Study Type
The CONSORT (Consolidated Standards of Reporting Trials) guideline and flowchart for randomized clinical trials was followed in conducting this study.10 This is a clinical trial, randomized, prospective longitudinal, parallel, analytical, experimental, with random allocation, in which 50% of the participants were in the intervention group and the remaining 50% in the control group. The study included the participation of post-operated patients, with reconstruction of the anterior cruciate ligament. The study was submitted for evaluation by the Ethics Committee of the Federal University of São Paulo and started after approval (Project CEP/UNIFESP n: 0507/2019; CAAE: 12902219.4.0000.5505). Designed by the World Health Organization, its Universal Evaluation Number (UTN): U1111-1242-8567 in 31/10/2019 and also submitted to the Brazilian Registry of Clinical Trials with approval opinion: RBR-9bdgxh in 28/01/2020. All patients signed a written informed consent form prior to the start of the study. The recruitment of participants was conducted after approval by the ethics committee in January 2020 until May 2020.
Participants
A pilot study was conducted, on a small scale, with the same objectives, procedures, materials and methods proposed in the research. Thus, a total of 24 patients were needed, 12 per group. Therefore, to represent a dropout rate of up to 10%, a total of 28 patients were recruited, 14 in each group.
Men and women with anterior cruciate ligament rupture confirmed by medical examination in the last 2 months and planned reconstructive surgery, aged between 18 and 59 years, were included in the study. Exclusion criteria were: new injury or reconstruction of the anterior cruciate ligament due to laxity in the involved limb, patients who needed additional surgical procedures after reconstruction of the anterior cruciate ligament, or infections in the operated region. Previous surgeries on the affected knee, injury to another knee ligament or meniscus were not included. Participants were recruited from the Cohen Institute of Orthopedics, Rehabilitation and Sports Medicine and underwent training at the same location.
Randomization and Allocation
For allocation into groups, randomization was performed in blocks. Each participant had a number that was generated by computer through the website <https://www.random.org/sequences/>, where the smallest value was 1 and the largest, 28. The hidden allocation format was automatically generated through the website, divided into two columns: the first, control group and the second, intervention.
Interventions
The experimental protocol is represented in the figure below, followed by the description of each step of the experiment.
The intervention group used pressure gauges measuring 10 cm x 80 cm in width and a 7 cm x 52 cm pneumatic bag (Cuff Scientific Leg® – WCS, Curitiba, Paraná, Brazil) in the region close to the inguinal ligament of the right and left leg. The control group performed the same proposed exercises without any occlusion material.
To determine the vascular occlusion pressure a portable vascular doppler device (DV-610B®; MEDMEGA, Franca, São Paulo, Brazil) was used. The patient rested for 15 minutes in neutral supine position before the procedure. After resting, with a portable vascular Doppler pen that has an infrared signal transducer, the blood flow over the posterior tibial artery was captured with the aid of a gel. Then the cuff was inflated at the lowest pressure until the arterial pulse was no longer detected, at this time, the pressure value marked on the manometer represented the total occlusion pressure, after this the cuff was deflated.11,12 During training, 80% of the total occlusion pressure was used. The maximum load tests (1RM) were performed through maximum repetitions in the leg press and flexor chair exercises for training with BFR. 30% of a repetition maximum (RM) was used in each selected exercise in the intervention group and 70% of 1RM in the control group.
In exercises with BFR, four sets were performed, the first always of 30 repetitions, followed by three more sets of 15 repetitions each. With two seconds for concentric contraction and two seconds for eccentric contraction. The rest time in each series was 30 seconds, marked by a digital clock. The pressure must be maintained during the performance of all sets of exercise, if necessary, the pressure can be adjusted at intervals of sets, after the end of the exercise. Between each exercise block, the BFR is released for five minutes, to normalize the blood flow in that region and therefore ensure reperfusion.13 The training frequency was twice a week for 12 weeks. When the patient could not perform the full movement of one repetition due to fatigue, the exercise was stopped. The control group performed the same exercises, with 70% of 1RM, with three sets of 10 repetitions, with all other parameters, but without BFR. The adherence of the research subjects was registered by an attendance list in each session.
Outcomes
The primary outcome of this study was the assessment of muscle strength using an isometric dynamometer on the first day of assessment and repeated on the last training day of the 4th, 8th and 12th week. The maximum isometric strength of the knee extensors and flexors in both legs was measured using the MICROFET2® digital handheld dynamometer (Hoggan Health Industries, West Jordan, UTAH, USA). The maximum force was recorded in Newtons (N) and collected three times in each leg. Between each contraction there were 60-second intervals. Participants were instructed to remain seated very erect on a chair with their arms crossed in front of their chests and their legs hanging over the edge, knee and hip flexed at 90 degrees. To measure the quadriceps through isometric knee extension, the dynamometer was positioned in the anterior region of the tibia and secured with an inelastic band, in a location identical to the location used in isokinetic dynamometers (five cm proximal to the lateral malleolus), generating stabilization and resistance against movement during the test. To assess the strength of the hamstrings for isometric knee flexion, the dynamometer was attached to the sural triceps region (five cm proximal to the lateral malleolus), and to a table located in front of the participant by an inelastic strap. A demonstration was performed to familiarize the patient with the device before measurements for isometric knee extension and flexion strength began to be collected. Each participant was instructed to perform maximal isometric contractions for five seconds and the maximal strength was recorded.
The secondary outcomes were to analyze physical function of the knee using the Lysholm, IKDC (International Knee Documentation Committee) and KOOS (Knee injury and Osteoarthritis Outcome Score) questionnaires. All questionnaires were completed by the patient in order to reduce application bias on the first day of assessment and repeated on the last training day of the 4th, 8th and 12th week.
Sample Size
Sample size calculation was performed using the G*power® 3.1.9.2 Software (Heinrich-Heine-Universität, Düsseldorf, Germany), considering a statistical power of 95% and α=0.05, based on the pilot study and previous studies with distinct populations7, therefore stipulating a total of 24 patients, 12 in each group. However, a total of 28 patients were recruited, considering an abstinence rate of at least 10%.
Statistical analysis
Data normality and homogeneity were tested using the Shapiro-Wilk and Levene tests, respectively, verifying that the sample did not follow a Gaussian distribution. Descriptive data analysis was performed and expressed as mean and standard deviation. To verify possible differences in the sampling characteristics between the two groups, ANOVA with Greenhouse-Geisser adjustment was used. To analyze the effect of training with vascular occlusion, the Generalized Linear Model of Generalized Estimating Equations (GEE) statistical test was used in which the independent factor was group allocation (control or intervention) and time (week 0, week 4, week 8 and week 12), dependent factor. Effect size was described using Cohen's d as small if 0.20-0.30, medium if 0.40-0.70 and large if greater than or equal to 0.80.