Minimal Clinically Important Differences in the Six-Minute Walk Test and the Incremental Shuttle Walking Test

Abstract

Simple walking tests are widely used for the assessment of functional status in patients with cardiorespiratory disorders. These tests require far less instrumentation than formal cardiopulmonary exercise tests, but they do require standardization of procedures to achieve reproducible results. The most widely used tests for patients with COPD are the 6-minute walking test (6MWT) and the incremental shuttle walking test (SWT). The 6MWT has been characterized in COPD patients with respect to reproducibility and responsivity to change in health status. The 6MWT results are correlated with pulmonary function, health-related quality of life, maximum exercise capacity, and mortality. The minimal clinically important difference (MCID) for the 6MWT is conservatively estimated to be 54–80 meters using both distributional and discriminative methods. For an individual patient, the 6MWT would need to change by about 86 meters to be statistically confident that there has been a change. The SWT has been less extensively validated than the 6MWT, but has similar reproducibility in COPD (CV = approximately 20%). The SWT results improve with pulmonary rehabilitation and bronchodilation, and are highly correlated with maximum oxygen consumption. There are no studies that address the issue of MCID for the SWT. In addition to the MCID, the design and interpretation of COPD clinical trials should take into account the severity of initial impairment, the asymmetry between positive and negative changes, the proportion of patients who show substantial improvement, and the costs and risks of the treatment.

• Exercise test
• Minimal clinically important difference
• Shuttle test
• COPD
• Differential threshold
• Lung diseases/obstructive/diagnosis
• MeSH

Introduction

Exercise tests are commonly used to evaluate disability and response to treatment in COPD patients. In recent years, there has been increasing interest in simple exercise tests that do not require complex technology to perform, and which are more closely relevant to daily activities than cycle ergometer oxygen consumption tests. The most widely used simple exercise test is the six-minute walk test (6MWT), which can be performed in most clinical settings and does not ordinarily require close medical supervision. In Europe, there is also widespread use of the shuttle walking test (SWT), which is a progressive exercise test that uses a taped metronome to coach patients to perform incrementally faster walking speeds to the point of intolerance. Both the 6MWT and the SWT have been validated against formal exercise tests and are responsive to therapeutic interventions.

The purpose of this manuscript is to describe the performance, validity, reliability, and interpretation of the two widely used walking tests. In particular, the goal of this review is to provide information that may be used to determine minimal clinically important differences in these tests forpatients with chronic obstructive lung disease. Understanding the minimal clinically important difference for these tests will permit clinicians to interpret changes in individual patients, and will facilitate the planning and interpretation of clinical trials that compare treatment effects on these exercise tests.

Description of the Six-Minute Walk Test (6MWT)

In 1968, Cooper introduced the 12-minute running test as a measure of physical fitness in young men [[1]]. Soon thereafter, the test was modified to a 12-minute walking test to assess functional impairment in patients with respiratory diseases [[2]]. Recognizing that many patients with chronic respiratory illness had difficulty completing a 12-minute walking test, a group of investigators demonstrated that the 6-minute walking distances were highly correlated to those obtained during a 12-minute walking test [[3]]. Since that time, the 6-minute walk test (6MWT) has been adopted as the most widely used functional test of exercise capacity given its ease of administration and its ability to reflect function during daily activities. Additionally, the 6-minute walking distance is now widely used for measurement of treatment response in clinical practice and clinical trials [[4]].

The 6MWT is used as a measure of functional status in patients with a wide variety of diseases, but has been particularly helpful for determining prognosis and response to treatment in patients with pulmonary hypertension, congestive heart failure, and COPD [5-10]. The test is conducted by having the patient walk as far as possible on a flat indoor course for a period of 6 minutes. The major outcome is the distance walked during the 6 minutes. Secondary outcomes may include oxygen saturation by pulse oximetry, ratings of dyspnea using a Borg or visual analog scale, and heart rate.

The major advantages of the 6MWT are that the test requires minimal technical resources and involves a familiar daily activity. Pacing and strategy are thought to play an important role in the performance on the 6MWT as there is a small but definite improvement on sequential tests over a short time interval in certain populations. This likely reflects learning effect rather than true physical training. For example, in patients with COPD, the 6-minute walk distance improved by a statistically significant average of 66 feet on the second consecutive day of testing. Similar effects have been demonstrated in patients with heart failure when tested on consecutive days [11&12]. It is also thought that coaching affects the 6-minute walk distance, and therefore this component of the test should be standardized, including the explanation of the test and the content of periodic messages [13&14].

Despite the apparent simplicity of this test, there may be many factors that influence outcomes including walking surface, length of the laps, shape of the walking course, patient explanations and coaching, use of oxygen and whether the oxygen device is carried by the patient, time of day, relationship to meals and bronchodilators, number of tests performed and rest interval between tests, and how the data from sequential tests is analyzed (e.g., mean, maximum, or final test). Recently, the ATS established pulmonary laboratory guidelines for performance of the 6MWT [[14]].

Validity of the Six-Minute Walk Test

Although the 6-minute walk distance is often considered to be a test of endurance rather than maximum work capacity, it does correlate with maximum oxygen uptake in patients with pulmonary hypertension, congestive heart failure, and COPD [[6]][8-10][[15]]. The 6-minute walk distance also correlates moderately with measures of health-related quality of life in COPD, measures of dyspnea, peak work capacity by cycle ergometry and pulmonary function [[4]][15-19]. The correlation coefficients of 6MWT with FEV₁, FVC, and DLCO are typically in the range of 0.5 to 0.6, whereas the correlation coefficient with maximum oxygen consumption is better, ranging from 0.5 to 0.8 [[15]][17&18]. The 6-minute walk is used as an outcome measure in clinical trials and has been shown to improve with bronchodilators [[20]], exercise training [[21]], mucus clearance devices [[22]], and lung volume reduction surgery in COPD [[5]][[23]]. The effect of long-acting bronchodilators on 6-minute walk distance has been inconsistent [[24]]. When the 6MWT is severely reduced, it has prognostic value. In patients with mild or moderate heart failure, a 6-minute walk distance of < 300 meters has been demonstrated to predict excess mortality of 20% at 34 months when compared to those walking greater than this distance [[25]]. Additionally, patients with a 6MWT less than 200 meters had an 84% specificity for predicting 6-month mortality following lung volume reduction surgery [[26]]. There were no deaths among those walking greater than 200 meters. In a mixed cohort of patients with far-advanced lung disease, those with less than 300-meter 6MWT had a 80% mortality [[8]]. Recently, a group of investigators have validated a prediction score for the risk of death among patients with COPD that includes 6-minute walk distance as one of its four components with a value of less than 350 meters being associated with increased risk of death [[27]].

Minimal Clinically Important Difference in 6MWT

Patient-Based Estimate of MCID

Some data are available upon which to establish minimally clinically important differences in the 6-minute walk test. Redelmeier and colleagues compared changes in actual distance walked in 6 minutes to patients' perceptions of their changes in their functional status over a period of several months [[28]]. Patients were found to have poor correlation of perceived improvement or worsening in terms of walking this objective measure (r = 0.20). Additionally, more than half of patients who perceived that they had improved actually showed deterioration in walking test. This result was attributedto poor memory of prior health status. Therefore, patients were asked to rate changes in themselves in comparison to eight other patients with COPD whom they observed in a rehabilitation program. The differences in 6MWD were computed for each perceived level of contrast: about the same, a little bit worse, somewhat worse, much worse, a little bit better, somewhat better, and much better. The minimally clinically important difference was determined from the difference between about the same and a little bit better or worse. The mean difference for those suggesting that they were “about the same” was − 10 m; for “a little bit better” was 30 m; and “a little bit worse” was − 80 m. Thus, there tended to be an optimistic bias, whereby patients would rank themselves equal or better than their peers with smaller differences than was required to rank themselves as worse. The average change need to rank oneself as better than one's peers was about 40 m (30 minus − 10 m) and to rank oneself as worse was 70 m (− 80 minus − 10m). Based on this analysis, the mean difference between those ranking themselves as a bit better or worse was 54 m (95% CI = 37–71 m). The magnitude of the difference did not vary significantly based on clinical or demographic measures or severity of baseline impairment. Based on this analysis, an improvement of 54 meters has been the value most commonly used as the least change in 6MWT that results in a clinically significant change in functional status [[14]].

Distribution-Based Estimate of MCID

Statistical approaches can also be used to define the MCID of the 6MWT by applying the notion of reproducibility of the measure to mathematically determine the least change in an individual that can be confidently assumed to be outside the range of random variation. Different approaches to this stochastic methodology have been proposed including the range of 1 standard error of measurement, one-half standard deviation, or 95% confidence limits. The largest data set of COPD patients undergoing repeat 6MW tests has been collected by Sciurba and colleagues in the National Emphysema Treatment Trial (NETT) [[11]]. On two separate occasions, 470 patients with severe emphysema undertook 6-minute walk tests. The second test was 23 m ± 44 m longer than the first with an intra-class correlation of 0.88. The standard error of the measurement, which was directly measured in this study is 44 meters. The indirect calculation of the standard error of the measurement, calculated as the revised Jacobsen formula gives a similar value of 45.3 meters, where S is the population standard deviation and r is an estimate of the reliability coefficient based on the intraclass correlation [[29]]. A direct calculation of the reliability coefficient is considerably lower than the intraclass correlation presumably, in part, because of the learning effect that makes the second walking distance longer than the first if repeated within a short interval. The reliability coefficient for repeated measurements for the 6MWT is 0.63 [[30]], which gives a calculated MCID of 80 meters with the revised Jacobsen formula, taking into account the short-term learning effect which averaged about 20 meters [[11]].

The coefficient of repeatability, then, is 86 m (1.96 × 44) which is 23% of the mean value of 371 m. This is similar to the upper 95% confidence bounds for the minimal clinically important difference based on self-comparison with other patients, which is recommended by the ATS as a threshold for comparison of change in a single patient between two test sessions [[14]].

The population of NETT patients had a mean 6MWT of 93 m, yielding a MCID based on the ½ SD of the population distribution of 47 m. Thus, the estimate of MCID is similar, whether it is based on self-assessment or distributional approaches [[31]]. As stated previously, 54 meters is often used as the MCID for the 6MWT based on the work of Redelmeier and colleagues [[30]]. However, it is important to point out that this is a mean improvement and that the 95% confidence interval includes a range from 37–71 m [[30]]. This range is comparable to the values that were statistically determined before here.

Incremental Shuttle Walk Test (SWT)

The shuttle walk test is derived from field tests of maximum exercise capacity and has been demonstrated to correlate with VO₂ max [[32]]. The test is performed by having the subject walk at a pace dictated by a tone recorded on a cassette player. The subject walks back and forth around two traffic cones placed 10 meters apart, and the walking pace is increased by a rate of 0.17 m/sec every minute for a maximum of 12 stages [33&34]. The number of laps (“shuttles”) increases each minute until the patient is no longer able to keep up with the pace because of breathlessness or other symptoms. The primary outcome is distance walked, calculated from the number of laps completed. During the test, the technician provides coaching to encourage the patient to maintain the pace. Although both 6MWT and SWT are both categorized as “simple walking tests,” they are aimed at measuring different functional abilities. The 6MWT was proposed to evaluate disability and functional status. In contrast, the SWT was initially applied in physical education as a field test similar to progressive ergometer or treadmill exercise tests, and as such is designed to be a symptom-limited maximum incremental exercise test. The putative advantage of the SWT is that it is externally paced and therefore does not require strategy or self-pacing by the patient. This is counterbalanced by the criticism that the SWT does not reflect common daily activities that require endurance and pacing. In part to overcome this, Ravill and colleagues has developed an Endurance SWT (ESWT) which has a slower increment in pacing. Because the ESWT bases its incremental increases on a percentage of the patient's prior maximal performance, it requires a preliminary SWT to estimate maximal capacity. In this test the patient exercises to 85% of the previously determined maximum capacity on the SWT [[35]]. The ESWT is more responsive to pulmonary rehabilitation in patients with COPD than the maximal effort SWT [[35]].

Validation of SWT

The SWT has been less extensively studied in COPD than the 6MWT. Singh and colleagues found that patients with COPD have a strong correlation of SWT with VO₂Max on treadmill testing with an r-value of 0.88.but a weak correlation with FEV₁ (r-value of 0.36) [[34]]. Administration of bronchodilators to elderly patients with chronic airflow limitation shows improvement in SWT that correlates with improvement in FEV₁ [35&36]. The change in SWT following rehabilitation is significant but modest compared to the endurance version of the test.

Measurement Variability of SWT in Chronic Airflow Limitation

Minimal clinically important differences in symptoms or functional status have not been correlated with changes in the SWT as they have been for the 6MWT. There are, however, several studies that have addressed the variability of SWT in patients with COPD [[36]]. It has been suggested that the SWT should have less variability than the 6MWT because it is externally paced rather than self-paced. Dyer and colleagues performed SWT in elderly patients with chronic airflow limitation (mean FEV₁ = 1.2 L) and normal elderly subjects. The SD of the difference between the two tests was 17.9 m with a coefficient of repeatability of 35.1 m (1.96 × 17.9) This represents 20% of the mean value of 178 m, which is similar to the 6MWT.

Unanswered Questions

There is adequate information to accept the validity of the 6MWT and SWT as measures of impairment in COPD. Because both tests are also validated for patients with cardiac disease and pulmonary hypertension, they are of particular use in comparing severity of impairment across populations with chronic cardiopulmonary diseases. Although the 6MWT was designed as an endurance test to establish disability and impairment and the SWT was designed to measure maximum exercise capacity, both tests show similar reproducibility and correlation with symptoms, lung function, and formal tests of exercise capacity. Because it is more commonly used in clinical trials, more data is available to establish the responsivity of the 6MWT to interventions than the SWT. There is also more evidence available, based on self-report and statistical variability, to establish MCID for 6MWT (e.g., 54–80 m) than for SWT.

Establishment of MCID for clinical tests involves several other issues that are not addressed in this review. The mean change in a population that is of clinical importance is often much less than what is required for one to be confident than an individual patient's response is outside of the range of random variation. For example, a small change in the mean value of a population may have large relative treatment effects at the tail of the distribution. Furthermore, a small change observed within the timeframe of a clinical trial but which is sustained over time may lead to increasing treatment effects over time. It is also important to recognize that MCID may not be symmetrical in terms of the value of improvements and declines. Moreover, the importance of a change in a test may also be importantly determined by the severity of the baseline impairment, so it is important to be mindful of the degree of impairment of the population under consideration. Finally, the magnitude and confidence in assessing benefits in a patient must take into account the cost and morbidity of the proposed treatment. Those treatments that are more costly and morbid should require larger clinical benefits.