Major health coach activities:

Patient COPD education; Correct use of inhalers and nebulizers; Identifying red flags and when to seek medical care; Dyspnea management; Patient decision making and action plans around, exercise, smoking cessation, nutrition, exacerbations; Ensuring appropriate preventive services (pneumovax, flu); Depression screening; Reinforcing clinician education and use of treatment guidelines by primary care providers; Helping patient obtain prescriptions; Identifying gaps in care, areas where care not in line with care plan; Helping patients to make and keep appointments and obtain needed testing; Facilitating communication between patients, pulmonary specialists and primary care providers; Connecting with community resources; Helping to access to psychosocial services as needed; Conducting exercise capacity assessment and working with pulmonary specialist to provide recommended exercise program; Working with patient family members a

Usual care includes access to specialist consultation via referral by the primary care clinician, access to patient education classes, smoking cessation classes, psychosocial medicine and nutritional counseling.

Outcomes were compared by group assignment using generalized linear models (GLMs) with a normal distribution for continuous outcomes), Poisson distribution for count outcomes (e.g. exacerbations and hospitalizations), and binomial distribution for binary outcomes. In all models, baseline levels of the outcome were entered as a predictor and follow-up levels as the dependent variable, with use of a robust standard error to account for clustering.

Outcomes were by group assignment (intention to treat) using generalized linear models adjusted for baseline levels of variable and clustering.

Value is for mean score of participants in Health Coached arm minus mean score in Usual Care, adjusted for baseline values and for clustering.

Major health coach activities:

Patient COPD education; Correct use of inhalers and nebulizers; Identifying red flags and when to seek medical care; Dyspnea management; Patient decision making and action plans around, exercise, smoking cessation, nutrition, exacerbations; Ensuring appropriate preventive services (pneumovax, flu); Depression screening; Reinforcing clinician education and use of treatment guidelines by primary care providers; Helping patient obtain prescriptions; Identifying gaps in care, areas where care not in line with care plan; Helping patients to make and keep appointments and obtain needed testing; Facilitating communication between patients, pulmonary specialists and primary care providers; Connecting with community resources; Helping to access to psychosocial services as needed; Conducting exercise capacity assessment and working with pulmonary specialist to provide recommended exercise program; Working with patient family members a

Usual care includes access to specialist consultation via referral by the primary care clinician, access to patient education classes, smoking cessation classes, psychosocial medicine and nutritional counseling.

Outcomes were compared by group assignment using generalized linear models (GLMs) with a normal distribution for continuous outcomes), Poisson distribution for count outcomes (e.g. exacerbations and hospitalizations), and binomial distribution for binary outcomes. In all models, baseline levels of the outcome were entered as a predictor and follow-up levels as the dependent variable, with use of a robust standard error to account for clustering.

Outcomes were by group assignment (intention to treat) using generalized linear models adjusted for baseline levels of variable and clustering.

Value is for rate of exacerbations for participants in Health Coached arm minus rate for Usual Care, adjusted for baseline values and for clustering.

Major health coach activities:

Patient COPD education; Correct use of inhalers and nebulizers; Identifying red flags and when to seek medical care; Dyspnea management; Patient decision making and action plans around, exercise, smoking cessation, nutrition, exacerbations; Ensuring appropriate preventive services (pneumovax, flu); Depression screening; Reinforcing clinician education and use of treatment guidelines by primary care providers; Helping patient obtain prescriptions; Identifying gaps in care, areas where care not in line with care plan; Helping patients to make and keep appointments and obtain needed testing; Facilitating communication between patients, pulmonary specialists and primary care providers; Connecting with community resources; Helping to access to psychosocial services as needed; Conducting exercise capacity assessment and working with pulmonary specialist to provide recommended exercise program; Working with patient family members a

Usual care includes access to specialist consultation via referral by the primary care clinician, access to patient education classes, smoking cessation classes, psychosocial medicine and nutritional counseling.

Outcomes were compared by group assignment using generalized linear models (GLMs) with a normal distribution for continuous outcomes), Poisson distribution for count outcomes (e.g. exacerbations and hospitalizations), and binomial distribution for binary outcomes. In all models, baseline levels of the outcome were entered as a predictor and follow-up levels as the dependent variable, with use of a robust standard error to account for clustering.

Outcomes were by group assignment (intention to treat) using generalized linear models adjusted for baseline levels of variable and clustering.

Major health coach activities:

Patient COPD education; Correct use of inhalers and nebulizers; Identifying red flags and when to seek medical care; Dyspnea management; Patient decision making and action plans around, exercise, smoking cessation, nutrition, exacerbations; Ensuring appropriate preventive services (pneumovax, flu); Depression screening; Reinforcing clinician education and use of treatment guidelines by primary care providers; Helping patient obtain prescriptions; Identifying gaps in care, areas where care not in line with care plan; Helping patients to make and keep appointments and obtain needed testing; Facilitating communication between patients, pulmonary specialists and primary care providers; Connecting with community resources; Helping to access to psychosocial services as needed; Conducting exercise capacity assessment and working with pulmonary specialist to provide recommended exercise program; Working with patient family members a

Usual care includes access to specialist consultation via referral by the primary care clinician, access to patient education classes, smoking cessation classes, psychosocial medicine and nutritional counseling.

Outcomes were compared by group assignment using generalized linear models (GLMs) with a normal distribution for continuous outcomes), Poisson distribution for count outcomes (e.g. exacerbations and hospitalizations), and binomial distribution for binary outcomes. In all models, baseline levels of the outcome were entered as a predictor and follow-up levels as the dependent variable, with use of a robust standard error to account for clustering.

Outcomes were by group assignment (intention to treat) using generalized linear models adjusted for baseline levels of variable and clustering.

Value is for mean score of participants in Health Coached arm minus mean score in Usual Care, adjusted for baseline values and for clustering.

Major health coach activities:

Patient COPD education; Correct use of inhalers and nebulizers; Identifying red flags and when to seek medical care; Dyspnea management; Patient decision making and action plans around, exercise, smoking cessation, nutrition, exacerbations; Ensuring appropriate preventive services (pneumovax, flu); Depression screening; Reinforcing clinician education and use of treatment guidelines by primary care providers; Helping patient obtain prescriptions; Identifying gaps in care, areas where care not in line with care plan; Helping patients to make and keep appointments and obtain needed testing; Facilitating communication between patients, pulmonary specialists and primary care providers; Connecting with community resources; Helping to access to psychosocial services as needed; Conducting exercise capacity assessment and working with pulmonary specialist to provide recommended exercise program; Working with patient family members a

Usual care includes access to specialist consultation via referral by the primary care clinician, access to patient education classes, smoking cessation classes, psychosocial medicine and nutritional counseling.

Outcomes were compared by group assignment using generalized linear models (GLMs) with a normal distribution for continuous outcomes), Poisson distribution for count outcomes (e.g. exacerbations and hospitalizations), and binomial distribution for binary outcomes. In all models, baseline levels of the outcome were entered as a predictor and follow-up levels as the dependent variable, with use of a robust standard error to account for clustering.

Outcomes were by group assignment (intention to treat) using generalized linear models adjusted for baseline levels of variable and clustering.

Value is for mean score of participants in Health Coached arm minus mean score in Usual Care, adjusted for baseline values and for clustering.

Health coaching included the following activities Patient COPD education; Correct use of inhalers and nebulizers; Red flags and when to seek medic al care; Dyspnea management; Patient decision making and action plans around, exercise, smoking cessation, nutrition, exacerbations; Ensuring appropriate preventive services (pneumovax, flu); Depression screening; Reinforcing clinician education and use of treatment guidelines by primary care providers; Helping patient obtain prescriptions Identifying gaps in care, areas where care not in line with care plan;Helping patients to make and keep appointments and obtain needed testing; Facilitating communication between patients, pulmonary specialists and primary care providers; Connecting with community resources; Helping to access to psychosocial services ; Conducting exercise capacity assessment and working with pulmonary specialist to provide recommended exercise program ; Working with patient family members and caregivers

Usual care was chosen as the comparison group to provide maximum generalizability of the study, as usual care is the practical alternative for the target population. Usual care includes patient education classes, smoking cessation classes, psychosocial medicine and nutritional counseling.

Outcomes were compared by group assignment using generalized linear models (GLMs) with a normal distribution for continuous outcomes), Poisson distribution for count outcomes (e.g. exacerbations and hospitalizations), and binomial distribution for binary outcomes. In all models, baseline levels of the outcome were entered as a predictor and follow-up levels as the dependent variable, with use of a robust standard error to account for clustering.

Outcomes were by group assignment (intention to treat) using generalized linear models adjusted for baseline levels of variable and clustering.

Value is for mean score of participants in Health Coached arm minus mean score in Usual Care, adjusted for baseline values and for clustering.

Major health coach activities:

Patient COPD education; Correct use of inhalers and nebulizers; Identifying red flags and when to seek medical care; Dyspnea management; Patient decision making and action plans around, exercise, smoking cessation, nutrition, exacerbations; Ensuring appropriate preventive services (pneumovax, flu); Depression screening; Reinforcing clinician education and use of treatment guidelines by primary care providers; Helping patient obtain prescriptions; Identifying gaps in care, areas where care not in line with care plan; Helping patients to make and keep appointments and obtain needed testing; Facilitating communication between patients, pulmonary specialists and primary care providers; Connecting with community resources; Helping to access to psychosocial services as needed; Conducting exercise capacity assessment and working with pulmonary specialist to provide recommended exercise program; Working with patient family members a

Usual care includes access to specialist consultation via referral by the primary care clinician, access to patient education classes, smoking cessation classes, psychosocial medicine and nutritional counseling.

Outcomes were compared by group assignment using generalized linear models (GLMs) with a normal distribution for continuous outcomes), Poisson distribution for count outcomes (e.g. exacerbations and hospitalizations), and binomial distribution for binary outcomes. In all models, baseline levels of the outcome were entered as a predictor and follow-up levels as the dependent variable, with use of a robust standard error to account for clustering.

Outcomes were by group assignment (intention to treat) using generalized linear models adjusted for baseline levels of variable and clustering.

Value is for mean percent predicted of participants in Health Coached arm minus mean percent predicted in Usual Care, adjusted for baseline values and for clustering.

Major health coach activities:

Patient COPD education; Correct use of inhalers and nebulizers; Identifying red flags and when to seek medical care; Dyspnea management; Patient decision making and action plans around, exercise, smoking cessation, nutrition, exacerbations; Ensuring appropriate preventive services (pneumovax, flu); Depression screening; Reinforcing clinician education and use of treatment guidelines by primary care providers; Helping patient obtain prescriptions; Identifying gaps in care, areas where care not in line with care plan; Helping patients to make and keep appointments and obtain needed testing; Facilitating communication between patients, pulmonary specialists and primary care providers; Connecting with community resources; Helping to access to psychosocial services as needed; Conducting exercise capacity assessment and working with pulmonary specialist to provide recommended exercise program; Working with patient family members a

Usual care includes access to specialist consultation via referral by the primary care clinician, access to patient education classes, smoking cessation classes, psychosocial medicine and nutritional counseling.

Outcomes were compared by group assignment using generalized linear models (GLMs) with a normal distribution for continuous outcomes), Poisson distribution for count outcomes (e.g. exacerbations and hospitalizations), and binomial distribution for binary outcomes. In all models, baseline levels of the outcome were entered as a predictor and follow-up levels as the dependent variable, with use of a robust standard error to account for clustering.

Outcomes were by group assignment (intention to treat) using generalized linear models adjusted for baseline levels of variable and clustering.

Value is for proportion (%) of participants in Health Coached arm minus proportion in Usual Care, adjusted for baseline values and for clustering.

Major health coach activities:

Patient COPD education; Correct use of inhalers and nebulizers; Identifying red flags and when to seek medical care; Dyspnea management; Patient decision making and action plans around, exercise, smoking cessation, nutrition, exacerbations; Ensuring appropriate preventive services (pneumovax, flu); Depression screening; Reinforcing clinician education and use of treatment guidelines by primary care providers; Helping patient obtain prescriptions; Identifying gaps in care, areas where care not in line with care plan; Helping patients to make and keep appointments and obtain needed testing; Facilitating communication between patients, pulmonary specialists and primary care providers; Connecting with community resources; Helping to access to psychosocial services as needed; Conducting exercise capacity assessment and working with pulmonary specialist to provide recommended exercise program; Working with patient family members a

Usual care includes access to specialist consultation via referral by the primary care clinician, access to patient education classes, smoking cessation classes, psychosocial medicine and nutritional counseling.

Value is for mean number of days for participants in Health Coached arm minus mean score in Usual Care, adjusted for baseline values and for clustering.

Major health coach activities:

Patient COPD education; Correct use of inhalers and nebulizers; Identifying red flags and when to seek medical care; Dyspnea management; Patient decision making and action plans around, exercise, smoking cessation, nutrition, exacerbations; Ensuring appropriate preventive services (pneumovax, flu); Depression screening; Reinforcing clinician education and use of treatment guidelines by primary care providers; Helping patient obtain prescriptions; Identifying gaps in care, areas where care not in line with care plan; Helping patients to make and keep appointments and obtain needed testing; Facilitating communication between patients, pulmonary specialists and primary care providers; Connecting with community resources; Helping to access to psychosocial services as needed; Conducting exercise capacity assessment and working with pulmonary specialist to provide recommended exercise program; Working with patient family members a

Usual care includes access to specialist consultation via referral by the primary care clinician, access to patient education classes, smoking cessation classes, psychosocial medicine and nutritional counseling.

Outcomes were compared by group assignment using generalized linear models (GLMs) with a normal distribution for continuous outcomes), Poisson distribution for count outcomes (e.g. exacerbations and hospitalizations), and binomial distribution for binary outcomes. In all models, baseline levels of the outcome were entered as a predictor and follow-up levels as the dependent variable, with use of a robust standard error to account for clustering.

Outcomes were by group assignment (intention to treat) using generalized linear models adjusted for baseline levels of variable and clustering.

Value is for proportion (%) of participants in Health Coached arm minus proportion in Usual Care, adjusted for baseline values and for clustering.

Major health coach activities:

Patient COPD education; Correct use of inhalers and nebulizers; Identifying red flags and when to seek medical care; Dyspnea management; Patient decision making and action plans around, exercise, smoking cessation, nutrition, exacerbations; Ensuring appropriate preventive services (pneumovax, flu); Depression screening; Reinforcing clinician education and use of treatment guidelines by primary care providers; Helping patient obtain prescriptions; Identifying gaps in care, areas where care not in line with care plan; Helping patients to make and keep appointments and obtain needed testing; Facilitating communication between patients, pulmonary specialists and primary care providers; Connecting with community resources; Helping to access to psychosocial services as needed; Conducting exercise capacity assessment and working with pulmonary specialist to provide recommended exercise program; Working with patient family members a

Usual care includes access to specialist consultation via referral by the primary care clinician, access to patient education classes, smoking cessation classes, psychosocial medicine and nutritional counseling.

Outcomes were compared by group assignment using generalized linear models (GLMs) with a normal distribution for continuous outcomes), Poisson distribution for count outcomes (e.g. exacerbations and hospitalizations), and binomial distribution for binary outcomes. In all models, baseline levels of the outcome were entered as a predictor and follow-up levels as the dependent variable, with use of a robust standard error to account for clustering.

Value is for proportion of participants in Health Coached arm minus proportion in Usual Care, adjusted for baseline values and for clustering.

Major health coach activities:

Patient COPD education; Correct use of inhalers and nebulizers; Identifying red flags and when to seek medical care; Dyspnea management; Patient decision making and action plans around, exercise, smoking cessation, nutrition, exacerbations; Ensuring appropriate preventive services (pneumovax, flu); Depression screening; Reinforcing clinician education and use of treatment guidelines by primary care providers; Helping patient obtain prescriptions; Identifying gaps in care, areas where care not in line with care plan; Helping patients to make and keep appointments and obtain needed testing; Facilitating communication between patients, pulmonary specialists and primary care providers; Connecting with community resources; Helping to access to psychosocial services as needed; Conducting exercise capacity assessment and working with pulmonary specialist to provide recommended exercise program; Working with patient family members a

Usual care includes access to specialist consultation via referral by the primary care clinician, access to patient education classes, smoking cessation classes, psychosocial medicine and nutritional counseling.

Outcomes were compared by group assignment using generalized linear models (GLMs) with a normal distribution for continuous outcomes), Poisson distribution for count outcomes (e.g. exacerbations and hospitalizations), and binomial distribution for binary outcomes. In all models, baseline levels of the outcome were entered as a predictor and follow-up levels as the dependent variable, with use of a robust standard error to account for clustering.

Outcomes were by group assignment (intention to treat) using generalized linear models adjusted for baseline levels of variable and clustering.

Major health coach activities:

Patient COPD education; Correct use of inhalers and nebulizers; Identifying red flags and when to seek medical care; Dyspnea management; Patient decision making and action plans around, exercise, smoking cessation, nutrition, exacerbations; Ensuring appropriate preventive services (pneumovax, flu); Depression screening; Reinforcing clinician education and use of treatment guidelines by primary care providers; Helping patient obtain prescriptions; Identifying gaps in care, areas where care not in line with care plan; Helping patients to make and keep appointments and obtain needed testing; Facilitating communication between patients, pulmonary specialists and primary care providers; Connecting with community resources; Helping to access to psychosocial services as needed; Conducting exercise capacity assessment and working with pulmonary specialist to provide recommended exercise program; Working with patient family members a

Usual care includes access to specialist consultation via referral by the primary care clinician, access to patient education classes, smoking cessation classes, psychosocial medicine and nutritional counseling.

Outcomes were compared by group assignment using generalized linear models (GLMs) with a normal distribution for continuous outcomes), Poisson distribution for count outcomes (e.g. exacerbations and hospitalizations), and binomial distribution for binary outcomes. In all models, baseline levels of the outcome were entered as a predictor and follow-up levels as the dependent variable, with use of a robust standard error to account for clustering.

Outcomes were by group assignment (intention to treat) using generalized linear models adjusted for baseline levels of variable and clustering.

Value is for proportion (%) of participants in Health Coached arm minus proportion in Usual Care, adjusted for baseline values and for clustering.

Major health coach activities:

Patient COPD education; Correct use of inhalers and nebulizers; Identifying red flags and when to seek medical care; Dyspnea management; Patient decision making and action plans around, exercise, smoking cessation, nutrition, exacerbations; Ensuring appropriate preventive services (pneumovax, flu); Depression screening; Reinforcing clinician education and use of treatment guidelines by primary care providers; Helping patient obtain prescriptions; Identifying gaps in care, areas where care not in line with care plan; Helping patients to make and keep appointments and obtain needed testing; Facilitating communication between patients, pulmonary specialists and primary care providers; Connecting with community resources; Helping to access to psychosocial services as needed; Conducting exercise capacity assessment and working with pulmonary specialist to provide recommended exercise program; Working with patient family members a

Usual care includes access to specialist consultation via referral by the primary care clinician, access to patient education classes, smoking cessation classes, psychosocial medicine and nutritional counseling.

Outcomes were compared by group assignment using generalized linear models (GLMs) with a normal distribution for continuous outcomes), Poisson distribution for count outcomes (e.g. exacerbations and hospitalizations), and binomial distribution for binary outcomes. In all models, baseline levels of the outcome were entered as a predictor and follow-up levels as the dependent variable, with use of a robust standard error to account for clustering.

Outcomes were by group assignment (intention to treat) using generalized linear models adjusted for baseline levels of variable and clustering.

Value is for proportion (%) of participants in Health Coached arm minus proportion in Usual Care, adjusted for baseline values and for clustering.

Major health coach activities:

Patient COPD education; Correct use of inhalers and nebulizers; Identifying red flags and when to seek medical care; Dyspnea management; Patient decision making and action plans around, exercise, smoking cessation, nutrition, exacerbations; Ensuring appropriate preventive services (pneumovax, flu); Depression screening; Reinforcing clinician education and use of treatment guidelines by primary care providers; Helping patient obtain prescriptions; Identifying gaps in care, areas where care not in line with care plan; Helping patients to make and keep appointments and obtain needed testing; Facilitating communication between patients, pulmonary specialists and primary care providers; Connecting with community resources; Helping to access to psychosocial services as needed; Conducting exercise capacity assessment and working with pulmonary specialist to provide recommended exercise program; Working with patient family members a

Usual care includes access to specialist consultation via referral by the primary care clinician, access to patient education classes, smoking cessation classes, psychosocial medicine and nutritional counseling.

Outcomes were compared by group assignment using generalized linear models (GLMs) with a normal distribution for continuous outcomes), Poisson distribution for count outcomes (e.g. exacerbations and hospitalizations), and binomial distribution for binary outcomes. In all models, baseline levels of the outcome were entered as a predictor and follow-up levels as the dependent variable, with use of a robust standard error to account for clustering.

Outcomes were by group assignment (intention to treat) using generalized linear models adjusted for baseline levels of variable and clustering.

Value is for rate of outpatient visits for participants in Health Coached arm minus rate for Usual Care, adjusted for baseline values and for clustering.

Major health coach activities:

Patient COPD education; Correct use of inhalers and nebulizers; Identifying red flags and when to seek medical care; Dyspnea management; Patient decision making and action plans around, exercise, smoking cessation, nutrition, exacerbations; Ensuring appropriate preventive services (pneumovax, flu); Depression screening; Reinforcing clinician education and use of treatment guidelines by primary care providers; Helping patient obtain prescriptions; Identifying gaps in care, areas where care not in line with care plan; Helping patients to make and keep appointments and obtain needed testing; Facilitating communication between patients, pulmonary specialists and primary care providers; Connecting with community resources; Helping to access to psychosocial services as needed; Conducting exercise capacity assessment and working with pulmonary specialist to provide recommended exercise program; Working with patient family members a

Usual care includes access to specialist consultation via referral by the primary care clinician, access to patient education classes, smoking cessation classes, psychosocial medicine and nutritional counseling.

Outcomes were compared by group assignment using generalized linear models (GLMs) with a normal distribution for continuous outcomes), Poisson distribution for count outcomes (e.g. exacerbations and hospitalizations), and binomial distribution for binary outcomes. In all models, baseline levels of the outcome were entered as a predictor and follow-up levels as the dependent variable, with use of a robust standard error to account for clustering.

Value is for rate of COPD-related ED visits for participants in Health Coached arm minus rate for Usual Care, adjusted for baseline values and for clustering.

Major health coach activities:

Patient COPD education; Correct use of inhalers and nebulizers; Identifying red flags and when to seek medical care; Dyspnea management; Patient decision making and action plans around, exercise, smoking cessation, nutrition, exacerbations; Ensuring appropriate preventive services (pneumovax, flu); Depression screening; Reinforcing clinician education and use of treatment guidelines by primary care providers; Helping patient obtain prescriptions; Identifying gaps in care, areas where care not in line with care plan; Helping patients to make and keep appointments and obtain needed testing; Facilitating communication between patients, pulmonary specialists and primary care providers; Connecting with community resources; Helping to access to psychosocial services as needed; Conducting exercise capacity assessment and working with pulmonary specialist to provide recommended exercise program; Working with patient family members a

Usual care includes access to specialist consultation via referral by the primary care clinician, access to patient education classes, smoking cessation classes, psychosocial medicine and nutritional counseling.

Outcomes were compared by group assignment using generalized linear models (GLMs) with a normal distribution for continuous outcomes), Poisson distribution for count outcomes (e.g. exacerbations and hospitalizations), and binomial distribution for binary outcomes. In all models, baseline levels of the outcome were entered as a predictor and follow-up levels as the dependent variable, with use of a robust standard error to account for clustering.

Outcomes were by group assignment (intention to treat) using generalized linear models adjusted for baseline levels of variable and clustering.

Value is for rate of non-COPD-related ED visits for participants in Health Coached arm minus rate for Usual Care, adjusted for baseline values and for clustering.

Major health coach activities:

Patient COPD education; Correct use of inhalers and nebulizers; Identifying red flags and when to seek medical care; Dyspnea management; Patient decision making and action plans around, exercise, smoking cessation, nutrition, exacerbations; Ensuring appropriate preventive services (pneumovax, flu); Depression screening; Reinforcing clinician education and use of treatment guidelines by primary care providers; Helping patient obtain prescriptions; Identifying gaps in care, areas where care not in line with care plan; Helping patients to make and keep appointments and obtain needed testing; Facilitating communication between patients, pulmonary specialists and primary care providers; Connecting with community resources; Helping to access to psychosocial services as needed; Conducting exercise capacity assessment and working with pulmonary specialist to provide recommended exercise program; Working with patient family members a

Usual care includes access to specialist consultation via referral by the primary care clinician, access to patient education classes, smoking cessation classes, psychosocial medicine and nutritional counseling.

Outcomes were compared by group assignment using generalized linear models (GLMs) with a normal distribution for continuous outcomes), Poisson distribution for count outcomes (e.g. exacerbations and hospitalizations), and binomial distribution for binary outcomes. In all models, baseline levels of the outcome were entered as a predictor and follow-up levels as the dependent variable, with use of a robust standard error to account for clustering.

Outcomes were by group assignment (intention to treat) using generalized linear models adjusted for baseline levels of variable and clustering.

Value is for rate of COPD-related hospital visits for participants in Health Coached arm minus rate for Usual Care, adjusted for baseline values and for clustering.

Major health coach activities:

Patient COPD education; Correct use of inhalers and nebulizers; Identifying red flags and when to seek medical care; Dyspnea management; Patient decision making and action plans around, exercise, smoking cessation, nutrition, exacerbations; Ensuring appropriate preventive services (pneumovax, flu); Depression screening; Reinforcing clinician education and use of treatment guidelines by primary care providers; Helping patient obtain prescriptions; Identifying gaps in care, areas where care not in line with care plan; Helping patients to make and keep appointments and obtain needed testing; Facilitating communication between patients, pulmonary specialists and primary care providers; Connecting with community resources; Helping to access to psychosocial services as needed; Conducting exercise capacity assessment and working with pulmonary specialist to provide recommended exercise program; Working with patient family members a

Usual care includes access to specialist consultation via referral by the primary care clinician, access to patient education classes, smoking cessation classes, psychosocial medicine and nutritional counseling.

Outcomes were compared by group assignment using generalized linear models (GLMs) with a normal distribution for continuous outcomes), Poisson distribution for count outcomes (e.g. exacerbations and hospitalizations), and binomial distribution for binary outcomes. In all models, baseline levels of the outcome were entered as a predictor and follow-up levels as the dependent variable, with use of a robust standard error to account for clustering.

Outcomes were by group assignment (intention to treat) using generalized linear models adjusted for baseline levels of variable and clustering.

Value is for rate of non-COPD-related hospitalizations for participants in Health Coached arm minus rate for Usual Care, adjusted for baseline values and for clustering.

Major health coach activities:

Patient COPD education; Correct use of inhalers and nebulizers; Identifying red flags and when to seek medical care; Dyspnea management; Patient decision making and action plans around, exercise, smoking cessation, nutrition, exacerbations; Ensuring appropriate preventive services (pneumovax, flu); Depression screening; Reinforcing clinician education and use of treatment guidelines by primary care providers; Helping patient obtain prescriptions; Identifying gaps in care, areas where care not in line with care plan; Helping patients to make and keep appointments and obtain needed testing; Facilitating communication between patients, pulmonary specialists and primary care providers; Connecting with community resources; Helping to access to psychosocial services as needed; Conducting exercise capacity assessment and working with pulmonary specialist to provide recommended exercise program; Working with patient family members a

Usual care includes access to specialist consultation via referral by the primary care clinician, access to patient education classes, smoking cessation classes, psychosocial medicine and nutritional counseling.

Outcomes were compared by group assignment using generalized linear models (GLMs) with a normal distribution for continuous outcomes), Poisson distribution for count outcomes (e.g. exacerbations and hospitalizations), and binomial distribution for binary outcomes. In all models, baseline levels of the outcome were entered as a predictor and follow-up levels as the dependent variable, with use of a robust standard error to account for clustering.

Outcomes were by group assignment (intention to treat) using generalized linear models adjusted for baseline levels of variable and clustering.

Value is for proportion (%) of participants in Health Coached arm minus proportion in Usual Care, adjusted for baseline values and for clustering.

Major health coach activities:

Patient COPD education; Correct use of inhalers and nebulizers; Identifying red flags and when to seek medical care; Dyspnea management; Patient decision making and action plans around, exercise, smoking cessation, nutrition, exacerbations; Ensuring appropriate preventive services (pneumovax, flu); Depression screening; Reinforcing clinician education and use of treatment guidelines by primary care providers; Helping patient obtain prescriptions; Identifying gaps in care, areas where care not in line with care plan; Helping patients to make and keep appointments and obtain needed testing; Facilitating communication between patients, pulmonary specialists and primary care providers; Connecting with community resources; Helping to access to psychosocial services as needed; Conducting exercise capacity assessment and working with pulmonary specialist to provide recommended exercise program; Working with patient family members a

Usual care includes access to specialist consultation via referral by the primary care clinician, access to patient education classes, smoking cessation classes, psychosocial medicine and nutritional counseling.

Outcomes were compared by group assignment using generalized linear models (GLMs) with a normal distribution for continuous outcomes), Poisson distribution for count outcomes (e.g. exacerbations and hospitalizations), and binomial distribution for binary outcomes. In all models, baseline levels of the outcome were entered as a predictor and follow-up levels as the dependent variable, with use of a robust standard error to account for clustering.