Validation of a New Shortness of Breath With Daily Activities Questionnaire in Patients With Chronic Obstructive Pulmonary Disease

Primary outcomes

1. Internal Consistency (IC) of the Shortness of Breath With Daily Activities (SOBDA) Questionnaire in Participants With Chronic Obstructive Pulmonary Disease (COPD) Assessed as Cronbach's Alpha Value

   Time frame: Day 1 of the 2-week Run-in Period

   Cronbach's alpha (CA) is a measure of the IC of the 13-item SOBDA questionnaire (completed via electronic diary by a sample of participants). It is the ratio of the variance (var.) of the sum of the individual scores and the var. of the total score. The var. of the sum of a group of independent variables is the sum of their var.; thus, if the variables are positively correlated, the var. of the sum will be increased. If the items making up the score are identical and so perfectly correlated, CA=1. If the items are independent, CA=0. Higher scores indicate a more reliable (precise) instrument.

2. Test-retest Reliability (T-RR) of SOBDA Scores Measured as the Difference in the SOBDA Weekly Score Between Week 1 and Week 2 of the 2-week Run-in Period

   Time frame: Week 1 and Week 2 of the 2-week Run-in Period

   T-RR=stability during repeat measures over time in a stable population. SOBDA score was determined by the 13-item (it.) scoring algorithm, assigning a weekly mean score of 1-4 (higher scores=more severe breathlessness with daily activities) based on the mean of 7 days of data (or >=4 days). Daily total score is computed from the mean of the participant's (par.) scores on the 13 it. (>=7 it. must have non-missing responses). Only scores of stable par. (indicating no change [score=3] on the par.-completed Patient Global Assessment of Change [PGAC]; 1 [ much worse] to 5 [much better]) were used.

3. Convergent Validity for the SOBDA Questionnaire Measured as Correlations of the Baseline SOBDA Score With Participant-completed Modified Medical Research Council (mMRC) and Physician-completed mMRC Scores at Visit 2

   Time frame: Baseline (last week of the 2-week Run-in Period) and pre-treatment on Visit 2 (Day 1 of the 6-week Treatment Period)

   Convergent validity is defined as the ability of the SOBDA questionnaire to measure required information and was assessed by examining the relationship between the SOBDA score and the participant/physician-completed mMRC Dyspnea Scale assessments. The physician/participant rated the degree of the participant's dyspnea (trouble breathing) on the 5-point mMRC scale (0, none; 4, very severe). Spearman's rank correlation coefficient assesses if the relationship between two variables is monotone. A correlation of +1 or -1 will occur if one variable is a perfect monotone of the other.

4. Convergent Validity for the SOBDA Questionnaire Measured as the Correlation of the Baseline SOBDA Score With the Clinician Global Assessment of Dyspnea Severity (CGI-S) Score at Visit 2

   Time frame: Baseline (last week of the 2-week Run-in Period) and pre-treatment on Visit 2 (Day 1 of the 6-week Treatment Period)

   Convergent validity is defined as the ability of the SOBDA questionnaire to measure the required information and was assessed by examining the relationship between the SOBDA score with the CGI-S score. Spearman's rank correlation coefficient assesses if the relationship between two variables is monotone. A correlation of +1 or -1 will occur if one variable is a perfect monotone of the other. Clinicians were asked to assess the severity of the participant's dyspnea on the CGI-S scale. This was evaluated on a 1-4 Likert scale: 1 (mild) to 4 (very severe).

5. Convergent Validity (CV) for the SOBDA Questionnaire Measured as the Correlation of the Baseline SOBDA Score With the Chronic Respiratory Disease Questionnaire-Self-Administered Standardized (CRQ-SAS) Dyspnea Domain Score at Visit 2

   Time frame: Baseline (last week of the 2-week Run-in Period) and pre-treatment on Visit 2 (Day 1 of the 6-week Treatment Period)

   Convergent validity is defined as the ability of the SOBDA questionnaire to measure required information and was assessed by examining the relationship between the SOBDA score and the CRQ-SAS dyspnea domain score. Pearson's correlation coefficient is a measure of the linear dependence between 2 variables. A correlation of +1 or -1 will occur if the data from the 2 variables lie exactly on a line. The CRQ is a 20-item instrument measuring 4 domains (each measured on a scale of 1 [maximum impairment] to 7 [no impairment]) of functioning: mastery, fatigue, emotional function, and dyspnea.

6. Known Group Validity for the SOBDA Questionnaire Measured as the Comparison of the Baseline SOBDA Score in the Indicated Categories of the Physician-completed (PyC) mMRC Score at Visit 2

   Time frame: Baseline (last week of the 2-week Run-in Period) and pre-treatment on Visit 2 (Day 1 of the 6-week Treatment Period)

   SOBDA known group validity refers to the extent to which scores from the SOBDA questionnaire should differentiate participants with varying levels of dyspnea severity. It was assessed by comparing summary measures for the SOBDA score for each indicated level (0, 1, 2, 3, and 4) of the PyC mMRC. The physician rated the degree of the participant's dyspnea on the 5-point mMRC scale: 0 (none) to 4 (very severe). Known group validity was confirmed if the SOBDA score increased with increasing values of PyC mMRC, both indicating increased levels of breathlessness.

7. Known Group Validity for the SOBDA Questionnaire Measured as the Comparison of the Baseline SOBDA Score in the Indicated Categories of the Participant-completed (ParC) mMRC Score at Visit 2

   Time frame: Baseline (last week of the 2-week Run-in Period) and pre-treatment on Visit 2 (Day 1 of the 6-week Treatment Period)

   SOBDA known group validity refers to the extent to which scores from the SOBDA questionnaire should differentiate participants with varying levels of dyspnea severity. It was assessed by comparing summary measures for the SOBDA score for each indicated level (0, 1, 2, 3, and 4) of the ParC mMRC. The participant rated the degree of his/her dyspnea on the 5-point mMRC scale: 0 (none) to 4 (very severe). Known group validity was confirmed if the SOBDA score increased with increasing values of ParC mMRC, both indicating increased levels of breathlessness.

8. Known Group Validity (KGV) for the SOBDA Questionnaire Measured as the Comparison of the Baseline SOBDA Score in the Indicated Categories of CGI-S Scores at Visit 2

   Time frame: Baseline (last week of the 2-week Run-in Period) and pre-treatement on Visit 2 (Day 1 of the 6-week Treatment Period)

   SOBDA KGV refers to the extent to which scores from the SOBDA questionnaire should differentiate participants with varying levels of dyspnea severity. It was assessed by comparing summary measures for the SOBDA score for each indicated level (0, 1, 2, 3, and 4) of the CGI-S score. Clinicians were asked to assess the severity of the participant's dyspnea on the CGI-S scale. This was evaluated on a 1-4 Likert scale: 1 (mild) to 4 (very severe). KGV was confirmed if the SOBDA score increased with increasing values of CGI-S, both indicating increased levels of breathlessness.

9. Participants (Par.) Classified as Responders/Non-responders According to the Patient Global Assessment of Change (PGAC) Response at Days 8, 15, 22, 29, 36, and 43 and at Visit 3/Premature Discontinuation (PD) (the End of the 6-week Treatment Period or PD)

   Time frame: Days 8, 15, 22, 29, 36, and 43 and Visit 3/PD (end of 6-week Treatment Period or earlier up to Week 8)

   The PGAC is par. completed on a 1-5 scale: 1, much worse; 2, worse; 3, no change; 4, better; 5, much better. Responders were defined as par. with a rating of "better" or "much better" (score of 4 or 5) on the PGAC at the relevant week; non-responders were defined as par. with a response of "much worse," "worse," or "no change" on the PGAC. As pre-specified in the study protocol, results are presented independent of treatment allocation . The study objectives were to assess the measurement properties and validity of the SOBDA questionnaire independent of specific treatment effect.

10. Change From the Previous Week to the Current Week's SOBDA Score by Participant-completed PGAC Response at Days 8, 15, 22, 29, 36, and 43 and at Visit 3/PD (End of the 6-week Treatment Period or PD)

    Time frame: Baseline; Days 8, 15, 22, 29, 36, and 43 and Visit 3/PD (end of 6-week Treatment Period or earlier up to Week 8)

    Responsiveness reflects the ability of the SOBDA questionnaire to detect change under conditions of known change. Responders (Rs)=participants (par.) with a rating of "better"/"much better" (score of 4/5) on the PGAC (range; 1 [much worse] to 5 [much better]) at the relevant week; NRs=par. with a response of "much worse," "worse," or "no change" (score of 3). Mean difference between Rs and NRs in the change from the previous week to the current week's SOBDA score was calculated. For Visit 3/PD, the change from Baseline to the last treatment week's SOBDA score for Rs and NRs was calculated.

11. Number of Participants Classified as Responders and Non-responders by Clinician Global Impression of Change Question (CGI-C) Response at Visit 3/PD

    Time frame: Visit 3/PD (end of 6-week Treatment Period or earlier up to Week 8)

    Clinicians were asked to provide their clinical impression regarding change in the participant's shortness of breath by CGI-C. This was evaluated on a 1-5 Likert scale: 1 (much worse) to 5 (much better), with 3 being no change. A CGI-C responder was defined as a participant who had a response of "better" (4) or "much better" (5), and a non-responder was defined as a participant who had a response of "much worse" (1), "worse" (2), or "no change" (3).

12. Number of Participants Classified as Responders and Non-responders by CRQ-SAS Dyspnea Domain Response at Visit 3/PD

    Time frame: Visit 3/PD (end of 6-week Treatment Period or earlier up to Week 8)

    A CRQ-SAS dyspnea domain responder was defined as a participant who had a score increase of 0.5 units or more for the dyspnea domain of the CRQ-SAS between Visit 2 and Visit 3/Premature Discontinuation. A non-responder was defined as a participant who had a decrease in the score, or an increase of less than 0.5 units.

13. Number of Participants Classified as Responders and Non-responders by Physician-completed and Participant-completed mMRC Response at Visit 3/PD

    Time frame: Visit 3/PD (end of 6-week Treatment Period or earlier up to Week 8)

    A Physician-completed and Participant-completed mMRC responder was defined as a participant who had a score decrease of one unit or more between Visit 2 and Visit 3/Premature Discontinuation. A non-responder was defined as a participant who had the same score or an increase in score.

14. Change From Baseline to Last Treatment Week in the SOBDA Score by CGI-C Responses at Visit 3/PD

    Time frame: Baseline (2-week Run-in Period) and Week Prior to Visit 3/PD (end of 6-week Treatment Period or earlier up to Week 8)

    The responsiveness of the SOBDA questionnaire was assessed by comparing score changes between responders and non-responders. The CGI-C is clinician completed on a 1 to 5 scale: 1, much worse; 2, worse; 3, no change; 4, better; 5, much better. Changes in mean SOBDA scores during the last week of treatment in responders and non-responders using definitions based on the CGI-C conducted at Visit 3/Premature Discontinuation were assessed.

15. Change From Baseline to Last Treatment Week in the SOBDA Score by CRQ-SAS Dyspnea Domain (DD) Responses at Visit 3/PD

    Time frame: Baseline (2-week Run-in Period) and Week Prior to Visit 3/PD (end of 6-week Treatment Period or earlier up to Week 8)

    The responsiveness of the SOBDA questionnaire was assessed by comparing score changes of responders (Rs) versus non-responders (NRs). The CRQ-SAS DD includes 5 questions (q.) scored 1 (maximum impairment) to 7 (no impairment). Individual q. were equally weighted, and domain scores (DSs) (range=1-7) were calculated as the mean across the non-missing items within each domain (DSs were calculated although an individual item score was missing). Changes in mean SOBDA scores during the last treatment week in Rs and NRs using definitions based on the CRQ-SAS DD conducted at Visit 3/PD were assessed.

16. Change From Baseline to Last Treatment Week in the SOBDA Score by Physician-completed mMRC and Participant-completed mMRC Responses at Visit 3/PD

    Time frame: Baseline (2-week Run-in Period) and Week Prior to Visit 3/PD (end of 6-week Treatment Period or earlier up to Week 8)

    The responsiveness of the SOBDA questionnaire was assessed by comparing score changes between responders and non-responders. The mMRC ranges from 0 (no breathlessness except with strenous exercise) to 4 (too breathless to leave the house; breathless when dressing/undressing) and is completed by the clinician or the participant as indicated. Changes in mean SOBDA scores during the last week of treatment in responders and non-responders using definitions based on the Physician-completed (Ph-C) and Participant-completed (Pa-C) mMRC conducted at Visit 3/Premature Discontinuation were assessed.

17. SOBDA Threshold for Response Assessed as Mean Change From the Previous Week's SOBDA Score Based on a Participant-completed PGAC Score Rated of "Better"

    Time frame: Baseline (last week of the 2-week Run-in Period) and Weeks 1, 2, 3, 4, 5, and 6 (6-week Treatment Period)

    Changes from Baseline in the SOBDA score for responders (Rs) and non-responders (NRs) (using the PGAC assessment; 1 [much worse] to 5 [much better]), together with the cumulative proportions of Rs and NRs, was used to establish the threshold for defining SOBDA questionnaire Rs. The threshold of response is a score change in the SOBDA questionnaire that is demonstrated to have a perceivable benefit for the participant. The threshold of response was evaluated as the change from Baseline in the SOBDA score based on PGAC scores pre-specified as "better" or demonstrating meaningful improvement.

18. SOBDA Threshold for Response as Assessed by Mean Change From Baseline to the Last Treatment Week in the SOBDA Score Based on a CGI-C Response Rated as "Better"

    Time frame: Baseline and Week Prior to Visit 3/PD (end of 6-week Treatment Period or earlier up to Week 8)

    The threshold of response is a score change in the SOBDA questionnaire that is demonstrated to have a perceivable benefit for the participant. The threshold of response was evaluated as the change from Baseline in the SOBDA score based on CGI-C scores pre-specified as "better" or demonstrating meaningful improvement. The CGI-C is clinician completed on a 1 to 5 scale: 1, much worse, 2, worse; 3, no change; 4, better; 5, much better.

19. SOBDA Threshold for Response as Assessed by Mean Change From Baseline to the Last Treatment Week in the SOBDA Score Based on a CRQ-SAS Dyspnea Domain (DD) Response Rated as "Better"

    Time frame: Baseline and Week Prior to Visit 3/PD (end of 6-week Treatment Period or earlier up to Week 8)

    The threshold of response (TOR) is a score change in the SOBDA questionnaire that is demonstrated to have a perceivable benefit. The TOR was evaluated as the change from Baseline in the SOBDA score based on CRQ-SAS scores pre-specified as "better" or demonstrating meaningful improvement. The CRQ-SAS DD includes 5 questions (q.) scored 1 (maximum impairment) to 7 (no impairment). Individual q. were equally weighted, and domain scores (DSs) (range=1-7) were calculated as the mean across the non-missing items within each domain (DSs were calculated although an individual item score was missing).

20. SOBDA Threshold for Response Assessed as Mean Change From Baseline to Last Treatment Week in the SOBDA Score Based on Forced Expiratory Volume in One Second (FEV1) Change From Baseline of 50 Milliliters (mL) to <100 mL

    Time frame: Baseline and Week Prior to Visit 3/PD (end of 6-week Treatment Period or earlier up to Week 8)

    FEV1 response was rated as 1=No change or worse (i.e., change of <50 mL); 2=Better (i.e., change of 50 to <100 mL); 3=Much better (i.e., change of >=100 mL). The threshold of response is a score change in the SOBDA questionnaire that is demonstrated to have a perceivable benefit for the participant. The threshold of response was evaluated as the change from Baseline in the SOBDA score based on study assessment (FEV1) scores pre-specified as "better" or demonstrating meaningful improvement.