Should ice application be replaced with neurocryostimulation for the treatment of acute lateral ankle sprains? A randomized clinical trial

Potential participants were recruited using a university community’s electronic mailing list (Université Laval, Quebec City) and at the university’s physiotherapy clinic. A physiotherapist in charge of recruitment and evaluations contacted all interested individuals for pre-screening and scheduled a first appointment at the clinic with potential participants to confirm the following criteria: 1) had experienced a moderate to high grade LAS in the 72 h preceding recruitment, 2) were aged 18 and above, and 3) were available to participate in eight physiotherapy treatments and five evaluation sessions over a 6-week period. Potential participants were excluded if they: 1) had experienced a previous LAS in the 12 months preceding recruitment, 2) had ever suffered a foot or ankle fracture, 3) had residual signs or symptoms of a previous foot or ankle injury, 4) had any lacerations, wounds or other conditions affecting skin integrity at the treatment site, 5) presented any contraindications to cryotherapy treatment (such as peripheral vascular disease, Raynaud’s syndrome, cold urticaria, or cryoglobulinemia) or altered skin sensitivity. Potential participants were also excluded if the principal injury, as diagnosed in the baseline evaluation, was not to the lateral ankle ligaments, but rather to the distal tibiofibular syndesmosis (assessed with the squeeze test, the dorsiflexion-external rotation test, and palpation) [18‐20]. For the purpose of this study, a moderate to high grade LAS diagnosis was confirmed with the presence of all of the followings: 1) an history of traumatic onset, 2) pain and limping (or incapacity) at walking, 3) visible edema on the lateral aspect of the ankle, and finally, 4) pain at palpation of the anterior talofibular ligament or the calcaneofibular ligament, and/or an augmented range of movement in at least one test among the anterior drawer, the anterior talofibular ligament stress test and the calcaneofibular ligament stress test. As there is no clear consensus on how to diagnose and classify LAS, these criteria were chosen to reflect current clinical guidelines [1, 5, 6, 21, 22], Potential participants presenting with positive Ottawa ankle rules who had not undergone radiography evaluation were sent for medical examination in order to exclude possible fractures [23]. The sectorial health sciences research ethics committee of Université Laval approved this study (#2015–053).

A single-blind, parallel group randomized clinical trial with blinded evaluator was realized (see Fig. 1. Study Design). Following the baseline evaluation, participants were randomly assigned to either the NCS (experimental: standardized rehabilitation program + NCS) or the ice (comparison: same standardized rehabilitation program + ice application) group. Both groups received standardized physiotherapy treatments, consisting of eight treatments over a four-week period (three treatments per week for two weeks, followed by one treatment per week for two weeks). They were evaluated five times in total: at baseline (T0), one week (T1), two weeks (T2), four weeks (T4) and six weeks (T6). Sociodemographic data were collected at T0. Functional capacity (the primary outcome), pain, ankle dorsiflexion ROM and edema measurements were taken at each of the five evaluation sessions.

The primary outcome was functional capacity, as measured by the Lower Extremity Functional Scale (LEFS). The LEFS is a self-administered questionnaire consisting of 20 items [24]. It has been validated in individuals with lateral ankle sprain [25], and its reliability, construct validity and responsiveness have all been demonstrated [24‐27]. The validated French Canadian version of the questionnaire was used for the current study [28, 29]. The minimal detectable change (MDC) and the minimal clinically important difference (MCID) are both 6 points [27].

Secondary dependent variables included pain intensity, ankle edema, and dorsiflexion ROM. Two visual analog scales (VAS) were filled to evaluate perceived pain intensity at rest and during usual activities in the last 48 h (10 cm scale varying from 0 [“no pain”] to 10 [“worst pain imaginable”]) [30]. The reliability of this scale has been demonstrated, and its MCID is 1.3 cm in individuals with acute pain [31]. The Figure of Eight method measures ankle size and edema using precise anatomical landmarks, and has been found to be reliable and valid in populations with ankle sprains or other injuries of the ankle or foot [32‐34]. The minimal detectable change (MDC_95%) is 0.96 cm for swollen ankles [35]. Finally, the employed dorsiflexion range of motion (ROM) measurement, the Weight-Bearing Lunge Test (WBLT), is standardized, simple, fast and reliable. The distance between the big toe and a wall is measured in a forward lunge in weight-bearing (with the knee touching the wall). Its MDC has been established at 1.6 cm (inter-evaluator) or 1.9 cm (intra-evaluator) [36].

Both groups received the same rehabilitation program. The only between-group difference was the application of ice in the comparison group and the use of NCS in the experimental group. Three physiotherapists, accustomed to working with individuals with acute ankle injuries, administered the treatments in an outpatient physiotherapy clinic on the university campus. All three physiotherapists treated participants within both allocation groups. They were properly trained in the study protocol and rehabilitation program, and they practiced the cryotherapy procedures often enough to be at ease and effective before the onset of the study. While they were asked to respect their own clinical judgment for the elaboration of their treatment plan, they were required to prescribe and teach stretching, strengthening and balance exercises and to perform joint mobilizations. They were also advised to use a compression and support modality such as a laced brace, taping or elastic bandaging. On the contrary, they were not permitted to use ultrasound, electrotherapy (such as transcutaneous electrical nerve stimulation or interferential current), heat application or dry needling modalities. This rehabilitation program follows current clinical guidelines [1, 5, 6, 37]. In-clinic sessions therefore consisted mainly of education on the injury and the healing process, teaching and practicing exercises, joint mobilization and soft tissue techniques, and management of compression and support methods. The parameters and details of each of these interventions were adjusted at the physiotherapist’s discretion according to each participant’s condition and progress.

In both groups, cryotherapy treatments were administered at the end of each physiotherapy session. For the NCS group, NCS was applied following the manufacturer’s instructions (Cryofos Medical GmbH, Germany), which is similar to application methods used in other studies [16, 17]. Slow sweeping motions of the medical gun were used over zones approximately the size of a credit card, for a maximum of 2 min, until a skin temperature of 4° Celsius was reached. Skin temperature was measured constantly by a thermometer contained within the medical gun, and the desired temperature of 4° Celsius was usually attained within 30 s of application. This procedure was applied to two distinct zones in order to cover the full surface area of the lateral ankle and was then repeated on the medial ankle and posterior to the knee, due to the presence of lymphatic ganglions at the latter site. For the ice group, the cryotherapy consisted of two bags of crushed ice (each approximately 25 × 25 cm) applied around the injured ankle in order to cover the joint entirely for 15 min, with the legs elevated. Skin temperature of the lateral treated ankle was measured before and immediately following each cryotherapy treatment in both groups, using the thermometer contained within the NCS medical gun.

The required sample size was established at 36 subjects, 18 in each group, based on a standard deviation of 12.5 points and a MCID of 6.3 points on the primary outcome (LEFS questionnaire) [26] as well as a 15% loss to follow-up, using G*Power 3.1.7 software. Alpha and beta levels were set at α = 0.05, β = 0.80, with effect sizes of 0.80.

Baseline sociodemographic data were compared between the groups using independent t-tests and chi-squared tests. Normality of the distributions was verified for each of the variables. Nonparametric Analysis for Longitudinal Data (nparLD Package 2.1, R-software, v.3.3.3) for repeated measures were used since distributions were normal at baseline (as there was a relatively wide range of scores for every single outcome) and gamma at the latest assessment times (as most participants improved close to optimal values). The nparLD is a procedure that manages a change of distribution between groups and measurement times [38]. A two-way (2-Groups [NCS, ice] × 5-Time [T0, T1, T2, T4 and T6]) nparLD was used to compare NCS and ice effects on each of the dependent variables. Intention-to-treat procedures were followed, and on the rare occasions that data were missing, the result from the previous evaluation was used (Last-Observation-carried-Forward method). The α criterion was always set at 5%.