Previous studies have associated mobilisation in foot conditions other than shoes (such as slippers, sandals, socks, bare-feet and other 'non-ideal' foot conditions) with an increased risk of falling [13–15]. The poorer relative performance of non-slip socks compared to barefoot conditions (a 'non-ideal' foot condition) in our study suggest that non-slip socks do not represent an adequate alternative to well-fitting rubber soled footwear or even to mobilisation in bare feet.
Additionally, ensuring that non-slip socks are being worn appropriately (with tread pattern aligned with the sole of the foot) would most likely require periodic checks by clinical staff, especially if provided to patients with cognitive impairment. Aside from the resource implications, poorly fitted socks or socks that are mis-aligned could constitute a trip or slip hazard for patients. It is suggested that these risks might outweigh the clinical benefits of marginally improved underfoot traction over compression stockings.
All non-slip sock products tested in this study had a tread pattern on the ventral surface (sole) of the sock (Figure 2). This tread pattern is three-dimensional in nature resulting in a series of peaks (1–2 mm high) and troughs. In phase one, specimens cut from the three samples were adhered to the rubber slider using double sided tape. As the Wet Pendulum Test generates limited downward force during the 'kick' phase of the test, force at the point of contact is predominantly along the horizontal plane. As a result, the rubber slider does not press down on the troughs. Consequently, only the raised portion of the sock specimen would make contact with the floor, reducing what could be termed as the 'effective contact area' and therefore reducing slip resistance. This phenomenon is avoided in the compression stocking sample due to the absence of a tread pattern resulting in a level surface. In this case, the effective contact area would be equal to the size of the specimen adhered to the rubber slider.
When tested in-situ during phase two, the combined effect of participant weight and pliant characteristics of soft tissue of the foot are likely to force contact between the troughs of the non-slip sock and the floor thereby ensuring contact is made between the whole foot and floor across all testing conditions. This difference is proposed a plausible explanation for the apparent lack of congruence between phase one and phase two results.
Nagata, Watanabe, Inoue and Kim (2008) studied the validity of five different friction testing methods as an index of the risk of slipping with seventy subjects and concluded that of the five methods tested, the ramp test was the most reliable, and the pendulum tester the least reliable . These results appear to validate the incongruence between results of two phases of our study.
There is also a possibility that the relative performance of non-slip socks and compression stocking is altered in the presence of a fluid contaminant. This hypothesis would require further investigation and if verifiable, has potential clinical implications when using non-slip socks with older persons having issues with bladder continence. Given previous findings that slips associated with standing in urine were reduced amongst nursing home residents wearing non-slip socks, one would have expected these socks to display greater slip resistance in the Wet Pendulum Test condition. However, this was not the case.
This study tested a convenience sample of non-slip socks, however it is recognised that there may be alternative products which perform differently.
This study tested non-slip sock performance on hospital grade vinyl which is the preferred floor covering as per AS 2055.1 . However, it is possible that relative results might vary over other surfaces such as tile, polished concrete or carpet. Foot anatomy, biomechanics and skin characteristics of the relatively young and healthy participants in this study are also likely to be different to hospital patients who are older and frail. Some variation in performance across foot conditions could be expected with a sample of older hospital patients.
The testing protocol employed in phase two, although not standardised or previously validated, is very similar in method to the ramp test recommended in the Australian Standard . However, the testing protocol still provides a reliable method to compare performance of various foot conditions within the same participant.
It needs to be acknowledged that the phase two ramp test collected slippage data with participants in a static standing position. It is conceivable that slippage characteristics, and therefore performance, of these foot conditions might vary during dynamic walking on a level surface.
The small number of participants can be considered a limitation of this study. However, we would like to highlight that the unit of analysis is not the individual participant but rather the results of the test in each foot condition, which is a product of the unique characteristics of the contact material (compression stocking, non-slip sock, conventional sock or skin), the fit of the particular sock (or compression stocking sample) to the participant's feet, and the weight of the participant. Additionally, we tested subjects with both large and small feet, as well as significant difference in weight, and found the results to follow a consistent pattern across all participants.