The proportion of sputum samples obtained at each visit (confirmed stable or AECOPD visits) and positive for specific bacterial pathogens by bacteriological methods (overall and by bacterial species). Numerator is the number of sputum samples positive for a given pathogen and denominator is the number of visits with a sputum sample tested for a given pathogen. Proportion is computed with 95% confidence intervals. A confirmed stable visit is defined as a scheduled study visit for which the investigator confirms that the subject is stable/ has recovered from a previous exacerbation. Bacterial pathogens include Haemophilus influenzae (Hi), Non-typeable Haemophilus influenzae (NTHi), non-Haemophilus influenzae (Non-Hi), Moraxella catarrhalis (M. catarrhalis), Streptococcus pneumoniae (S. pneumoniae), Staphylococcus aureus (S. aureus), Pseudomonas aeruginosa (P. aeruginosa), Klebsiella pneumoniae (K. pseumoniae), Acinetobacter baumannii (A. baumannii).

Analysis was performed on the Full Analysis Set which included all screened subjects who were enrolled except for those subjects who discontinued the study at Month 0 visit.

The proportion of sputum samples obtained at each visit (confirmed stable visits and AECOPD visits) and positive for specific viral pathogens by PCR (overall and by viral species). The numerator is the number of sputum samples positive for a given pathogen and the denominator is the number of visits with a sputum sample tested for a given pathogen. The proportion is calculated with 95% confidence intervals. Viral pathogens, as identified by PCR, include respiratory syncytial virus (RSV), parainfluenza virus, enterovirus, human rhinovirus (HRV), metapneumovirus, influenza virus, adenovirus, bocavirus and coronavirus.

Analysis was performed on the Full Analysis Set which included all screened subjects who were enrolled except for those subjects who discontinued the study at Month 0 visit.

The proportion of sputum samples obtained at each confirmed stable/AECOPD visit and positive for specific bacterial pathogens as measured by real-time qualitative PCR/quantitative PCR, (overall and by bacterial species,) are computed with 95% confidence intervals. The numerator is the number of sputum samples positive for a given pathogen and the denominator is the number of visits with a sputum sample tested for a given pathogen. The bacterial pathogens include H. influenzae, M. catarrhalis, S. pneumoniae, S. aureus, P. aeruginosa and S. pyogenes. A confirmed stable visit is defined as a scheduled study visit for which the investigator confirms that the subject is stable / has recovered from a previous exacerbation.

Analysis was performed on the Full Analysis Set which included all screened subjects who were enrolled except for those subjects who discontinued the study at Month 0 visit.

Concordance between bacteriological methods (culture) and PCR sputum results are described for all the combinations of bacterial presence by both measures. Each category name includes the following parameters: Bacteria species-Culture (yes/no)- PCR (yes/no). Concordance is expressed as the number of sputum samples in a given category among the total number of sputum samples assessed for the presence of bacterial pathogens by both culture and PCR

Analysis was performed on the sputum samples taken from Full Analysis Set which included all screened subjects who were enrolled except for those subjects who discontinued the study at Month 0 visit.

Number of sputum samples obtained at each confirmed stable visit, and positive for bacterial pathogens by bacteriological methods (overall and by bacterial species). The spirometric classification of airflow limitation in COPD patients is based on post-bronchodilator forced expiratory volume in 1 second (FEV1) and can be divided into four GOLD grades [GOLD, 2013]: GOLD 1 (MILD): FEV1 ≥ 80% predicted; GOLD 2 (Moderate)= 50% ≤ FEV1 < 80% predicted; GOLD 3 (Severe) = 30% ≤ FEV1 < 50% predicted; GOLD 4 (Very Severe) = FEV1 < 30% predicted.

Analysis was performed on the sputum samples taken from Full Analysis Set which included all screened subjects who were enrolled except for those subjects who discontinued the study at Month 0 visit.

Number of sputum samples obtained at each confirmed stable visit, and positive for bacterial pathogens by PCR (overall and by bacterial species). The spirometric classification of airflow limitation in COPD patients is based on post-bronchodilator forced expiratory volume in 1 second (FEV1) and can be divided into four GOLD grades [GOLD, 2013]: GOLD 1 (MILD): FEV1 ≥ 80% predicted; GOLD 2 (Moderate)= 50% ≤ FEV1 < 80% predicted; GOLD 3 (Severe) = 30% ≤ FEV1 < 50% predicted; GOLD 4 (Very Severe) = FEV1 < 30% predicted.

Analysis was performed on the sputum samples taken from Full Analysis Set which included all screened subjects who were enrolled except for those subjects who discontinued the study at Month 0 visit.

Number of sputum samples obtained at each confirmed stable visit, and positive for virus pathogens by PCR (overall and by viral species). The spirometric classification of airflow limitation in COPD patients is based on post-bronchodilator forced expiratory volume in 1 second (FEV1) and can be divided into four GOLD grades [GOLD, 2013]: GOLD 1 (MILD): FEV1 ≥ 80% predicted; GOLD 2 (Moderate)= 50% ≤ FEV1 < 80% predicted; GOLD 3 (Severe) = 30% ≤ FEV1 < 50% predicted; GOLD 4 (Very Severe) = FEV1 < 30% predicted.

Analysis was performed on the sputum samples taken from Full Analysis Set which included all screened subjects who were enrolled except for those subjects who discontinued the study at Month 0 visit.

Number of sputum samples obtained at each AECOPD visit, and positive for bacterial pathogens by bacteriological methods (overall and by bacterial species). Classification of severity of AECOPD as follows: Mild = controlled with an increase in dosage of regular medications; Moderate = requires treatment with systemic corticosteroids and/or antibiotics; Severe = requires hospitalization.

Analysis was performed on the sputum samples taken from Full Analysis Set which included all screened subjects who were enrolled except for those subjects who discontinued the study at Month 0 visit.

Number of sputum samples obtained at each AECOPD visit, and positive for bacterial pathogens by PCR (overall and by bacterial species). Classification of severity of AECOPD is as follows: Mild = controlled with an increase in dosage of regular medications; Moderate = requires treatment with systemic corticosteroids and/or antibiotics; Severe = requires hospitalization.

Analysis was performed on the sputum samples taken from Full Analysis Set which included all screened subjects who were enrolled except for those subjects who discontinued the study at Month 0 visit.

Number of sputum samples obtained at each AECOPD visit, and positive for viral pathogens by PCR (overall and by viral species). Classification of severity of AECOPD is as follows: Mild = controlled with an increase in dosage of regular medications; Moderate = requires treatment with systemic corticosteroids and/or antibiotics; Severe = requires hospitalization.

Analysis was performed on the sputum samples taken from Full Analysis Set which included all screened subjects who were enrolled except for those subjects who discontinued the study at Month 0 visit.

Incidence rate is estimated by the mean number of exacerbations per subject and per year from Negative Binomial model (or Poisson model in case of under dispersion) without covariates and computed with 95% confidence intervals (CI). Confirmed AECOPDs include AECOPD events plus missed AECOPD events (i.e.: all morning alerts confirmed by phone call (as well as cases with no morning alert) for which there has been no site visit but for which AECOPD medical records are available). Potential AECOPDs include all morning alert confirmed by phone call for which there has been no site visit and for which no medical records are available.

Analysis was performed on the Full Analysis Set which included all screened subjects who were enrolled except for those subjects who discontinued the study at Month 0 visit.

Classification of severity of AECOPD is as follows: Mild- Controlled AECOPD with an increase in dosage of regular medications; Moderate- Requires treatment with systemic corticosteroids and/ or antibiotics; Severe- Requires hospitalisation.

Analysis was performed on the Full Analysis Set which included all screened subjects who were enrolled except for those subjects who discontinued the study at Month 0 visit.

The spirometric classification of airflow limitation in COPD patients is based on post-bronchodilator forced expiratory volume in 1 second (FEV1) and can be divided into four GOLD grades [GOLD, 2013]: GOLD 1 (MILD): FEV1 ≥ 80% predicted; GOLD 2 (Moderate)= 50% ≤ FEV1 < 80% predicted; GOLD 3 (Severe) = 30% ≤ FEV1 < 50% predicted; GOLD 4 (Very Severe) = FEV1 < 30% predicted.

Analysis was performed on the Full Analysis Set which included all screened subjects who were enrolled except for those subjects who discontinued the study at Month 0 visit.

Descriptive statistics (mean, standard deviation) on the number of days of AECOPD episodes are presented, overall and by AECOPD severity.

Analysis was performed on the Full Analysis Set which included all screened subjects who were enrolled except for those subjects who discontinued the study at Month 0 visit.

The CAT is a patient-completed instrument to assess the heath-related quality of life (HRQOL) and symptom burden in patients with COPD. Descriptive statistics (mean, standard deviation) on the CAT scores are tabulated at each stable visit. The CAT index is derived as the sum of the ratings recorded for each of the eight individual items. Each of these items has 6 possible scores (0, 1, 2, 3, 4 or 5), leading to a range of 0 (best score) to 40 (worst score) for CAT score.

Analysis was performed on the Full Analysis Set which included all screened subjects who were enrolled except for those subjects who discontinued the study at Month 0 visit.

The CAT is a patient-completed instrument to assess the HRQOL and symptom burden in patients with COPD. Descriptive statistics (mean, standard deviation) on the CAT scores are tabulated by frequency of exacerbations). The CAT index is derived as the sum of the ratings recorded for each of the eight individual items. Each of these items has 6 possible scores (0, 1, 2, 3, 4 or 5), leading to a range of 0 (best score) to 40 (worst score) for CAT score.

Analysis was performed on the Full Analysis Set which included all screened subjects who were enrolled except for those subjects who discontinued the study at Month 0 visit.

The SGRQ-C is designed to assess HRQOL and current health of the patients. Descriptive statistics (mean, standard deviation) on the SGRQ-C scores are tabulated at each stable visit. The SGRQ-C total score is derived as the weighted sum of the forty individual items leading to a range of 0 (best score) to 100 (worst score) as detailed in the reference manual [St George's Respiratory Questionnaire for COPD patients, version 1.3, 2016].

Analysis was performed on the Full Analysis Set which included all screened subjects who were enrolled except for those subjects who discontinued the study at Month 0 visit.

Summary statistics (mean, standard deviation) on post bronchodilator FEV1% of predicted normal value is tabulated at enrolment and final visit.

Analysis was performed on the Full Analysis Set which included all screened subjects who were enrolled except for those subjects who discontinued the study at Month 0 visit.

Healthcare utilization includes all unscheduled visits to a physician office, visits to urgent care, visits to emergency department, and hospitalizations. The impact of AECOPD on healthcare utilization is assessed during the stable periods. Hospitalizations that were associated with the disease being studied were not collected as Adverse Events (AEs) or as serious AEs (SAEs) as per protocol.

Analysis was performed on the Full Analysis Set which included all screened subjects who were enrolled except for those subjects who discontinued the study at Month 0 visit.

Healthcare utilization includes all unscheduled visits to a physician office, visits to urgent care, visits to emergency department, and hospitalizations. The impact of AECOPD on healthcare utilization is assessed during exacerbation periods. Hospitalizations that were associated with the disease being studied were not collected as AEs or as SAEs as per protocol.

Analysis was performed on the Full Analysis Set which included all screened subjects who were enrolled except for those subjects who discontinued the study at Month 0 visit.