CC, Hopstaken RM, Hood K, Dinant GJ. Effect of point of care testing for C reactive protein and training in communication skills on antibiotic use in lower respiratory tract infections: cluster randomized trial. BMJ 2009; 338: b1374 doi:10.1136/bmj.b1374.

21. Llor C, Sierra N, Hernndez S, et al. Impact of C-reactive protein testing on adherence to thrice-daily antibiotic regimens in lower respiratory tract infections. Prim Care Respir J 2010; 19: 358-362.

Downloaded From: http://journal.publications.chestnet.org/ by a Instituto Nacional de Enfermedades Respirat User on 07/02/2013

Page 20 of 31

19

22.Hurst JR, Donaldson GC, Perera WR, et al. Use of plasma biomarkers at exacerbation of chronic obstructive pulmonary disease. Am J Respir Crit Care Med 2006;174:867874.

23. Koutsokera A, Kiropoulos TS, Nikoulis DJ, et al. Clinical, functional and biochemical changes during recovery from COPD exacerbations. Respir Med 2009; 103: 919-926.

24. Perera WR, Hurst JR, Wilkinson TMA, et al. Inflammatory changes, recovery and recurrence at COPD exacerbations. Eur Respir J 2007; 29: 527-534.

25.Gompertz S, OBrien C, Bayley DL, Hill SL, Stockley RA. Changes in bronchial inflammation during acute exacerbations of chronic bronchitis. Eur Respir J 2001; 17: 1112-1119.

26. Weis N, Almdal T. C-reactive protein-can it be used as a marker of infection in patients with exacerbations of chronic obstructive pulmonary disease? Eur J Intern Med 2006; 17: 88-91.

27. Dev D, Wallace E, Sankaran R, Cunniffe J, Govan JRW, Wathen CG, Emmanuel FXS. Value of C-reactive protein measurements in exacerbations of chronic obstructive pulmonary disease. Respir Med 1998; 92: 664-667.

Downloaded From: http://journal.publications.chestnet.org/ by a Instituto Nacional de Enfermedades Respirat User on 07/02/2013

Page 21 of 31

20

28. Daniels JMA, Schoorl M, Snijders D, et al. Procalcitoninvs C-reactive protein as predictive markers of response to antibiotic therapy in acute exacerbations of COPD. Chest 2010; 138: 1108-1115.

29. Bafadhel M, McKenna S, Terry S, et al. Acute exacerbations of chronic obstructive pulmonary disease. Identification of biologic clusters and their biomarkers. Am J Respir Crit Care Med 2011; 184: 662-671.

30.Sethi S, Wrona C, Eschberger K, Lobbins P, Cai X, Murphy TF. Inflammatory profile of new bacterial strain exacerbations of chronic obstructive pulmonary disease. Am J Respir Crit Care Med 2008;177:491-497.

31.Sethi S, Evans N, Grant BJ, Murphy TF. New strains of bacteria and exacerbations of chronic obstructive pulmonary disease. N Engl J Med 2002;347:465-471.

32. Rohde G, Borg I, Wiethege A, et al. Inflammatory response in acute viral exacerbations of COPD. Infection 2008; 36: 427-433.

33. Kherad O, Kaiser L, Bridevaux PO, et al. Upper-respiratory viral infection, biomarkers, and COPD exacerbations. Chest 2010; 138: 896-904.

Downloaded From: http://journal.publications.chestnet.org/ by a Instituto Nacional de Enfermedades Respirat User on 07/02/2013

Page 22 of 31

21

34. Miravitlles M, Guerrero T, Mayordomo C, Snchez-Agudo L, Nicolau F, Seg JL onBehalf of the EOLO StudyGroup. Factors associated with increased risk of exacerbation and hospital admission in a cohort of ambulatory COPD patients: a multiple logistic regression analysis. Respiration 2000; 67: 495-501.

35. Miravitlles M, Espinosa C, Fernndez-Laso E, Martos JA, Maldonado JA, Gallego M and Study Group of Bacterial Infection in COPD. Relationship between bacterial flora in sputum and functional impairment in patients with acute exacerbations of COPD. Chest 1999; 116: 40-46.

Downloaded From: http://journal.publications.chestnet.org/ by a Instituto Nacional de Enfermedades Respirat User on 07/02/2013

Page 23 of 31

22

TABLE 1. Baseline characteristics of the population treated without antibiotics according to the exacerbation type. Type I n= 45 Age (years), mean (SD) Male sex, n (%) Smoking status: Current, n (%) Former, n (%) 30 (66.7) 15 (33.3) 41.8 (13.7) 24 (53.3) 10 (22.2) 8 (17.8) 2552.0 (865.0) 69.4 (17.9) 1523.1 (574.9) 62.6 (11.2) 59.5 (5.8) 43 (59.7) 29 (40.3) 36.1 (21.0) 33 (45.8) 12 (16.7) 7 (9.7) 2961.5 (1017.2) 73.9 (18.6) 1879.0 (673.5) 68.1 (13.4) 63.5 (5.8) 16 (45.7) 19 (54.3) 36.4 (21.0) 16 (45.7) 6 (17.1) 4 (11.4) 2627.7 (874.4) 68.9 (18.2) 1656.0 (586.3) 65.5 (9.5) 63.0 (4.7) 0.260 0.697 0.734 0.429 0.049 0.280 0.010 0.056 0.001 0.162 71.3 (8.7) 41 (91.1) Type II n = 72 65.9 (11.8) 52 (72.2) Type III n = 35 68.1 (10.7) 26 (74.3) 0.058 0.044 p

Pack-years, mean (SD) High blood pressure, n (%) Diabetes mellitus, n (%) Coronary heart disease, n (%) FVC (ml), mean (SD) FVC (%), mean (SD) FEV1 (ml), mean (SD) FEV1 (%), mean (SD) FEV1/FVC ratio, mean (SD) Treatment of the exacerbation: Short-acting (%) Oral corticosteroids, n (%) -agonists, n

20 (44.4)

19 (26.4)

14 (40.0)

0.105

15 (33.3) 42.0 (36.5)

9 (12.5) 21.9 (25.5)

3 (8.6) 20.7 (18.8)

0.004 <0.001

CRP, median (IQR)

CRP: C-reactive protein; IQR: interquartile range

Downloaded From: http://journal.publications.chestnet.org/ by a Instituto Nacional de Enfermedades Respirat User on 07/02/2013

Page 24 of 31

23

Type I: all the Anthonisen criteria present (increased dyspnoea, increased sputum volume and

purulent sputum); type II: only two criteria present; type III: only one criterion present

Downloaded From: http://journal.publications.chestnet.org/ by a Instituto Nacional de Enfermedades Respirat User on 07/02/2013

Page 26 of 31

25

Table 2. Univariate and multivariate logistic regression analysis of factors related to the exacerbation that predict clinical failure of exacerbations of mild to moderate COPD not treated with antibiotics. Variable OR Increased dyspnoea Increased volume of sputum Increased purulence of sputum CRP >40 mg/L 1.6 2.1 5.9 13.4 Univariate CI95% 0.6-3.8 0.7-6.5 1.7-20.7 5.3-34.3 P value 0.32 0.20 0.005 <0.001 OR 2.3 1.8 6.3 NA Multivariate CI95% 0.9-5.9 0.6-6.1 1.8-22.5 NA P value 0.078 0.32 0.005 NA Multivariate with CRP OR 1.3 0.6 6.1 13.4 CI95% 0.4-3.9 0.2-2.4 1.5-25.0 4.6-38.8 P value 0.32 0.20 0.005 <0.001

Footnote: CRP: C-reactive protein; NA= Not assessed

Downloaded From: http://journal.publications.chestnet.org/ by a Instituto Nacional de Enfermedades Respirat User on 07/02/2013

Page 28 of 31

27

Table 3. Univariate and multivariate logistic regression analysis of factors related to the patient that predict clinical failure of exacerbations of mild to moderate COPD not treated with antibiotics. Variable OR Pack-years (>20) High blood pressure Coronary heart disease FEV1(%) < 65% 1.4 2.0 2.9 3.0 Univariate CI95% 0.9-2.1 0.9-4.6 1.0-8.3 1.3-7.2 P value 0.094 0.096 0.042 0.012 3.0 1.3-7.2 0.012 OR Multivariate CI95% P value

Downloaded From: http://journal.publications.chestnet.org/ by a Instituto Nacional de Enfermedades Respirat User on 07/02/2013

Page 29 of 31

28

FIGURES

Figure 1. Percentage of failure rates in exacerbations of mild to moderate COPD not treated with antibiotics, according to Anthonisens criteria.

Figure 2. ROC curves showing the predictive value for clinical failure of a) Anthonisens criteria (dotted line) and b) with the addition of CRP 40 mg/l (continuous line), among mild-to-moderate COPD patients with exacerbations not treated with antibiotics.

Footnote: a) Anthonisens criteria AUC= 0.708 (95% confidence interval (CI): 0.6160.801); b) with the addition of CRP 40 mg/l AUC= 0.842 (95%CI: 0.760-0.924). Differences between curves were significant at p=0.033.

Downloaded From: http://journal.publications.chestnet.org/ by a Instituto Nacional de Enfermedades Respirat User on 07/02/2013

Page 30 of 31

254x190mm (72 x 72 DPI)

Downloaded From: http://journal.publications.chestnet.org/ by a Instituto Nacional de Enfermedades Respirat User on 07/02/2013

Page 31 of 31

254x190mm (72 x 72 DPI)

Downloaded From: http://journal.publications.chestnet.org/ by a Instituto Nacional de Enfermedades Respirat User on 07/02/2013