Lower Limb Blood Flow Geko vs Foot Pump

Ethical approval was obtained from the National Research Ethics Service (Edgbaston), reference 13/WM/0139. Upon admission to hospital for THA, eligible patients were invited to participate in the study. Inclusion criteria were: patients undergoing primary THA, in good general health and fitness, age 18 years or over and the ability to provide written informed consent. Patients were taken to the ultrasound suite within the first three days following surgery.

Baseline ultrasound Patients lay in bed, the back supported at approximately 45 degrees, legs lying horizontal with their knees extended. Duplex ultrasound was then performed on the superficial femoral vein (Acuson S2000, Siemens). Baseline bilateral venous assessments of peak velocity, mean velocity, volume flow and vessel diameter were taken in triplicate. All scans were performed by the same ultrasonographer for consistency. Gate size was adjusted to just within the vessel wall.

Randomisation Following baseline measurements, patients were fitted with either the NMES device or an IPC device, the order determined by randomisation using online software. A representative from the study sponsor provided sealed envelopes containing the arm to be randomised which was opened on the day of testing.

Patients were randomised into arm A (NMES first, IPC second) or arm B (IPC first, NMES second). Blinding was not possible because the obvious physical differences between the two devices. Following testing of the first device, a 30-minute rest period was then given. The second device was then tested. All devices were fitted by the first author. All patients underwent scanning by the same sonographer to ensure consistency in data recording.

NMES application The NMES device (Geko T-1, Firstkind Ltd., High Wycombe, United Kingdom) was applied to both legs as per the manufacturer's instructions. The skin over the fibula head was prepared with an exfoliation pad, followed by an alcohol wipe. The device was applied to the skin using an adhesive strip and positioned so that the built-in electrodes were positioned over the common peroneal nerve. An overlay strip was then used to improve skin contact. A charge balanced short single pulse with a current of 27mA was delivered at a frequency of once per second. The pulse width adjustment has seven selectable stimulation settings ranging from 70 - 560 μs and was increased until a visible foot twitch was produced. Following successful application of the device for ten minutes, venous blood flow measurements were repeated.

IPC application

An IPC device (Vadoplex, OPED UK, UK) was fitted to each foot as per the manufacturer's instructions and activated for ten minutes prior to recording venous flow. The IPC provides a pump inflation pressure of 120mmHg, inflating once every 20 seconds. Each inflation was maintained for approximately 1 second, before deflation occurred. The maximum recording duration of the ultrasound scanner was 14 seconds. In order to capture 20 seconds of data, two contiguous measurements of 14 and 6 seconds were made. The machine records flow within the vessel. Since flow varies within the vessel, the machine is able to use a complex algorithm to sample velocities and flow over time and hence calculate the mean. In order to calculate mean velocity and volume flow over a complete 20 second cycle, we used the following equation:

1. Measure the blood flow during inflation (F1) and record the time duration of this flow (A).
2. Measure the blood flow during the opposite, deflated, part of the cycle (F2) and record the total cycle duration (B).

The calculation below provides a summated value for flow.

Total flow = [(A/B) x F1)] + [(1-(A/B)) x F2]

Device acceptability Patients were asked to rate comfort of the devices using a verbal rating score: 1 extremely comfortable, 2 comfortable, 3 uncomfortable, 4 very uncomfortable, 5 extremely uncomfortable.

Power analysis and statistical analysis Data from an unpublished study in healthy volunteers of the NMES device verses IPC comparing venous flow parameters, showed a significant benefit of NMES, with an effect size of 1.4 (Cohen's d). Sample size calculation based on this effect size gave a power in excess of 90% for a design of ten subjects per site.

We checked for significant association (p<0.05) of device, measurement order and leg on blood flow parameters of interest using generalized linear latent and mixed model with a random effects term to account for repeated measures on a given subject using Stata10 (Stata Corp, college station, TX). Blood flow parameters are reported as mean and 95% confidence intervals (CI). A Wilcoxon signed rank test was used to analyse the verbal rating scores for patient comfort.