Maximal Inspiratory Pressure: Test-Retest Reliability
Performance did not technically plateau, but systematic differences between Pimax measured at the third and fourth test sessions were very small; hence, data from these sessions were examined for test-retest reliability. The mean absolute difference, cumulative frequency distribution for the absolute difference, and the 95 percent confidence interval for the absolute difference indicated that Pimax fluctuated by approximately 5 cm HzO or less in most of the sample. This was consistent with other measures of strength where day-to-day fluctuations were reported to be ± 10 percent. However, a portion of the sample demonstrated wider fluctuations. Taking a conservative approach, we conclude that when monitoring individual patients at repeated intervals, Pimax must change by approximately 10 cm H20 or more before it can be interpreted as a clinical change in strength. Moreover, a repeatable trend must be established to confirm a change in strength.
A pressure transducer and strip chart recorder are frequently used in the research laboratory to reduce methodologic measurement error, but the aneroid gauge is more practical in field testing and in the clinical setting. Our data from 15 subjects suggested that experienced data collectors could acquire reliable data with an aneroid gauge. However, there are two notable limits to this technique. Maximal inspiratory pressure data must be interpreted with caution if patients have difficulty sustaining maximal pressures for 1 full second. Moreover, the aneroid gauge cannot be used reliably when patients generate wildly fluctuating pressures during each effort as this makes it difficult to identify the precise point at which the pressure was maintained for 1 full second.
In this study, the test-retest reliability for Pimax was high in part because we controlled for factors that might influence biologic variability in performance. Repeated measurements were taken at the same time of day to control for the effects of fatigue. We discarded data from all patients who experienced an upper respiratory tract infection because of potential effects on respiratory muscle strength as demonstrated by Mier-Jedrzejowicz et al. Both data collectors were experienced in conducting tests of Pimax. This was important because experienced data collectors provided clear unambiguous instructions and placed patients at ease allowing them to concentrate on generating a maximal inspiratory effort. Tests were conducted in a quiet environment to improve concentration and the individual Pimax trials were paced in a slow manner to reduce stress. It was not possible to control for subject motivation, but we tried to minimize the effects of different levels of motivation by using the same precise directions for each test.
We conclude it is necessary to control for the learning effect when measuring Pimax in naive COPD patients. To ignore this important source of systematic measurement error increases the chance of reporting an improved strength that does not exist. However, numerous practice sessions are time-consuming and more efficient methods are needed to speed up the learning process. For the group as a whole, test-retest reliability was high and day-to-day fluctuations in performance were consistent with other tests of strength.