Repeatability of Multiple Breath Washout in Pediatric Primary Ciliary Dyskinesia

ABSTRACT

Background

Primary ciliary dyskinesia (PCD) is a motile ciliopathy characterized by abnormal mucociliary clearance and progressive lung disease. Spirometry is commonly used to monitor lung health and response to treatment, but it is known to be insensitive to early subclinical lung disease in PCD. Multiple breath washout is more sensitive than spirometry, but its repeatability in PCD has not been assessed.

Objectives

To evaluate the (i) same‐day and (ii) 28‐day repeatability of lung clearance index 2.5% (LCI) in PCD.

Methods

Participants > 6 years old with a confirmed PCD diagnosis were recruited from two Canadian PCD centers. Participants completed baseline lung function tests to measure their forced expiratory volume in 1‐second, z‐score (FEV1z), and LCI. Tests were repeated either on the same day or after 28 days. No clinical interventions were performed during the same‐day repeat testing. Outpatient therapies were unchanged during 28‐day repeat testing. Repeatability was assessed using intraclass correlation (ICC), and Bland−Altman plots (B&A).

Results

Twenty‐three participants were enrolled (same‐day: 16; 28‐day: 13). The same‐day and 28‐day repeat testing ICC for FEV1z were 0.9 and 0.92, and LCI were 0.95 and 0.71, respectively. Baseline testing showed that most participants had abnormal LCI (18 of 29 tests), even in those with FEV1z in the normal range. FEV1z and LCI exhibited a weak inverse correlation.

Conclusions

LCI is a repeatable and sensitive lung function measure in PCD patients, and may be a suitable outcome metric for clinical trials, particularly in patients with early subclinical lung disease.

Abbreviations

BCCH

BC Children's Hospital

B&A

Bland−Altman limit of agreement plots

CR

coefficient of reproducibility

CT

computed tomography

CV%

within‐subject coefficient of variation percentage

CXR

chest X‐ray

EM defect

ciliary ultrastructural transmission electron microscopy defect

FEV1

forced expiratory volume in 1‐second

FEV1z

forced expiratory volume in 1‐second, z‐score

ICC

intraclass correlation

IQR

interquartile range

LCI

lung clearance index 2.5%

LOA

limits of agreements (1.96 × standard deviation)

MBW

multiple breath washout

PCD

primary ciliary dyskinesia

r

Pearson's correlation coefficient

SD

standard deviation

SickKids

hospital for sick children

1. Introduction

Primary ciliary dyskinesia (PCD) is a genetic disease with dysfunctional motile cilia and abnormal mucociliary clearance. Patients with PCD will often have chronic, suppurative, sino‐oto‐pulmonary disease [1] and recurrent pulmonary exacerbations, leading to progressive lung damage, pulmonary function decline, and early‐onset bronchiectasis [2]. Therefore, routine health assessments and pulmonary function testing are critical in PCD management to monitor the progression of lung disease [3].

Spirometry is the most available pulmonary function test and is used to measure the forced expiratory volume in 1‐second (FEV1) [3]. Despite its widespread use, FEV1 is known to be insensitive to early lung disease, especially in the small airways [4]. Consequently, this has made FEV1 a poor metric in PCD for early disease monitoring and as an endpoint in clinical trials [4, 5, 6]. An alternative pulmonary function test is multiple breath washout (MBW), which measures the kinetics of a tracer gas (e.g., nitrogen) as the patient performs tidal breathing. The primary reported metric is the lung clearance index 2.5% (LCI), which is the cumulative amount of expired volume required to reduce the tracer gas concentration below a pre‐defined threshold level of 2.5% or 1/40th of the initial normalized concentration [7]. While LCI is a sensitive measure for detecting early manifestations of peripheral airway pathology [7], the literature on its clinimetric properties in PCD is limited.

Cross‐sectional MBW studies have found LCI to be more sensitive than FEV1, with PCD patients demonstrating abnormal LCI despite having a FEV1 in the normal range [5, 6, 8, 9]. However, there are conflicting reports about the relationship between LCI and FEV1, with some studies showing no correlation [5, 6, 8, 9] and others finding an inverse linear relationship [10, 11]. Longitudinal studies assessing how LCI changes over different time periods (1‐ or 5‐year) also have seemingly differing results, with one study finding an increase in LCI and the others reporting no change [11, 12, 13]. There are also a couple of studies that have reported on the responsiveness of LCI to clinical status change. One article found LCI to be responsive during a pulmonary exacerbation and another reporting no change after a single session of chest physiotherapy [13, 14]. Overall, the current MBW studies have discordant results but no studies have assessed the repeatability of MBW in PCD, which may explain some of the reported findings.

Repeatability is an important measurement property that is specific to a disease‐test combination and assesses how consistent the results are when the test is repeated under identical conditions. Moreover, establishing repeatability is necessary before assessing the responsiveness of a test, otherwise it is unknown whether differences in the results are reflective of clinical status change or test variation [15].

Therefore, the objectives of our study were to evaluate the (i) same‐day and (ii) 28‐day within‐subject repeatability of LCI in a PCD cohort. The secondary objectives were to assess the (i) same‐day and (ii) 28‐day within‐subject repeatability of FEV1z. We hypothesize that LCI and FEV1z are repeatable outcome metrics in PCD.

2. Methods

2.1. Study Design

We performed a multicenter prospective observational cohort study assessing the repeatability of FEV1z and LCI in PCD.

2.2. Study Procedure

Patients were recruited from the SickKids and BC Children's Hospital (BCCH) PCD clinics. Written informed consents and, when applicable, assents were obtained from the participants and parents. The inclusion criteria were (i) > 6 years old, (ii) ability to perform reproducible spirometry, and (iii) with a confirmed PCD diagnosis (i.e., biallelic pathogenic/likely pathogenic variants in one known PCD gene, or a classic ciliary ultrastructural transmission electron microscopy defect) [16, 17]. The exclusion criteria were (i) intercurrent pulmonary exacerbation (based on the treating PCD physician using published criteria [18]), (ii) any other cardiac or respiratory disease, leading to cardiopulmonary compromise, and (iii) medical instability that would preclude the ability to undergo the required investigations. Only participants from BCCH were offered the opportunity to partake in the same‐day and/or 28‐day repeat testing.

All participants underwent baseline lung function testing with spirometry and MBW. Lung function tests were repeated in the same order, either within the same‐day (after at least 2 h had elapsed) or after 28‐days (±7 days). For same‐day repeat testing, no pulmonary interventions were conducted in between tests. For the 28‐day repeat testing, participants continued their daily routines without any modification to their therapies.

Spirometry and MBW were performed by qualified research team members according to ATS/ERS standards [19, 20]. Spirometry results were based on the Global Lung Function Initiative reference equations [21, 22]. Nitrogen MBW was performed using the Exhalyzer D (Eco Medics AG, Dürnten, Switzerland) and Spiroware Version 3.3.2, and required a minimum of two acceptable trials. MBW quality control was completed [23].

2.3. Statistical Analysis

Descriptive statistics were calculated for the entire cohort including sex, age (at time of diagnosis and at time of testing), ciliary defect, genetics, bronchiectasis (based on the most recent clinical chest X‐rays [CXR] or computed tomography [CT] scans of the chest), and situs inversus, and reported as number (n) and proportion (%) or median and interquartile range (IQR), as appropriate.

Repeatability was assessed using the intraclass correlation coefficient (ICC) and Bland−Altman limit of agreement plots (B&A) [24, 25]. ICC values of < 0.5, 0.5–0.75, 0.75–0.9, and > 0.90 indicated poor, moderate, good, and excellent repeatability, respectively [24]. The B&A were plotted with the limits of agreement (LOA) defined as the mean difference ±1.96 times the standard deviation (mean ± 1.96 × SD) [25]. The B&A were assessed for any biases or trends across the range of test results. Differences in the repeat test results were tested for normality using the Shapiro−Wilks test.

Repeat test results were compared using the Wilcoxon signed‐rank test to determine if they were statistically different. Correlations between repeat test results were visualized using linear regressions. The coefficient of reproducibility (CR) and within‐subject coefficient of variation percentage (CV%) for FEV1z and LCI were calculated (Supporting Information S1: Table 1). A CV% of 10% or less, more than 10% and up to 20%, more than 20% and up to 30%, and greater than 30%, were considered excellent, good, acceptable, and poor, respectively.

The calculated sample size was 15 participants to achieve an ICC of at least 0.9, with 95% confidence and a confidence interval (CI) width of 0.2 [26].

A p < 0.05 was considered statistically significant. Analyses were completed using RStudio [27] and GraphPad [28].

2.4. Research Ethics and Clinical Trials

This study was part of a multi‐center clinical study that was approved by the research ethics boards at the Hospital for Sick Children (SickKids; REB# 1000068639) and BC Children's Hospital (BCCH; REB# H21‐00142 and REB# H22‐03529). The study was registered with www.clinicaltrials.gov (NCT# 04858191).

3. Results

3.1. Demographics

A total of 23 participants (10 from SickKids and 13 from BCCH) were recruited for the study. The whole cohort had 12 (52.2%) male participants, and the median age was 13 years (IQR: 10.5 to 14). The participants from SickKids were on average younger (median age: 10.5, IQR: 8.5 to 13.75) compared to BCCH (median age: 13 years, IQR: 12 to 16). There were 17 (73.9%) participants with documented bronchiectasis (based on the most recent available chest CT or CXR) and 11 (47.8%) with situs inversus totalis or situs ambiguous. The ultrastructural ciliary transmission electron microscopy defects (EM defect), genotype, and lung function test results for the entire cohort are summarized in Table 1.

Table 1.

Participant baseline demographics.

Parameter	Total (n = 23)	Hospital for sick children (n = 10)	BC children's (n = 13)
Male (n, %)	12 (52.2%)	5 (50%)	7 (53.8%)
Age (years; median [IQR])
During study	13 [10.5, 14]	10.5 [8.5, 13.8]	13 [12, 16]
At diagnosis	6 [2, 9.5]	4.3 [0.8, 6.8]	7 [4, 11]
Bronchiectasis (n, %)a	17 (73.9%)	7 (70%)	10 (76.9%)
Situs inversus (n, %)	11 (47.8%)	4 (40%)	7 (53.8%)b
EM ciliary defect (n, %)
ODA	8 (34.9%)	3 (30%)	5 (38.5%)
ODA/IDA	4 (17.4%)	2 (20%)	2 (15.4%)
IDA/MTD	1 (4.3%)	1 (10%)	0 (0%)
Normal	3 (13%)	2 (20%)	1 (7.7%)
Non‐diagnostic	7 (30.4%)	2 (20%)	5 (38.5%)
Genotype (n, %)
CCDC103	1 (4.3%)	1 (10%)	0 (0%)
CCDC40	1 (4.3%)	1 (10%)	0 (0%)
DNAAF5	1 (4.3%)	0 (0%)	1 (7.7%)
DNAH11	2 (8.7%)	1 (10%)	1 (7.7%)
DNAH5	5 (21.8%)	1 (10%)	4 (30.8%)
DNAI1	3 (13%)	1 (10%)	2 (15.4%)
DNAI2	1 (4.3%)	0 (0%)	1 (7.7%)
Hydin	1 (4.3%)	1 (10%)	0 (0%)
LRRC6	2 (8.7%)	0 (0%)	2 (15.4%)
RPGR	2 (8.7%)	0 (0%)	2 (15.4%)
RSPH9	1 (4.3%)	1 (10%)	0 (0%)
ZMYD19	1 (4.3%)	1 (10%)	0 (0%)
Unknown	2 (8.7%)	2 (20%)	0 (0%)
Same‐day repeat testing
Participants (n, %)	16 (100%)	10 (62.5%)	6 (37.5%)
Test 1 (baseline)
FEV1z	−0.64 [−1.38, −0.05]	−0.21 [−0.94, 0.16]	−1.16 [−1.79, −0.92]
LCI	7.2 [6.36, 7.94]	6.5 [6.3, 7.81]	7.21 [6.62, 7.86]
Test 2 (same‐day repeat)
FEV1z	−0.5 [−1.27, −0.26]	−0.47 [−1.05, −0.26]	−1.12 [−1.75, −0.34]
LCI	7.35 [6.4, 9.03]	6.86 [6.18, 8.66]	7.56 [7.4, 8.96]
28‐day repeat testing
Participants (n, %)	13 (100%)	0 (0%)	13 (100%)
Test 1 (baseline)
FEV1z (n = 13, 100%)	−0.75 [−1.24, 0.11]	N/A	−0.75 [−1.24, 0.11]
LCI (n = 13, 100%)	8.04 [7.33, 8.81]	N/A	8.04 [7.33, 8.81]
Test 2 (28‐day repeat)
FEV1z (n = 10, 77%)	−1.47 [−1.66, −0.73]	N/A	−1.47 [−1.66, −0.73]
LCI (n = 10, 77%)	7.84 [7.12, 8.72]	N/A	7.84 [7.12, 8.72]

Abbreviations: CC = central complex, FEV1z = forced expiratory volume in 1‐second z‐score, IDA = inner dynein arm, IQR = interquartile range, LCI = lung clearance index 2.5, N/A = not applicable, ODA = outer dynein arm, VDP = ventilatory defect percentage.

^aDiagnosis of bronchiectasis was based on the most recent chest CT/CXR available at the time of study, with some studies being > 5 years old.

^b 1 participant had situs ambiguous (dextrocardia).

3.2. Same‐Day Repeat Testing

Sixteen participants (10 from SickKids and 6 from BCCH) completed same‐day repeat testing. Two participants did not satisfactorily complete repeat spirometry. Two participants were unable to complete one of the MBW tests to acceptable standards (Supporting Information S1: Table 2). The median baseline lung function results were FEV1z of −0.64 (IQR: −1.38 to −0.05) and LCI of 7.2 (IQR: 6.36 to 7.94), and better at SickKids (FEV1z: −0.21, LCI: 6.5) than BCCH (FEV1z: −1.16, LCI: 7.21), as shown in Table 1.

The repeat test results for FEV1z and LCI are plotted in Figure 1 and listed in Table 2. No statistical difference was found between the repeat test groups for both FEV1z and LCI, as indicated by the Wilcoxon signed‐rank test (p = 0.36 and 0.32, respectively). The Pearson's correlation coefficients were excellent for FEV1z and LCI (r = 0.91 and 0.95, respectively; Figure 1C,D and Table 2). The ICC for FEV1z and LCI were excellent (ICC = 0.9 and 0.95, respectively; Table 2). The B&A for FEV1z (Figure 1A) demonstrated agreement, with a small positive bias (0.15), and no notable trend. The B&A for LCI (Figure 1B) also demonstrated agreement, with a small negative bias (−0.19), and no notable trend. B&A using the relative difference in LCI had similar results (Supporting Information S1: Figure 1A). The differences in the same‐day test results of FEV1z and LCI were also found to be normal (Shapiro−Wilk test p= 0.24 and 0.99, respectively). For FEV1z, the CR was 0.36 and the CV% was 54.8%. For LCI, the CR was 0.38 and the CV% was 3.1%.

Figure 1.

Same‐day testing results visualized using Bland and Altman plots (A: FEV1z and B: LCI) and linear regressions (C: FEV1z and D: LCI). Participants from SickKids and BCCH are represented by filled circles and open squares, respectively. In (A) and (B), the mean difference and limits of agreement (mean ± 1.96 x SD) are shown by the solid and dashed lines, respectively. In (C) and (D), the linear regression line is represented by the dashed line and the 95% confidence interval by the shaded area. BCCH = BC Children's Hospital, FEV1z = forced expiratory volume in 1‐second z‐score, LCI = lung clearance index 2.5%, r = Pearson's correlation coefficient, SD = standard deviation, SickKids = hospital for sick children.

Table 2.

Lung function within‐subject repeatability test results.

Site	n	FEV1z			LCI
		ICC	CR	CV%	ICC	CR	CV%
Same‐day repeat testing
SickKids	10	0.93	0.36	54.8	0.95	0.38	3.1
BCCH	6	0.81			0.96
Both	16	0.90			0.95
28‐day repeat testing
BCCH	10	0.92	0.35	14.11	0.71	1.64	4.52

Note: One participant was unable to complete the MBW. Two participants were unable to complete spirometry.

Abbreviations: BCCH = BC Children's Hospital, CR = coefficient of reproducibility, CV% = within‐subject coefficient of variation percentage, FEV1z = forced expiratory volume in 1‐second z‐score, ICC = intraclass correlation, LCI = lung clearance index 2.5%, SickKids = hospital for sick children.

During baseline testing (Test 1), three participants had abnormal FEV1z (< 1.65) [22] and eight participants had abnormal LCI (> 7.1) [29], as shown in (Supporting Information S1: Table 2). After repeated lung function testing (Test 2), three participants had abnormal FEV1z with two participants being the same, and the LCI was persistently abnormal for the same eight participants. Comparison of spirometry and MBW results, for the baseline and repeat testing are shown in Supporting Information S1: Figure 2A,B. Baseline testing showed a weak inverse correlation that was statistically significant (n = 15, r = 0.56, p = 0.03), but the same‐day repeat testing demonstrated a weak inverse but statistically insignificant correlation (n = 14, r = 0.53, p = 0.051).

3.3. 28‐Day Repeat Testing

Thirteen participants were recruited for the 28‐day repeat testing with 11 participants completing both study visits and two participants only completing the baseline visit. Two participants were unable to complete spirometry at their baseline visit, therefore their clinical spirometries (completed within 2 weeks before the study visit) were used. Three participants were unable to complete the repeat spirometry, and one participant was unable to complete the repeat MBW to acceptable standards (Supporting Information S1: Table 3).

The median time interval between study visits was 33 days (range: 29 to 39). The median lung function for the baseline visit were a FEV1z of −0.75 (IQR: −1.24 to 0.11) and LCI of 8.04 (IQR: 7.33 to 8.81). The 28‐day repeat visit results were a FEV1z of −1.47 (IQR: −1.66 to −0.73) and LCI of 7.84 (IQR: 7.12 to 8.72).

The 28‐day repeat test results for FEV1z and LCI are plotted in Figure 2 and listed in Table 2. No statistical difference was found between the repeat test groups for both FEV1z and LCI, as indicated by the Wilcoxon signed‐rank test (p = 0.28 and 0.82, respectively). The Pearson's correlation coefficient for FEV1z and LCI were r = 0.94 and 0.72, respectively (Figure 2C,D and Table 2). The ICC for FEV1z was 0.92, and LCI was 0.71. The B&A for FEV1z (Figure 2A) demonstrated agreement, with a small positive bias (0.1), no notable trend, and with only one data point outside the LOA (−0.8, 1). Similarly, the B&A for LCI (Figure 2B) demonstrated agreement, with a positive bias (0.26), no notable trend, and with only one data point outside the LOA (−1.94, 2.45). B&A using the relative difference in LCI had similar results (Supporting Information S1: Figure 1B). The differences in the 28‐day test results of FEV1z and LCI did not satisfy normality (Shapiro−Wilk test p‐value of 0.01 and 0.02, respectively). For FEV1z, the CR was 0.35 and the CV% was 14.11%. For LCI, the CR was 1.64, and the CV% was 4.52%.

Figure 2.

28‐day testing results visualized using Bland and Altman plots (A: FEV1z and B: LCI) and linear regressions (C: FEV1z and D: LCI). Participants from BCCH are represented by open squares. In (A) and (B), the mean difference and limits of agreement (mean ± 1.96 x SD) are shown by the solid and dashed lines, respectively. In (C) and (D), the linear regression line is represented by the dashed line and the 95% confidence interval by the shaded area. BCCH = BC Children's Hospital, FEV1z = forced expiratory volume in 1‐second z‐score, LCI = lung clearance index 2.5%, r = Pearson's correlation coefficient.

During baseline testing, two participants had abnormal FEV1z (less than the lower limit of normal) [22] and 10 participants had abnormal LCI (> 7.1) [29] (Supporting Information S1: Table 3). During the 28‐day repeat testing, three participants had abnormal FEV1z, including the two identified during the baseline visit. The LCI was persistently abnormal for 7 participants, and one participant had improvement in LCI. Two other participants did not complete the repeat MBW. Comparison of spirometry and MBW results, for the baseline and 28‐day repeat testing are shown in Supporting Information S1: Figure 2C,D. Baseline testing showed a statistically significant and moderate negative correlation (n = 13; r = 0.62, p = 0.03) but the 28‐day repeat testing demonstrated a weakly positive and statistically insignificant correlation (n = 10; r = 0.47, p = 0.43).

4. Discussion

This study is the first to evaluate and establish the repeatability of FEV1z and LCI in pediatric patients with PCD. FEV1z exhibited excellent repeatability for both same‐day and 28‐day repeat testing, with ICC of 0.9 and 0.92, respectively (Table 2). Similarly, LCI demonstrated excellent same‐day repeatability (ICC: 0.95) and good 28‐day repeatability (ICC: 0.71; Table 2). We hypothesize that the lower ICC for 28‐day repeatability in LCI reflects the increased sensitivity of LCI as a lung function measure and its ability to detect unmeasured patient factors, such as subclinical pulmonary exacerbations. Notably, an outlier was identified during the 28‐day repeat testing of LCI and represented a participant who experienced a pulmonary exacerbation shortly after their first study visit. Excluding this outlier markedly improved the 28‐day ICC from 0.71 to 0.91.

In addition to the objectives, there were other notable findings. First, LCI was found to be abnormal in most participants, even in those with normal range FEV1z. This suggests that LCI is a more sensitive lung function marker at detecting lung disease, compared to FEV1z. This is consistent with other studies [5, 6, 13, 30]. Second, the correlation between FEV1z and LCI (Supporting Information S1: Figure 2) demonstrated a weak inverse correlation (similar to previous studies [10, 11, 13]) that was not statistically significant for all tests. We hypothesize that this could be, in part, due to our small study cohort and the inability of spirometry and MBW to differentiate regional lung function abnormalities and peripheral airway obstructions, like mucus plugging [31]. Ideally, the incorporation of structural and/or functional imaging would have been helpful to understand the impact of regional lung abnormalities on the lung function results. Third, the CR (based on the same‐day and 28‐day repeat testing) for FEV1z was calculated to be between 0.35 and 0.36, and for LCI, 0.38 and 1.64. These are some of the first estimates of the minimal detectable difference (MDD) for FEV1z and LCI in PCD patients [32]. However, it is important to remember that our CR was calculated using a distribution‐based approach and in a relatively small PCD cohort with generally mild lung disease. Interestingly, our findings fall within the 95% CI reported in a PCD study that assessed the change in FEV1z and LCI during a pulmonary exacerbation [13]. In that study, a change in FEV1z of −0.85 (95% CI: −1.29 to −0.41) and LCI of 0.89 (95% CI: 0.11 to 1.67) were considered statistically significant. Further studies assessing the responsiveness of FEV1z and LCI to clinical status change are still needed. Lastly, our study cohort had similar demographics and baseline FEV1z as the iPCD cohort [33] (n = 207, age range: 10−13, FEV1z = −1.00 (95% CI: −1.21 to −0.79), suggesting that our cohort may be representative of the larger pediatric PCD population.

There were study limitations. First, the overall sample size was relatively small, despite recruitment efforts at two Canadian PCD centers. Additionally, the 28‐day repeatability study was conducted at only one center. This may introduce sampling bias, as participants who completed both the same‐day and the 28‐day repeatability studies may represent a subset of patients with milder disease and less ventilation inhomogeneity, making them more willing to participate in multiple follow‐up MBW studies. However, we did not observe any statistical difference in baseline lung function between any of the participant groups, as shown in Supporting Information S1: Table 4. Second, the 28‐day repeat testing study arm was underpowered. For an ICC of 0.9, the calculated sample size should be at least 15 participants, but of the 13 participants recruited, only 10 participants completed the 28‐day repeat testing to acceptable levels. Interestingly, but not unexpected, the ICC improved from 0.72 to 0.95 when the participant with the pulmonary exacerbation was removed and this change underscores the sensitivity of the ICC to clinical variability, as it represents the proportion of variances attributable to the participants relative to the summed variances from the participants, test and other confounders (such as the change in clinical status from a pulmonary exacerbation). Third, most study participants had generally mild disease, based on their FEV1z and LCI, which may limit the generalizability of the study findings to patients with more severe disease. Future studies with larger sample sizes and participants with ranging disease severities are needed to better account for the variability across clinical states. Fourth, the completion of chest physiotherapy before the study visits was not standardized. It is possible that participants who completed airway clearance therapies before the study visit may have decreased mucus plugging and have more repeatable results. However, the contribution may be minimal as a recent study found that a single airway clearance therapy session did not significantly impact FEV1 or LCI in PCD patients [14]. Lastly, the protocoled order of lung function testing was spirometry followed by MBW, however this was not strictly enforced due to the timing of equipment and technical expertise. It is possible that the forced expiratory maneuvers during spirometry could have shifted mucus in the lungs, changing the location of mucus plugs and the LCI on MBW.

Overall, our study establishes the within‐subject repeatability of MBW over same‐day and 28‐day intervals, highlighting its reliability as a lung function measure in PCD. Importantly, the findings during same‐day repeat testing were most indicative of the repeatability of MBW, as they were impacted less by patient variability. In contrast, the 28‐day repeat testing results were influenced more by patient‐specific factors due to the longer interval between testing sessions, but nonetheless provided valuable insight into the sensitivity and responsiveness of MBW to clinical status changes. Our findings also confirmed that LCI is more sensitive than FEV1z at detecting early lung disease, further supporting the integration of MBW into clinical trials, longitudinal surveillance studies, and routine medical practice for PCD management. Future studies are needed to evaluate the responsiveness of MBW to clinical changes in PCD and to better understand the mechanisms driving the differences between FEV1z and LCI. Incorporating structural or functional chest imaging into these studies could shed light on how regional structural lung abnormalities and peripheral airway obstructions impact lung function outcomes.

Author Contributions

Wallace B. Wee: conceptualization, investigation, funding acquisition, writing – original draft, methodology, validation, visualization, writing – review and editing, software, formal analysis, data curation. Layan M. Bashi: writing – original draft, methodology, validation, visualization, writing – review and editing, formal analysis, data curation. Renee Jensen: methodology, data curation. Jonathan H. Rayment: methodology, data curation, writing – review and editing. Teresa To: writing – review and editing, conceptualization, methodology. Felix Ratjen: writing – review and editing, methodology, resources. Giles Santyr: conceptualization, writing – review and editing, methodology. Sharon D. Dell: supervision, resources, writing – review and editing, conceptualization.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

PCD MBW Supplement v2025‐03‐26.

Data Availability Statement

The data that supports the findings of this study are available in the supporting material of this article.