A Systematic Review of the Use of Physiological Tests Assessing the Acute Response to Treatment During Exacerbations of COPD (with a Focus on Small Airway Function)

Abstract

Exacerbations are prevalent in Chronic Obstructive Pulmonary Disease (COPD) patients and associated with poor clinical outcomes. Currently, there is a lack of sensitive and specific tools that can objectively identify exacerbations and assess their progress or treatment response. FEV₁ is often reported as a study outcome, but it has significant limitations. Studies have suggested that small airways measures might provide physiological biomarkers during exacerbations. Therefore, this study was done to assess which physiological tests of small airways function have been used in the acute setting during exacerbations of COPD and the evidence to support their use. An electronic databases search was conducted in April 2019. A standard systematic review methodology was used. Eligible studies were those of ≥10 participants that compared at least one small airway test with FEV₁ to assess response to treatment with baseline and a follow-up measurement ≤2 months after. Analyses were narrative. Of 1436 screened studies, seven studies were eligible. There was heterogeneity in which tests of small airways were used and three different small airways measures were reported. Studies were small (including 20 to 87 subjects). Six articles reported improvements in small airway measurements during the recovery from exacerbation which correlated with FEV₁. Included studies varied in their timing and duration of the assessment. There is some evidence to support the use of small airway tests in acute exacerbations of COPD. However, studies have been small with different tests being utilized. Further studies to determine the usefulness of each test may be of interest.

Keywords:

• COPD
• exacerbation
• monitoring
• small airways’ tests
• small airway dysfunction
• Lung function

Introduction

Chronic obstructive pulmonary disease (COPD) is a chronic inflammatory lung disease that is progressive [1] and characterized by persistent respiratory symptoms and progressive airflow limitation [2]. The combination of parenchymal destruction and small airways dysfunction (with small airways defined as those with an internal diameter of <2mm) lead to airflow limitation [2]. There has been increasing interest in the involvement of small airways in the pathophysiology of COPD since Hogg et al. described small airway resistance in this condition [3]. Pathological studies have shown that smokers with established airflow limitation had a generalized reduction in the diameter of the small airways [4] and, pathophysiologically, the small airways appear to be the major contributor to airflow limitation in COPD [5,6]. Furthermore, studies have shown that small airway disease may precede the onset of emphysema and airflow obstruction [6–8].

Many patients with COPD undergo exacerbations, which are associated with substantial mortality and morbidity [9]. Global Initiative for Chronic Obstructive Lung Disease (GOLD) [2] defines an exacerbation as “an acute event characterized by a worsening of the patient’s respiratory symptoms that is beyond normal day-to-day variations and leads to a change in medication”. Exacerbations are associated with a reduction in quality of life and lung function, and substantial healthcare cost and utilization [10]. Recently, the United Kingdom national COPD audit data identified difficulties in diagnosing exacerbations and poor care during exacerbations as significant, unfulfilled health needs [11]. Methods of diagnosing COPD exacerbations and assessing the response to treatment currently rely on self-reported symptoms in clinical practice and spirometry in research studies. Commonly, the forced expiratory volume in the first second (FEV₁) has been the primary physiological outcome measure for clinical trials during exacerbation. However, this forced respiratory maneuver is variable (even during stable COPD) [12] and insensitive to change over time [13], making it a poor biomarker for exacerbations. For example, bronchodilators have been shown to decrease hyperinflation, lessen the work of breathing and improve symptoms in the absence of a significant spirometric response [14].

Tests of small airway function might provide more sensitive and specific measures during exacerbations but there are a number of physiological measurements that have been proposed as measures of small airways function and a number of different devices available for each measurement [15–17]. The reported measurements include (but are not limited to) Mid-Maximal Expiratory Flow (MMEF which is also referred to as the forced expiratory flow between 25% and 75% of the forced vital capacity (FVC) (FEF_25-75%)), nitrogen washout tests, Forced oscillation technique (FOT) and airway resistance (Raw) obtained by body plethysmography.

While there remain some debate about the specificity of certain tests to evaluate small airways function [18], studies suggest that tests of small airways may be better able to identify physiological responses to treatment compared with conventional spirometric measures [19]. However, it is unclear whether such tests would be clinically valuable during an exacerbation of COPD as there are no previous comprehensive reviews of the evidence in this area. The aim of this review was to summarize the findings of studies comparing a measure of small airways function to FEV₁ during exacerbations of COPD to inform whether these tests could be incorporated as a primary outcome for further studies within this population. FEV₁ was chosen pragmatically as a comparator as it remains the most commonly reported physiological test in exacerbations of COPD, despite the potential limitations of this measure, as discussed above.

Methods and design

Protocol and registration

The systematic review protocol was prepared following Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines [20] and registered on PROSPERO (registration number: CRD42019131939).

Eligibility criteria

Study design

Randomized controlled trials (RCT), non-randomized interventional studies, observational studies, case series and uncontrolled studies of ≥10 that compared at least one small airway test with FEV₁ at both exacerbation onset (this was clinician-defined exacerbation) and a follow-up measurement up to and including two months after exacerbation onset.

Type of participants

Adult patients older than 18 years with a clinical diagnosis of COPD during exacerbation with no limitation on either COPD or exacerbation severity were included. Studies of COPD patients not experiencing an exacerbation (therefore, stable) were excluded. Studies of patients with a primary diagnosis of other lung diseases, including asthma, were also excluded.

Intervention and comparator

Studies were included if they contained at least one of the following commonly used tests of small airways in comparison to FEV₁: FOT, MMEF, Raw by body plethysmography and nitrogen washout tests.

Objective

The objective of this study was to assess treatment response using small airway tests in comparison to FEV₁ during exacerbation of COPD.

Outcome

Tests of small airways which included but (were not limited to) FOT, MMEF, Raw by body plethysmography, and nitrogen washout tests.

Search strategy

A comprehensive search was conducted using MEDLINE (via Ovid), EMBASE (via Ovid), CINAHL, and the Cochrane Central Register of Controlled Trials (CENTRAL). Different search strategies were used for each database (shown in the online supplement). ClinicalTrials.gov was used to search for completed and ongoing clinical trials. Where possible, articles not in English were translated. References lists of peer-reviewed published articles was also manually searched for additional references. Search results were downloaded to Rayyan [21] and duplicates were removed.

Study selection

Two authors independently screened studies to assess eligibility for inclusion in the study according to the pre-specified inclusion and exclusion criteria (NYA and MA). Disagreements were resolved through the third reviewer (RGE). Full-text studies were screened and reviewed by two independent reviewers using the pre-specified inclusion and exclusion criteria (NYA and MA). Disagreement was resolved through discussion with a third reviewer (RGE) and reasons for exclusion recorded. Study selection was done through Rayyan [21]. The selection process was reported using the PRISMA flow diagram (see Figure 1).

Figure 1. Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) flow diagram [16].

Data extraction

For each included study, data were extracted using customized electronic data extraction form, which was piloted by two reviewers (NYA and MA) before the data extraction phase. Data extraction was completed by one reviewer (NYA) and checked for accuracy by a second reviewer (MA). Corresponding authors were contacted if data were ambiguous or missing.

Risk of bias and quality assessment

Risk of bias was assessed using two quality assessment tools. RCTs were evaluated with the revised Cochrane tool for assessing the risk of bias in RCT [22], with the risk of bias classified as high, some concern or low for each study. Non-randomized studies were assessed using NIH Quality Assessment Tool for Before-After (Pre-Post) Studies with No Control Group [23], with the quality classified as good, fair or poor. Two reviewers independently checked each selected article (NYA and MA). Disagreement was resolved through discussion.

Data synthesis

Outcomes of the included studies were compared and a narrative analysis was performed. Due to the high degree of heterogeneity (with different tests of small airways, different definitions of exacerbations used and different outcome measures), a meta-analysis of included studies was not possible.

Results

Of 1436 citations (excluding duplicates), 154 relevant studies met inclusion criteria. Thirteen were non-English articles (four German, two Russian, two Chinese, one Korean, one Polish, one Bulgarian and one French) but were included for full text screening. After the full-text screening, seven articles were eligible to be included in the review and 147 were excluded with exclusion reasons documented. Of the seven articles included, only one was not in English and was written in Russian [24] (See Figure 1).

Of the included studies, six studies were non-randomized [24–29], and one was RCT [30]. Five studies were hospitalized exacerbations [24,26–29], one included both hospitalized and community exacerbations [25], and one included community exacerbations only [30]. Different definitions of exacerbation were used in most studies and they are reported in (Table 1). The duration of included studies ranged from 5 to 60 days, with a median time of 14 days (Interquartile Range (IQR) 8 to 42). The sample size across all studies ranged between 20 to 87 participants, with a median size of 29 participants (IQR 22.5 to 52.5). The characteristics and the main findings of the seven included studies are detailed in the online supplement.

Table 1. Definition of exacerbation in the included studies:.

Author	Definition of exacerbation
Stevenson et al. [25]	An increase in at least two major symptoms: dyspnea, sputum purulence, or increased sputum volume, sufficient to require hospitalization.
Sahn et al. [26]	Not explained
Yetkin et al. [28]	Episodes of worsening breathlessness and/or wheezing, often accompanied by greater volume or purulence of sputum and increased cough
Parker et al. [25]	According to the criteria adopted by the Canadian Thoracic Society, ‘‘a sustained worsening of dyspnea, cough or sputum production leading to an increase in the use of maintenance medication and/or supplementation with additional medications’’
Komlev et al. [24]	an increase in cough, an increase in the amount of sputum and a change in its character, an increase in shortness of breath and a decrease in exercise tolerance, moderate signs of intoxication syndrome that requires therapy.
Jetmalani et al. [29]	GOLD definition “an acute event characterized by a worsening of the patient’s respiratory symptoms that is beyond normal day-to-day variations and leads to a change in medication”
Johnson et al. [26]	Increased breathlessness for at least 24 h with at least two criteria of increased cough frequency or severity, increased sputum volume or purulence or increased wheeze.

Notes: This table demonstrates how exacerbations were defined among included studies. Each row provides the definition of exacerbations used in the stated study. Although they have defined exacerbations differently, most of the studies’ definitions met current international guidelines of exacerbation.

Of the seven included articles three studies used FOT [26,27,29], four studies used MMEF [24,25,28,30] and two studies used Raw by body plethysmography [24,25]. Using small airways tests and FEV₁, treatment response was assessed in all of the included studies. However, included studies varied in how they reported response to treatment. Some studies reported pre- and post-measure [24,27–30] while, in the others, pre-measure and the changes in the outcome measured were reported [25,26].

Quality assessment of included studies

All non-randomized studies had clearly stated their objective, research question and eligibility criteria. However, three of the studies did not justify the sample size of the study [24,28,29], although two had the highest number of participants among the included studies [24,28]. Participants of the included studies were representative of the general population of interest in all of the studies. Four of the studies had a number of visits where measurements were taken [25–27,30]. One study was felt to be of good quality [26], two of fair quality [24,28] and three of poor quality [25,27,29] (as shown in Table 2). The risk of bias of the only included RCT [30] was found to be high (as shown in Table 3).

Table 2. NIH quality assessment tool of included non-randomized studies.

NIH Quality assessment	Yetkin et al. [23]	Jetmalani et al. [29]	Parker et al. [25]	Johnson et al. [26]	Stevenson et al. [25]	Komlev et al. [24]
Was the study question or objective clearly stated?	Yes	Yes	Yes	Yes	Yes	Yes
Were eligibility/selection criteria for the study population prespecified and clearly described?	Yes	Yes	Yes	Yes	Yes	Yes
Were the participants in the study representative of those who would be eligible for the test/service/intervention in the general or clinical population of interest?	Yes	Yes	Yes	Yes	Yes	Yes
Were all eligible participants that met the prespecified entry criteria enrolled?	CD	CD	CD	CD	CD	CD
Was the sample size sufficiently large to provide confidence in the findings?	Yes	CD	No	Yes	No	Yes
Was the test/service/intervention clearly described and delivered consistently across the study population?	Yes	Yes	Yes	Yes	Yes	Yes
Were the outcome measures prespecified, clearly defined, valid, reliable, and assessed consistently across all study participants?	Yes	Yes	Yes	Yes	Yes	Yes
Were the people assessing the outcomes blinded to the participants' exposures/interventions?	NA	NA	NA	NA	NA	NA
Was the loss to follow-up after baseline 20% or less? Were those lost to follow-up accounted for in the analysis?	CD	No/NA	Yes/CD	Yes/Yes	Yes/Yes	CD
Did the statistical methods examine changes in outcome measures from before to after the intervention? Were statistical tests done that provided p values for the pre-to-post changes?	Yes	Yes	Yes	Yes	Yes	Yes
Were outcome measures of interest taken multiple times before the intervention and multiple times after the intervention (i.e., did they use an interrupted time-series design)?	Yes	Yes	CD	Yes	Yes	CD
If the intervention was conducted at a group level (e.g., a whole hospital, a community, etc.) did the statistical analysis take into account the use of individual-level data to determine effects at the group level?	CD	CD	CD	CD	CD	CD
Overall Quality rating	Fair Quality	Poor Quality	Poor Quality	Good Quality	Poor Quality	Fair Quality

Notes: Quality assessment was done using NIH tool. Each question was assessed by two reviewers separately (NYA. and MA), and overall rating was blindly determined. Results were then discussed, and Conflicts in questions and overall rating resolved through discussion. The overall quality rating for included studies is the agreed rating between reviewers.

Abbreviations: CD = Cannot determine, NA = Not applicable.

Table 3. Risk of bias assessment of the included RCT.

Study ID	Randomization process	Deviations from intended interventions	Missing outcome data	Measurement of the outcome	Selection of the reported result	Overall
Sahn et al. [26]	High	High	High	Some concerns	Some concerns	High risk of bias

Notes: Risk of bias was evaluated using Revised Cochrane tool for RCT. The tool assesses risk of bias on several domains: Randomization process, deviations from intended interventions, missing outcome data, measurement of the outcome and selection of the reported results. The risk of bias was evaluated as recommended: high risk, some concerns, or low risk. Each domain was assessed by two reviewers independently (NYA and MA). Conflict in domains and overall risk of bias were resolved through discussion.

Forced oscillation technique (FOT)

FOT was used to assess treatment response in COPD exacerbation in three observational studies. Impulse oscillometry (Jaeger MasterScreen (IOS)), which is a later version of FOT, was used in one study. Stevenson et al. [27] used IOS and FEV₁ to assess treatment response during exacerbation in a longitudinal study. Here, after the baseline assessment (day 1 of the exacerbation), IOS and FEV₁ were repeated at day 2, day 3, discharge day and day 42. The median length of hospital stay was 7 days (range 3-10 days). The authors only reported R5 (total respiratory resistance) and X5 (total respiratory reactance), with both reported as mean ± standard deviation (SD). R5 did not show any improvement during the treatment period. X5 showed an average improvement of 26% from baseline (-0.42 ± 0.03 kPa/L/s) to discharge day (-0.31 ± 0.03 kPa/L/s; p < 0.001) and a 33% improvement from baseline to day 42 (-0.28 ± 0.04 kPa/L/s; p < 0.01). FEV₁ also improved during the treatment period from a mean of 1.03 ± 0.08 L at baseline to 1.08 ± 0.08 L (p < 0.05) at day 2 with no changes at day 3. At discharge, FEV₁ increased to 1.12 ± 0.09 L (p < 0.05) (an average of 8% improvement) but this did not meet the accepted Minimal Clinically Important Difference ((MCID) of at least 100 ml) for FEV₁ [31]. The 100 ml MCID was not observed until day 42 (1.26 ± 0.10 L; p < 0.01), where there was an average 22% improvement from baseline.

Jetmalani et al. [29] who used an in-house-built FOT device at 6 Hz, assessed if expiratory flow limitation ((EFL) representing a physiological condition in which expiratory flow cannot rise by increasing respiratory effort [32]) related to symptoms during hospitalized exacerbations of COPD. The authors categorized patients into two groups: those with EFL at admission and those with no EFL at admission. They performed assessments on admission and prior to discharge, with mean length of stay being 5 days (±1 SD). All FOT results were reported as mean ± SD. Resistance parameters (Rrs (total respiratory resistance), Rrs_insp (resistance during inspiration), Rrs_exp (resistance during expiration)) did not change in either group. Reactance parameters improved in the EFL group prior to discharge. Although Xrs_insp (reactance during inspiration) showed improvement (increasing from −3.73 ± 0.97 cmH₂O⋅s⋅L⁻¹ to −2.90 ± 0.87 cmH₂O⋅s⋅L⁻¹, p = 0.01), Xrs (total respiratory resistance), Xrs_exp (reactance during expiration), ΔXrs (expiratory flow limitation index, which is used to measure EFL (Xrs_insp - Xrs_exp)) demonstrated the most pronounced improvement (Xrs increased from −7.37 ± 2.27 cmH₂O⋅s⋅L⁻¹ to −4.41 ± 1.92 cmH₂O⋅s⋅L⁻¹ (p = 0.008), Xrs_exp from −8.70 ± 3.19 cmH₂O⋅s⋅L⁻¹ to −5.12 ± 2.33 cmH₂O⋅s⋅L⁻¹ (p = 0.008), and ΔXrs from 4.97 ± 2.64 cmH₂O⋅s⋅L⁻¹ to 2.21 ± 1.51 cmH₂O⋅s⋅L⁻¹ (p = 0.008)). FEV₁ did not change in either group.

Johnson et al. [26] used FOT (with an oscillation frequency of 5 Hz) and FEV₁ to assess treatment response across three visits (<48 h of admission, after one week and after 6 weeks). All results are reported as mean and standard error of mean [8]. No significant changes were seen in resistance parameters (Rrs, Rrs_insp, Rrs_exp) across timepoints. Following one week of admission, mean Xrs_insp increased by 13% (4.5 SEM) (p < 0.05); Xrs by 27.9% (7 SEM) and Xrs_exp 31.5% (7.8 SEM) (p < 0.001 for both). After 6 weeks, Xrs_insp increased by 27.4% (6.7 SEM) (p < 0.001), Xrs by 35.2% (8.9 SEM) and Xrs_exp by 37.1% (10.0 SEM) (p < 0.005 for both). FL% (flow limitation percentage, which represents the proportion of breaths for which ΔXrs indicated the flow limitation) improved at visit two (decreasing by 19.2 ± 6.1%, p < 0.005) and decreasing by 19.5 ± 7%, (p < 0.05) at visit 3. FEV₁ also improved, meeting the MCID criterion at visit two (improvement of 0.153 L ± 0.046 (6.4 ± 1.7%; p < 0.005)) and three (0.274 L ± 0.064 (11.4 ± 2.3%; p < 0.001)). When expressed as a percentage changes, reactance parameters improved more than FEV₁.

MMEF

MMEF was used in four studies. Sahn et al. [30] (who included only male patients in their RCT) used MMEF and FEV₁ to assess the response to high (1000 mg three times a day (TID)) or standard (500 mg TID) dose amoxicillin during an episode of acute bronchitis in patients with COPD (reflecting the terminology of exacerbations at that time). However, they reported the results as only one mean and SD irrespective of the different doses of antibiotic being given. MMEF and FEV₁ (pre and post-bronchodilator) were collected at day 0, day 2 and day 7 with MMEF changing throughout the treatment period (p < 0.05). At day 2, pre-bronchodilator MMEF increased from 1.25 ± 0.26 L/s to 1.46 ± 0.28 L/s (p < 0.05) but there was no difference in pre-bronchodilator MMEF when day 7 (1.39 ± 0.29 L/s) was compared to baseline. Post bronchodilator MMEF improved only at day 7, where it rose from 1.35 ± 0.28 L/s to 1.67 ± 0.32 L/s (p < 0.05) (an average 24% increase). FEV₁ also improved during the treatment period, meeting the MCID criterion. Pre-bronchodilator FEV₁ increased from 1.65 ± 0.22 L to 1.82 ± 0.22 L at day 2 and to 1.81 ± 0.23 L at day 7 (p < 0.05). Post-bronchodilator FEV₁ rose from 1.80 ± 0.23 L to 1.94 ± 0.22 L (p < 0.05) at day 2 and to 1.98 ± 0.23 L (p < 0.05) at day 7 (an average 10% increase).

MMEF was also used in two observational studies. Yetkin et al. [28] used MMEF (termed FEF_25-75 in their study (L/s and %)) and FEV₁ to assess treatment response during an exacerbation, collecting readings on admission day and discharge day (mean length of stay 9 days ± 2). All results were reported as mean and SD. MMEF and FEV₁ changed at discharge (p < 0.05): MMEF increased from 0.43 ± 0.14 L/s to 0.52 ± 0.14 L/s (p < 0.05) (an average 20% increase) and FEV₁ increased from 1.14 ± 0.29 L to 1.22 ± 0.32 L, (p-value <0.05) (an average 7% increase), which FEV₁ did not meet the MCID.

Parker et al. [25] assessed both Inspiratory Capacity (IC) and MMEF in comparison to FEV₁ to evaluate the severity and recovery of acute exacerbations of COPD in a longitudinal study with multiple visits (admission day, day 7, day 14, day 30 and day 60). All results were reported as mean and SEM. MMEF did not show any change throughout the treatment period. In contrast, FEV₁ improved during the recovery period of the exacerbation although changes above the MCID criterion were only seen from day 14 (FEV₁ increased by 0.12 ± 0.06 L at day 14, 0.13 ± 0.06 L at day 30 (p < 0.05) and by 0.24 ± 0.06 L at day 60 (p < 0.01))

In an interventional study, Komlev et al. [24] also used MMEF and FEV₁, assessing the effect of systemic glucocorticoid steroids (SGS) during acute exacerbations of COPD from admission to day 14 of treatment. All results were reported as mean and SD. In their study, MMEF demonstrated improvements at day 14 (rose from 15.7 ± 7.0% to 21.5 ± 14%, p < 0.05; an average improvement of 37%). FEV₁ also demonstrated improvements (increased from 41.1 ± 14% to 50.1 ± 16%, p < 0.05; an average 22% increase). They also separated patients into two groups: group 1 (>15% change in FEV₁ at day 14 of SGS) and group 2 (<15% changes in FEV₁ at day 14 of SGS). At day 14, group 1 demonstrated improvements in MMEF and FEV₁ (p < 0.05) with no change in group 2. In group 1, MMEF increased from 12.4 ± 7.0% predicted to 28.8 ± 18.0% predicted at day 14 of SGS treatment (p < 0.05), and FEV₁ increased from 34.2 ± 11.0% to 60.5 ± 14.0%. Although it could not be determined whether the improvement in FEV₁ met MCID criteria as actual values were not given, a 25% change represents a clinically relevant change.

Raw by body plethysmography

Raw obtained by body plethysmography was used to assess treatment response in two studies. Parker et al. [25] reported MMEF, FEV₁ and specific airway resistance (sRaw) in their observational study at admission, day 7, day 14, day 30 and day 60 (as described previously). All results were reported as mean and SEM. Although sRaw decreased at day 7 and day 30, it only demonstrated significant improvement at day 14 and day 60 (p < 0.05). Compared to the baseline assessment, sRaw changed by 133 ± 57% and 128 ± 51% at day 14 and 60, respectively (p < 0.05) with FEV₁ improving most at day 60 (p < 0.01).

As described previously, Komlev et al. [24] assessed airway resistance during inspiration [18] and expiration (R_ex) at admission day and day 14 of SGS. All results were reported as mean and SD. R_ex showed significant improvement (decreasing from 1.05 ± 0.66 kPa/L/s to 0.76 ± 0.5 kPa/L/s; p < 0.05), while R_in did not. When separated into the two groups, R_in did not change at day 14 in either group. In contrast, R_ex decreased both in group 1 (from 1.02 ± 0.5 kPa/L/s to 0.57 ± 0.4 kPa/L/s; p < 0.05) and group 2 (from 1.06 ± 0.7 kPa/L/s to 0.84 ± 0.5 kPa/L/s; p < 0.05), with FEV₁ improving only in group 1 (p < 0.05).

Discussion

This is the first systematic review evaluating the use of physiological tests of small airway function during exacerbations of COPD. The aim of this review was to assess whether there was sufficient evidence to incorporate measures of small airways function as a primary outcome for studies of COPD exacerbations. In summary, the small number of studies, low participant numbers, study heterogeneity and general mixed quality of the studies limits any conclusions about the utility of these tests during exacerbations. However, in most studies, there were early signals of small airways measurement change following the diagnosis and treatment of an exacerbation. This was reported (albeit not consistently) for FOT, MMEF and sRaw, often at an increased magnitude or prior to an FEV₁ change. This suggests further studies are warranted to determine if these physiological tests provide information about the early recovery phase of exacerbations. However, more information is needed to determine population reference ranges, which test should be used in which setting and which devices provide comparable results.

Exacerbations are common and serious events and are associated with reductions in quality of life, health status and lung function [10]. Diagnostic tools and treatment approaches have not changed significantly for over 30 years [33] but there is great interest in testing new therapies to prevent or treat exacerbations and finding more sensitive and specific tools to diagnose exacerbations and map their recovery. It is increasingly recognized that small airways dysfunction and damage is an early feature of COPD [6,8] and these processes may also be implicated in exacerbations [34,35]. Exacerbations are usually associated with an increase in airway inflammation, impacting on small airways through airway narrowing, mucus hypersecretion and sometimes plugging [34,35]. In light of this, small airways tests have been proposed as potential tools that could be used in studies of COPD [36].

In general, published studies of tests of small airways function during exacerbations were small and heterogeneous, using different small airway tests and different definitions of an exacerbation. The methodological quality of these studies was variable and mainly of fair to poor quality. There was only one study of good quality [26], with two of fair quality [24,28], three of poor quality [25,27,29] and one trial rated as having a high risk of bias [30]. Small airways tests along with FEV₁ demonstrated improvement following the recovery from exacerbation in the majority of the included studies but different studies reported improvements at different time points, even with the same test.

Despite these important limitations, there were suggestions of an earlier signal of change from exacerbation onset in the tests of small airways compared to FEV₁. For example, for FOT: while Stevenson et al. [27] reported a difference in FEV₁ across all time points, there was only an average improvement of 8% in FEV₁ at discharge in contrast to an average improvement of 26% in X5 (with average length of stay 7 days). Johnson et al. [26] reported improvements in FEV₁ as well as FOT, with FOT having a greater percentage change than FEV₁. Jetmalani et al. [29] did not find a difference in FEV₁ at discharge (day 5) despite describing difference in measures of FOT. For MMEF, Sahn et al. [30] described an average increase of 24% in MMEF but only a 10% increase in FEV₁ at day 7 post-exacerbation. Komlev et al. [24] reported a greater percentage increase in MMEF than FEV₁ at day 14 post-exacerbation and improvements in R_ex. Yetkin et al. [28] described a 20% increase in MMEF but only a 7% increase in FEV₁ between admission and hospital discharge (with an average admission length of 9 days). Parker et al. [35] described no significant change in MMEF or FEV₁ within the first two weeks following exacerbation onset but significant improvements in FEV₁ and sRaw after day 14. When interpreting these data, the variance of results should be carefully considered alongside the average percentage change, but these results support some small airways signal during exacerbations for at least some COPD patients.

Although tests of small airway and FEV₁ seems to mirror each other in terms of physiological improvement, (especially in the later recovery phases of an exacerbation), obtaining FEV₁ measurement can be challenging (as spirometry is a maximal forced maneuver) [2] compared to some small airways’ tests (for example, IOS, FOT) where patients perform tidal breathing. Furthermore, patients during exacerbation may be too unwell to perform spirometry. Potentially, FOT may be a more acceptable test to perform at the bedside but, unfortunately, the included studies did not report completeness of data at each testing time for each test. Therefore, it is unclear if patients were more likely to complete tidal breathing assessments as opposed to forced maneuvers.

Three studies used FOT with the most significant changes were seen in reactance parameters (Xrs, Xrs_insp, Xrs_exp, ΔXrs, X5). All studies assessed respiratory impedance at a lower oscillation frequency. In Stevenson et al. IOS assessed resistance and reactance at an oscillation frequency of 5 Hz, describing that both X5 and FEV₁ improved from baseline to follow up, with no change in R5 [27]. Jetmalani et al. [29] used FOT to measure respiratory impedance at an oscillation frequency of 6 Hz. This study showed that reactance parameters (but not resistance parameters or FEV₁) changed significantly from onset to recovery in patients with EFL on admission but no changes in reactance, resistance or FEV₁ in those without EFL. These findings and others [27] suggest that not all COPD patients have EFL during exacerbations. In those that do, EFL is associated with changes in reactance parameters at lower frequency and EFL may identify a sub group most likely to provide a signal of improvement using FOT. Johnson et al. [26] used FOT at 5 Hz oscillation frequency and also reported that FEV₁ and reactance parameters changed significantly while resistance did not. Although Jetmalani et al. have used an atypical FOT frequency, there are no marked differences between them as frequencies in the 3 − 10 Hz range, with all appearing to demonstrate a good sensitivity to airway caliber [37]. Furthermore, several studies have used FOT in this range, and the three included studies reported the same findings of FOT. Reactance parameters in all FOT studies (but not in those without EFL at baseline [29]) showed improvements during the recovery from exacerbation. Reactance is frequency-dependent (changes as oscillation frequency changes) due to the heterogeneity across tissue viscoelasticity, airway tree, airway wall shunt, and time constant [18]. At a lower oscillation frequency, reactance becomes more negative as the mechanical characteristics of the lung and chest wall dominate, which reflects greater elastance (or stiffness) of the oscillation mechanics in accordance with the impedance equation. Furthermore, frequency dependence may be affected by the presence of heterogeneous ventilation resulting from airway diseases [18], and at any given frequency, effective reactance decreases as heterogeneity increases. The reason for the improvement of reactance described in the included studies is unclear but it may be due to the reduction in hyperinflation after bronchodilators, increasing lung compliance over tidal breathing range. Although several studies have indicated that resistance at lower frequency (especially R5) is also sensitive to change after bronchodilators [19,38], R5 represents the total airway resistance and does not exclusively assess small airways. FOT may be of use during exacerbations of COPD but assessing resistance at different oscillation frequencies (up to 20 Hz) might provide more information as, although resistance at higher oscillation frequencies does not changes after bronchodilator therapies [38], it can be used to calculate the difference between R5 and R20 (R5-20). R5-20 is used to determine the contribution of peripheral airways which might give an insight about small airway function. However, the anatomical location of the transition between the distal and proximal airways has not been identified [18,39]. Therefore, future studies should aim to include both R5 and R20, so R5-20 can be calculated. Impedance measurements often differ when measured by different devices, making comparison between studies difficult, therefore future work should consider if a gold standard technique can be identified, to allow inter-study learning [40].

Three out of four studies that included MMEF described an improvement in readings over the course of an exacerbation despite differences in the treatments being given and the place of care. Despite the need to correct MMEF for lung volumes (to increase reproducibility) studies did not report if this adjustment had been made. Sahn et al. [30] evaluated two different doses of amoxicillin and described improvements in both MMEF and FEV₁ during the treatment period. However, the treatment groups were combined into one reported outcome, suggesting a high risk of reporting bias. Yetkin et al. [28] also did not indicate whether MMEF was adjusted for lung volumes but both MMEF and FEV₁ improved over the course of an exacerbation. Komlev et al. [24] described an improvement in MMEF in patients when reported results as a whole. However, when they did a post hoc analysis the improvements were seen in patients who had increase in FEV₁ by >15% after 14 days of SGS treatment (group 1) but not in group 2, where no improvement in FEV₁ was seen.

Only one study did not report an improvement in MMEF [25], although an improvement in FEV₁ was described. Potential reasons for this might include that their study was the smallest among those including MMEF (with 20 patients and three patients who withdrew from the study). Secondly, they included older and more severe patients than the other studies, where significant small airways destruction is likely to already have occurred. Lastly, they did not indicate whether MMEF was adjusted for lung volumes and changes in FVC may impact the reproducibility of MMEF.

MMEF has a wide reference range in clinical practice which might limit its interpretation, and results prone to considerable variability [41]. Furthermore, MMEF is obtained by performing spirometry; a forced maneuver that COPD patients may struggle to perform during an exacerbation. Additionally, established COPD is associated with severely lowered MMEF (approximately 17% predicted in one study [17]) perhaps limiting its ability to identify change.

Two studies reported airway resistance by body plethysmography. Parker et al. [25] reported that sRaw showed significant % changes during the treatment period (with no change in MMEF) but did not demonstrate the method of obtaining sRaw. Komlev et al. [24] reported Raw differently (assessing Raw in two phases; expiration and inspiration). Their study showed that R_ex changed significantly at day 14 of treatment while R_in did not change. Furthermore, when looking at the analysis of the two groups, R_ex changed significantly at day 14 of treatment in both groups while R_in only changed in group 1 (FEV₁ >15% at day 14 of SGS treatment). However, the method of obtaining Raw was not described. There are limits to the interpretation of these findings. Measures of airway resistance may be influenced by increased resistance of the small or large airways, which may be variable during exacerbation and prone to alteration with breathing effort. Furthermore, inflammation and airflow obstruction in the larger airways may render the small airways less accessible for physiological testing and although airway resistance by body plethysmography may be a valuable tool, the evidence of its sensitivity for small airway changes is still uncertain. Moreover, performing body plethysmography in the acute setting may be challenging as the test cannot be done at the bedside.

Limitations and implication for research and clinical practice

This systematic review was limited by different exacerbation definitions being utilized across included studies and the practical challenges of summarizing and synthesizing data due to their heterogeneity. Tests of small airways function appear to change during the course of an exacerbation of COPD; hence, more well-structured studies are needed to determine the usefulness of each test. Although most of the included studies showed some changes in small airway tests, it remains unclear which tests of small airways function are the most sensitive to change, which are best tolerated by patients and which are the most practical to deliver. Better designed, larger studies that compare a number of tests and better characterize the exacerbation and the patient with COPD are needed.

Conclusion

Exacerbations are common and life-threatening events that may have an impact on health status and lung function. Despite this, diagnostic methods and management approaches have been largely unchanged significantly for over 30 years [33]. There is a critical need for sensitive physiological measures to objectively assess and map exacerbations recovery. Small airway dysfunction is a pathophysiological feature of COPD and it may be worsened by the increased inflammation present during exacerbations. Small airway tests may be valuable in recognizing exacerbations and following their recovery or response to treatment (either as a complimentary test to FEV₁ or in specific phases of the recovery process), although further evidence is needed to support their use.

Authors contributions

NYA and ES designed the study. NYA performed the initial search and data extraction. NYA, RGE and MA assessed the eligibility of the included studies. NYA and MA performed the quality assessment for the included studies. NYA wrote the initial manuscript. RGE, ES and JS revised the manuscript. All authors read and approved the final version of the manuscript.

Declaration of interest

The authors report no conflict of interest.

Acknowledgement

The authors would like to thank Adam Seccombe for his help and support in refining the search strategy. RGE was supported by a Clinical Trials Fellowship (NIHR300008), supported by the National Institute for Health Research (NIHR), outside of the scope of this work. The views expressed are those of the authors and not necessarily those of the NHS, the NIHR, or the Department of Health.

Additional information

Funding

The current study was undertaken as part of NYA’s PhD. NYA is supported by King Saud bin Abdulaziz University for health sciences through the Saudi Arabian Cultural Bureau in the UK.