Clinimetric Properties of Outcome Measures in Bronchiectasis

Abstract

Rationale: There is a lack of outcome measures with robust clinimetric properties in bronchiectasis. Objectives: To determine the clinimetric properties (reliability over 1 year during clinical stability and responsiveness over the course of antibiotics for pulmonary exacerbation) of objective and patient-reported outcome measures. Methods: This multicenter cohort study included adults with bronchiectasis from seven hospitals in the United Kingdom. Participants attended four visits, 4 months apart over 1 year while clinically stable and at the beginning and end of exacerbation and completed lung function (spirometry and multiple breath washout), provided a blood sample for C-reactive protein (CRP) measurement, and completed health-related quality of life (HRQoL) questionnaires (Quality of Life-Bronchiectasis, St. George's Respiratory Questionnaire, and EuroQoL 5-Dimensions 5-Levels). Results: Participants (n = 132) had a mean (standard deviation) age of 66 (11) years, and 64% were female. Lung function parameters (forced expiratory volume in one second [FEV₁], standard lung clearance index [LCI^2.5]) were reliable over time [coefficient of variation (CV): <10%]). Regarding responsiveness, FEV₁ demonstrated better properties than LCI^2.5; therefore, a clear justification for the use of LCI^2.5 in future trials is needed. CRP was less reliable (CV > 20%) over time than FEV₁ and LCI^2.5, and whereas CRP had a large mean change between the start and end of an exacerbation, this may have been driven by a small number of patients having a large change in CRP. Reliability of HRQoL questionnaires and questionnaire domains ranged from acceptable (CV: 20-30%) to good (CV: 10-20%), and HRQoL were responsive to treatment of exacerbations. Considering the specific questionnaire domain relevant to the intervention and its associated clinimetric properties is important. Additional statistics will support future power and/or sample size analysis. Conclusions: This information on the clinimetric properties of lung function parameters, CRP, and HRQoL parameters should be used to inform the choice of outcome measures used in future bronchiectasis trials.

Keywords: bronchiectasis; clinimetrics; outcome measures.

Figure 1.

Participant flow for the study. A total of 100 participants completed four stable visits, 17 participants completed three stable visits, 8 participants completed two stable visits, and 7 participants completed one stable visit only. Fifteen participants were excluded after consent as they were unable to perform an acceptable LCI, and one participant withdrew consent before the first visit. A total of 45 participants experienced a pulmonary exacerbation during the study period and completed visits at both the start and end of the exacerbation. Two participants only completed the start of the exacerbation visit, and 85 participants reported no exacerbations during the study period. CT = computed tomography; HRCT = high resolution computed tomography; IMP = investigational medicinal product; LCI = lung clearance index.

Figure 2.

Mean and standard error of the mean for lung function measures (A) forced expiratory volume in 1 second (FEV₁) percent predicted and (B) standard lung clearance index (LCI^2.5) (number of turnovers) collected across four consecutive visit time points from stable bronchiectasis participants (only participants with data from at least three visits were included). The error bars represent the standard error of the mean, derived from dividing the standard deviation by the square root of the sample size (FEV₁: n = 113; LCI^2.5: n = 95) in each measurement. Visits 1–4 = stable visit time points 3 months apart over 1 year.

Figure 3.

Mean and standard error of CRP serum concentrations (mg/L) collected across four consecutive visit time points from stable bronchiectasis participants (only participants with data from at least three visits were included). The error bars represent the standard error of the mean derived from dividing the standard deviation by the square root of the sample size (n = 96) in each measurement. Visits 1–4 = stable visit time points 3 months apart over 1 year. CRP = C-reactive protein.

Figure 4.

Mean and standard error of the mean scores for health-related quality of life (HRQoL) questionnaires were collected across four consecutive visit time points from stable bronchiectasis participants (only participants with data from at least three visits were included). The error bars represent the standard error of the mean, derived from dividing the standard deviation by the square root of the sample size (EQ-5D-5L, n = 113; QoL-B, n = 116; SGRQ, n = 117) in each HRQoL measurement. Visits 1–4 = stable visit time points 3 months apart over 1 year. EQ-5D-5L (VAS) = EuroQol 5-Dimensions 5-Levels (Visual Analogue Scale) – higher score equates to increased HRQoL; QoL-B–respiratory = Quality of Life-Bronchiectasis respiratory symptoms (higher score equates to increased HRQoL); SGRQ–Total = St. George’s Respiratory Questionnaire total score (higher score equates to decreased HRQoL).

Figure 5.

Mean and standard error of the mean for lung function measures (A) forced expiratory volume in 1 second (FEV₁) percent predicted and (B) standard lung clearance index (LCI^2.5) (number of turnovers) collected across two consecutive visit time points from bronchiectasis participants who experienced a pulmonary exacerbation (only participants with data from both pulmonary exacerbation visits were included). Change in objective measurements from start to end of exacerbation (mean [standard deviation]): FEV₁, 2.9 (3.8) %; LCI^2.5, −0.08 (0.6) turnovers. The error bars represent the standard error of the mean, derived from dividing the standard deviation by the square root of the sample size (FEV₁: n = 38; LCI^2.5: n = 26) in each measurement. PEx End = end of pulmonary exacerbation within 2 weeks of completing antibiotic therapy; PEx Start = start of pulmonary exacerbation within 24 hours of commencing antibiotic therapy.

Figure 6.

Mean and standard error of CRP serum concentrations (mg/L) collected across two consecutive visit time points from bronchiectasis participants who experienced a pulmonary exacerbation (only participants with data from both pulmonary exacerbation visits were included). The error bars represent the standard error of the mean derived from dividing the standard deviation by the square root of the sample size (n = 31) in each measurement. CRP = C-reactive protein; PEx End = end of pulmonary exacerbation within two weeks of completing antibiotic therapy; PEx Start = start of pulmonary exacerbation within 24 hours of commencing antibiotic therapy.

Figure 7.

Mean and standard error of the mean scores for health-related quality of life (HRQoL) questionnaires collected across two consecutive visit time points from bronchiectasis participants who experienced a pulmonary exacerbation (only participants with data from both pulmonary exacerbation visits were included). The error bars represent the standard error of the mean, derived from dividing the standard deviation by the square root of the sample size (EQ-5D-5L [EuroQol 5-Dimensions 5-Levels; higher score equates to increased HRQoL], n = 42; QoL-B [Quality of Life-Bronchiectasis; higher score equates to increased HRQoL], n = 44; and SGRQ [St. George’s Respiratory Questionnaire; higher score equates to decreased HRQoL], n = 44) in each HRQoL measurement. Change in HRQoL measurements from start to end of exacerbation (mean [standard deviation]): EQ-5D-5L (VAS), 13.522 (8.798); QoL-B respiratory domain, 16.506 (8.981); and SGRQ total, −3.156 (5.868). PEx End = end of pulmonary exacerbation within 2 weeks of completing antibiotic therapy; PEx Start = start of pulmonary exacerbation within 24 hours of commencing antibiotic therapy; VAS = Visual Analogue Scale.