Reliability and validity of a mobile home pelvic floor muscle trainer: The Elvie Trainer

Abstract

Aims Reliability and validity of force measurement and task detection by the Elvie Trainer were evaluated against an intravaginal dynamometer (IVD) and ultrasound (US) imaging. Methods Women were recruited from local physiotherapy clinics. At the first visit, pelvic floor muscle (PFM) strength and tone were assessed manually. Women performed two sets of three repetitions of rest, PFM maximal voluntary contraction (MVC), and maximal Valsalva maneuver (MVM) tasks in supine and standing, with the Elvie Trainer in situ. Women performed another set of rest and MVC repetitions with a custom IVD in situ. At the second visit, PFM strength and tone were reassessed manually. Women performed two sets of three repetitions of the rest, PFM MVC, and MVM tasks in supine and standing, with the Elvie Trainer in situ. Concurrent US imaging was then acquired during a final set of PFM MVC and MVM repetitions in supine and standing, while the Elvie Trainer remained in situ. Reliability was evaluated using intraclass correlation coefficients. Validity was evaluated using Spearman's/Pearson's correlations and receiver operator characteristic curves. Results Thirty women participated in the study. The Elvie Trainer MVC force outcomes exhibited excellent within‐day and good between‐day reliability, but were significantly lower than IVD measures, and exhibited poor relationships with IVD force outcomes. The Elvie Trainer was able to specify correct/incorrect performance of a PFM MVC. Conclusions The Elvie Trainer exhibits acceptable within‐day and between‐day reliability and can detect the correct performance of PFM MVCs; however, force measurements are not valid indicators of PFM strength and should not be used to measure outcomes.

Design and validation of an automated dual‐arm instrumented intravaginal dynamometer

“…Differences in measures of PFM function between supine and standing positions have been previously described, 32,33 thus valid comparisons between PFM MVC forces can only be made when data are acquired in the same position. Using our first-generation IVD, Czyrnyj et al 31 31 were similar to posterior IVD arm outcomes from the current study. While Czyrnyj et al 31 did not report rate of force development during active contraction, values computed retrospectively from that data set are not different from those reported in the current study (current study = 9.12 ± 6.58 N/s, Czyrnyj et al 31 =8.56 ± 6.08 N/s, p > .05).…”

“…Using our first-generation IVD, Czyrnyj et al 31 31 were similar to posterior IVD arm outcomes from the current study. While Czyrnyj et al 31 did not report rate of force development during active contraction, values computed retrospectively from that data set are not different from those reported in the current study (current study = 9.12 ± 6.58 N/s, Czyrnyj et al 31 =8.56 ± 6.08 N/s, p > .05).…”

“…In vivo data from the posterior arm of the novel IVD presented here were similar to those obtained using our first-generation IVD, which acquired force data only on the posterior arm. 25,30,31 In a group of 20 nulliparous participants, during a passive stretching task performed in supine and using the same testing parameters, Bérubé et al 25…”

“…Using our first‐generation IVD, Czyrnyj et al 31 acquired force outcomes during active PFM MVCs in a standing position in a sample of 30 female participants (age = 44.0 ± 12.8 years; BMI = 25.0 ± 4.3 kg/m 2 ; ICIQ‐FLUTS = 8.0 ± 4.1, parity = 2 1,5 ). Measures of standing baseline force (current = 8.77 ± 3.27 N, Czyrnyj et al 31 =10.23 ± 2.47 N, p > .05) and relative peak force (current = 4.04 ± 1.97 N, Czyrnyj et al 31 =5.18 ± 2.2 N, p > .05) acquired by Czyrnyj et al 31 were similar to posterior IVD arm outcomes from the current study. While Czyrnyj et al 31 did not report rate of force development during active contraction, values computed retrospectively from that data set are not different from those reported in the current study (current study = 9.12 ± 6.58 N/s, Czyrnyj et al 31 =8.56 ± 6.08 N/s, p > .05).…”

“…The system requires a computer to operate, compatible software, and has a large number of small components, making it vulnerable to damage. While this device presents higher functionality for a lower cost than previous iterations, 25,31 the benefits of this device come at a cost: the device is delicate and must be stored and handled carefully. While completely suitable to a research environment, a clinical environment would likely benefit from a device with fewer parts, less fragile components, and one that operates without a need to interface with a computer.…”