On the usefulness of Thompson´s test with a four-point scale after Achilles tendon rupture, a pilot study Authors: Anna-Lena Wetterteg, Jacques Riad, Per Nordin Published: 2026 ABSTRACT BackgroundOutcome after an acute Achilles tendon rupture often leaves individual variations in recovery and lingering impairment. Simple and accurate methods to assess healing during early rehabilitation are lacking. In this pilot study, we developed and evaluated a graded assessment procedure based on Thompson´s test. The aim was to examine if foot movement of the injured foot can be reliably assessed with the Four-point Scale of Thompson’s test (FST).MethodsThe observed foot movement on the injured side, was graded in four categories, zero to three, in relation to the foot movement on the non-injured side. Assessments of foot movement for 24, non-surgically treated, individuals were performed at 2, 4, 8, 10 and 12 weeks after the injury. The movement was visually measured by two examiners, physiotherapists, and simultaneously measured using a marker based three-dimensional motion analysis system (3D-MA). Visual FST measurements were compared with the 3D-MA–based measurements.For each method, intra- and inter-observer agreement as well as between method agreement was evaluated using AC2 agreement measure and difference plots.ResultsThe intra- and inter observer agreement for the FST showed that that the agreement for examiner 1 was 0.94 and for examiner 2, 0.84 while the agreement between 1 and 2 was 0.83. Intra-observer agreement for the 3D-MA measurements for examiner 1 was 0.84 and for examiner 2 were 0.85 while the agreement between examiner 1 and 2 was 0.70. Agreement between methods for examiner 1 was 0.67 and for examiner 2, 0.57. Both measurement methods showed a similar increase of the average foot movement over the follow-up period. ConclusionsOur findings suggests that the inter- and intra-examiner agreements were high indicating that the scale is reliable within and between raters. The agreements between FST and 3D-MA was moderate indicating that criterion validity needs further studies against a modified marker-based 3D motion analysis.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12891-026-10011-y. Introduction A rupture of the Achilles tendon (ATR) is an acute injury to the strongest tendon in the body and often has a major impact on capability for physical activity [1, 2]. Initial treatment is either surgical or non-surgical, and the rehabilitation process aims at restoring lower extremity functions [3, 4]. The injury is common. The incidence of ATR in Sweden was 56 per 100,000 person-years for men and 17 for women in 2021 [5, 6]. An ATR often leads to chronic impairment and decreased physical activity, irrespective of treatment [7–9]. There are large individual variations in patient recovery and the reasons for this are not fully understood [10, 11]. It has been suggested that the ensuing rehabilitation during the first year after ATR could be even more important for recovery than the choice of initial treatment [12, 13]. Early rehabilitation protocols shown to be beneficial for functional rehabilitation include removable orthosis instead of casts, early weight bearing and active controlled ankle motion exercises [14–20]. Rehabilitation can be divided into phases with increasing tendon load and ankle motion. Transitions between phases have traditionally been based on time from injury, but to achieve a more individualized rehabilitation, it has been suggested that transitions should be based on both time from injury and recovery of functional milestones [4, 19]. Early in the rehabilitation process, functional assessments using bodyweight are limited due to the frailty of the tendon. During the healing process, the Achilles tendon successively tolerates more load. Eight to twelve weeks into the process, it is usually possible to assess function using the weight of the body with tests such as the heel-rise test. To enable individualized rehabilitation throughout the healing process, methods adapted to assess development are required for clinical practice [21–23]. In recent years, elongation of the Achilles tendon during the healing process has been suspected to influence outcome after an ATR, and the Achilles tendon appears to elongate regardless of treatment [24]. The main changes of the tendon length seem to occur during the three first months after the injury, and minimizing tendon elongation could therefore be a treatment goal to restore lower extremity function [25, 26]. Besides ultrasonographic measurement, other possible assessment methods in the early healing process to assess tendon length could be the static metrics Achilles Tendon Resting Angle (ATRA) or Achilles Tendon Length Measure (ATLM) [27–30]. These could possibly provide indirect measurements of the length of the Achilles tendon and have been used in studies to evaluate tendon length during the early period after ATR [31]. Assessments with ATRA and ATLM require equipment such as goniometers and tape measures. To the best of our knowledge, there are no dynamic methods to assess foot movement during the early period after ATR which are easy to use and are designed for clinical practice. A well-known test used during this early period to diagnose ATR is Thompson´s test, a squeeze test. In Thompson’s test, movement in predominantly plantar flexion is assessed by compressing the calf muscle and comparing the amount of foot movement of the injured side to the non-injured side. An absent or clearly decreased plantar flexion on the injured side indicates a positive test [32–34]. Therefore, compression of a calf muscle with a non-injured Achilles tendon should generate an unaffected movement of the foot within the ankle joints range of motion [35]. In our clinical practice, the dichotomous Thompson´s test is used first for diagnostics and then repeatedly during rehabilitation after ATR in order to examine function of the injured side. The current form of the Thompson’s test is dichotomous; however, grading movement in relation to the non-injured side during early stages of rehabilitation may prove beneficial. A graded assessment procedure was designed where the steps in the scale were constructed to represent the observed degree of foot movement, the Four-point scale of Thompson´s test (FST). The four categories were defined as no movement to very small movement, a movement less than half of the reference foot, more than half of the movement of the reference foot, and nearly full to full movement, The overall aim of the study was to examine if the observed degree of foot movement of the injured foot can be reliably assessed using the FST assessment procedure during early rehabilitation after ATR. Materials and methods Participants Participants were identified and provided with written and oral information about the study at the Emergency Department at Skaraborg Hospital Skövde. Recruitment was conducted during a 15-month period. Participants were recruited within the first few days after their ATR. Inclusion criteria were ages 18–75 years, a complete ATR diagnosed by a palpable gap in the tendon, a positive Thompson´s test and non-surgical treatment. Exclusion criteria were previous ATR and other injury or disease that affected physical function of the lower extremity. The hospital’s treatment protocol for ATR recommends non-surgical treatment, except for cases of late diagnosed ATR and for active athletes where surgical treatment was recommended instead. Thus, we elected to exclude eight surgically treated individuals from the study, including one active athlete and seven individuals with late diagnosis. Eighty-two individuals were diagnosed with ATR during the recruitment period, 54 met the inclusion criteria and of these, 30 consented to participate. Six of these participants did not appear for their follow-up visits, twenty-four individuals, completed the study. Study design and clinical procedure This pilot study was a prospective longitudinal cohort study where participants’ healing process was followed for 12 weeks. Participants were examined at regular follow-up visits at 2, 4, 8, 10 and 12 weeks after the injury, the hospital’s routine follow-up after ATR. The passive foot movement was examined using FST and simultaneously captured with a marker based three-dimensional motion analysis system (3D-MA) as a reference method. When examining foot movement on the injured side, the movement on the non-injured side was used as reference. The two examiners, both physiotherapists performed FST independently. The examiners received no previous training of FST beside their clinical experience in performing Thompson´s test. Examiner one had worked as a physiotherapist for 25 years while examiner two had worked less than a year. The examiners were not present in the examination room at the same time, and they were blinded for each other’s assessment protocols where outputs from the FST assessment were documented. The output for 3D-MA needs calculation and was therefore blinded for the examiners. However, the examiners were aware of which week, and which follow-up visit the participant attended and could see their own scores from the previous visit in the protocol. At each visit, each participant was examined by Examiner 1 followed by Examiner 2. The hospital’s treatment protocol was followed. First, the foot was placed in plantar flexion using a foot and ankle orthosis with heel-lifts (Aircast 01P DJO Global). After two weeks, successive weight-bearing in the orthosis and early motion exercises began. The heel-lifts were removed gradually over eight weeks as the plantar flexion angle decreased down to neutral. At eight weeks, the orthosis was replaced by ordinary shoes with a 5 mm heel-lift which was used for another four weeks. The examiners performed 93 FST assessments, but due to technical and patient administrative problems, only 66 of these were captured by the digital camera system. Compression one and two, which were registered with both 3D-MA and FST during the same examination, were used to analyse the agreement within examiners. To achieve consistency, the second compression was used for analysis of agreement between examiners. Four-point Scale of Thompsons´s test (FST) During the FST examination, participants lay prone with their feet over the end of the examination table, as described by T.C Thompson [33, 34]. The FST consisted of two consecutive compressions on each calf using both hands which were lifted off the skin between compressions (Fig. 1). The compressions started with the non-injured side and were then repeated on the injured side. The passive foot movement generated from the calf muscle compression of the injured side was compared with the corresponding of the non-injured reference side and translated into an ordinal outcome with values ranging from zero to three. From a clinical view, grading for the injured side can be described as follows: no movement to very small movement categorized FST = 0, small movement less than half of the amount of the non-injured side categorized FST = 1, moderate movement more than half of the amount of the non-injured side categorized FST = 2, and the same or almost the same movement as on the non-injured side categorized FST = 3 (Table 1). Fig. 1Patient lying prone, examiner compressing the calf muscle and estimating foot movement and estimating the passive foot movement, with FST while 3D-MA cameras simultaneously captured foot movement Patient lying prone, examiner compressing the calf muscle and estimating foot movement and estimating the passive foot movement, with FST while 3D-MA cameras simultaneously captured foot movement Table 1Definition of the FSTb classificationDescription of foot movementClassification FSTbNo movement to very small movement0Small movement less than half of the amount of thenon-injured side1Moderate movement more than half of the amount of the non-injured side2The same or almost the same movement as on the non-injured side3b. FST: Four-point scale of Thompson´s test Table 1Definition of the FSTb classificationDescription of foot movementClassification FSTbNo movement to very small movement0Small movement less than half of the amount of thenon-injured side1Moderate movement more than half of the amount of the non-injured side2The same or almost the same movement as on the non-injured side3b. FST: Four-point scale of Thompson´s test Definition of the FSTb classification Small movement less than half of the amount of the non-injured side b. FST: Four-point scale of Thompson´s test Marker based three-dimensional motion analysis (3D-MA) Foot movement, plantar flexion in relation to the shank, was registered by a 3D-MA incorporating 41 retroreflective markers and 12 digital camera system (Oqus 400 Qualisys Medical AB, Gothenburg, Sweden). The markers were secured to specific anatomical landmarks in accordance with a modified Helen-Heyes Model [36]. Additional markers, three on the calcaneus and one on the distal metatarsals, 1, 2 and 5 respectively, were applied to achieve a foot model for three-plane assessment of the foot, which was part of another study (Fig. 1). Since participants were under ATR treatment, the static trial was modified to keep the foot in relative plantar flexion. This was ensured by having participants semi-standing, leaning back against the examination table with their feet positioned in plantar flexion. The movement variables for ankle movement were registered and expressed in degrees of plantar flexion of the ankle joint, thus in the sagittal plane. The outcome in degrees was calculated as a ratio, where the non-injured side was reference for the injured side, generating values between zero and one. This ratio constitutes the measure of plantar flexion movement used in subsequent analyses. A value approaching one implies that movement on the injured side is close to movement of the non-injured side. To enable comparisons between methods we created a categorical variable based on the continuous 3D-MA ratios. We used three cut-off points for the foot movement in the 3D-MA ratio scale. These cut-off points were defined a-priori corresponding to the rationale for the FST classification. The thresholds were selected to 15%, 50%, and 85% of full movement allowing for an interpretation in accordance with the FST values (Table 2). The choice of 15% in each end of the interval was based on the observation that passive movement induced by compression has a low range of movement up to around 10 degrees in our subjects on the non-injured side as measured with 3D-MA (Table 2). Table 2Categorizationa of 3D-MAb values3D-MAa values (ratio)3D-MA Interpretation3D-MA codes0.00-0.15Less than 15% of full movement00.16–0.50More than 15% but less than 50% of full movement10.51–0.85More than 50% but less than 85% of full movement20.86-1.00More than 85% of full movement and up to 100%3a. Categorization constructed to represent the same value as in the FST variableb. 3D-MA: Three-dimensional motion-analysis Table 2Categorizationa of 3D-MAb values3D-MAa values (ratio)3D-MA Interpretation3D-MA codes0.00-0.15Less than 15% of full movement00.16–0.50More than 15% but less than 50% of full movement10.51–0.85More than 50% but less than 85% of full movement20.86-1.00More than 85% of full movement and up to 100%3a. Categorization constructed to represent the same value as in the FST variableb. 3D-MA: Three-dimensional motion-analysis Categorizationa of 3D-MAb values a. Categorization constructed to represent the same value as in the FST variable b. 3D-MA: Three-dimensional motion-analysis Statistical methods Descriptive statistics include means, standard deviations, proportions, frequencies, minimums and maximums. All confidence intervals have a 95% confidence level. Agreement of FST and 3D-MA outcomes within and between examiners, as well as between methods, were analysed with agreement statistic AC2 in the vein of Kappa statistic. When the examiners did not agree, custom weights were used (details in Appendix 1). Continuous 3D-MA data were analysed for intra- and inter examiner agreement using difference plots based on means and differences of the foot movement ratios [37]. In the difference plots, values on the x-axis represent our best estimate of the plantar flexion movement ratios as defined in Marker based three-dimensional motion analysis (3D-MA). The values on the y-axis represent the difference between the plantar flexion movement ratios. Further information in Appendix 2. Means per week 2, 4, 8, 10 and 12 of all measurements for both methods, FST and categorized 3D-MA, are depicted in a line plot showing the general development over time (Fig. 5). Results The study group consisted of 20 males and four females, with mean age 44 years ranging from 24 to 71 years, The agreement within and between examiners for FST and categorized 3D-MA showed for the FST assessments within Examiner 1, 0.94 [0.89; 1.00] and for the agreement within Examiner 2, 0.84 [0.75; 0.93]. The agreement between examiners was 0.83 [0.76; 0.91]. (Table 3). The corresponding agreement for the 3D-MA assessments within examiners showed for Examiner 1, 0.84 [0.77; 0.92] and for Examiner 2, 0.85 [0.78; 0.92] (Table 3). The 3D-MA agreement between examiners showed a somewhat lower agreement, 0.70 [0.61; 0.80] (Table 3). Agreement between methods, FST and categorized 3D-MA, within one examiner showed, 0.67 for Examiner 1 and 0.57 for Examiner 2 (Table 3). Table 3Agreement for measurements analysed with Kappa statistic AC2 and 95% confidence intervals. The FSTb agreement analysis comprise 93 measurements based on 24 individuals and 3D-MAa analysis comprise 66 measurements based on 24 individualsMeasurementsAgreement AC2 [Cl 95%] cFSTb Within Examiner 10.94 [0.89; 1.00]FST Within Examiner 20.84 [0.75; 0.93]FST Between Examiner 1 and 20.83 [0.76; 0.91]3D-MAa Within Examiner 10.84 [0.77; 0.92]3D-MA Within Examiner 20.85 [0.78; 0.92]3D-MA Between Examiner 1 and 20.70 [0.61; 0.80]FST and 3D-MA Between methods Examiner 10.67 [0.55; 0.78]FST and 3D-MA Between methods Examiner 20.57 [0.44; 0.70]a. 3D-MA: Three-dimensional motion-analysisb. FST: Four-point scale of Thompson´s testc. CI 95%: 95% confidence interval Table 3Agreement for measurements analysed with Kappa statistic AC2 and 95% confidence intervals. The FSTb agreement analysis comprise 93 measurements based on 24 individuals and 3D-MAa analysis comprise 66 measurements based on 24 individualsMeasurementsAgreement AC2 [Cl 95%] cFSTb Within Examiner 10.94 [0.89; 1.00]FST Within Examiner 20.84 [0.75; 0.93]FST Between Examiner 1 and 20.83 [0.76; 0.91]3D-MAa Within Examiner 10.84 [0.77; 0.92]3D-MA Within Examiner 20.85 [0.78; 0.92]3D-MA Between Examiner 1 and 20.70 [0.61; 0.80]FST and 3D-MA Between methods Examiner 10.67 [0.55; 0.78]FST and 3D-MA Between methods Examiner 20.57 [0.44; 0.70]a. 3D-MA: Three-dimensional motion-analysisb. FST: Four-point scale of Thompson´s testc. CI 95%: 95% confidence interval Agreement for measurements analysed with Kappa statistic AC2 and 95% confidence intervals. The FSTb agreement analysis comprise 93 measurements based on 24 individuals and 3D-MAa analysis comprise 66 measurements based on 24 individuals a. 3D-MA: Three-dimensional motion-analysis b. FST: Four-point scale of Thompson´s test c. CI 95%: 95% confidence interval Agreement for 3D-MA within and between examiners was analysed with difference plots. The plot for 3D-MA within Examiner 1 showed a mean difference of -0.02, upper limit (UL) 0.23 and lower limit (LL) -0.28 (Fig. 2; Table 4). The difference plot for within Examiner 2 showed similar results (Fig. 3; Table 4). The mean difference was close to zero and the limits of agreement accounted for about a quarter of the possible variation. The differences between the measurements exhibited a narrow and homogenous variation along the mean difference for both examiners. Fig. 23D-MAa difference plot illustrating agreement within Examiner 1, with mean difference line, LLb, ULc. All estimates are given with 95% confidence limits. The 3D-MA analysis comprises 66 measurements based on 24 individuals. a. 3D-MA: Three-dimensional motion-analysis b. LL; Lower limit of agreement, mean difference – 1.96 x standard deviations c. UL: Upper limit of agreement, mean difference + 1.96 x standard deviations 3D-MAa difference plot illustrating agreement within Examiner 1, with mean difference line, LLb, ULc. All estimates are given with 95% confidence limits. The 3D-MA analysis comprises 66 measurements based on 24 individuals. a. 3D-MA: Three-dimensional motion-analysis b. LL; Lower limit of agreement, mean difference – 1.96 x standard deviations c. UL: Upper limit of agreement, mean difference + 1.96 x standard deviations Fig. 33D-MAa difference plot, illustrating agreement within Examiner 2, with mean difference line, LLb, ULc. All lines given with 95% confidence limits. The 3D-MA analysis comprise 66 measurements based on 24 individuals. a. 3D-MA: Three-dimensional motion-analysisb. LL; Lower limit of agreement, mean difference – 1.96 x standard deviationsc. UL: Upper limit of agreement, mean difference + 1.96 x standard deviations 3D-MAa difference plot, illustrating agreement within Examiner 2, with mean difference line, LLb, ULc. All lines given with 95% confidence limits. The 3D-MA analysis comprise 66 measurements based on 24 individuals. a. 3D-MA: Three-dimensional motion-analysisb. LL; Lower limit of agreement, mean difference – 1.96 x standard deviationsc. UL: Upper limit of agreement, mean difference + 1.96 x standard deviations Agreement between examiners for 3D-MA analysed with a difference plot showed a mean difference of -0.04, UL 0.32 and LL -0.40 (Fig. 4; Table 4). The mean difference is close to zero and the limits of agreement account for a little more than a third of the possible variation. However, the difference plot showed an increasing variation along the mean difference, indicating a slightly lesser agreement between examiners as the foot movement in plantar flexion increased. Fig. 43D-MAa difference plot, illustrating agreement between Examiner 1 and 2, with mean difference line, LLb, ULc. All lines given with 95% confidence limits. The 3D-MA analysis comprise 66 measurements based on 24 individuals. a. 3D-MA: Three-dimensional motion-analysisb. LL; Lower limit of agreement, mean difference – 1.96 x standard deviationsc. UL: Upper limit of agreement, mean difference + 1.96 x standard deviations 3D-MAa difference plot, illustrating agreement between Examiner 1 and 2, with mean difference line, LLb, ULc. All lines given with 95% confidence limits. The 3D-MA analysis comprise 66 measurements based on 24 individuals. a. 3D-MA: Three-dimensional motion-analysisb. LL; Lower limit of agreement, mean difference – 1.96 x standard deviationsc. UL: Upper limit of agreement, mean difference + 1.96 x standard deviations Table 4Agreement for measurements analysed with difference plots showing mean difference and upper and lower limits of agreement. All estimates given with 95% confidence intervals. The 3D-MAa analysis comprise 66 measurements based on 24 individualsMeasurementAgreement 3D-MAa / Difference plotsMean difference Limits of agreement3D-MA Within Examiner 1-0.02 [-0.05; -0.01]ULc 0.23 [0.23;0.24]LLb -0.28 [-0.28; -0.27]3D-MA Within Examiner 20.03 [0.01; 0.06]UL 0.23 [0.19;0.27]LL -0.16 [-0.20; -0.12]3D-MA Between Examiner 1 and 2-0.04 [-0.09; -0.01]UL 0.32 [0.25;0.40]LL -0.40 [-0.48; -0.32]a. 3D-MA: Three-dimensional motion-analysisb. LL; Lower limit of agreement, mean difference – 1.96 x standard deviationsc. UL: Upper limit of agreement, mean difference + 1.96 x standard deviations Table 4Agreement for measurements analysed with difference plots showing mean difference and upper and lower limits of agreement. All estimates given with 95% confidence intervals. The 3D-MAa analysis comprise 66 measurements based on 24 individualsMeasurementAgreement 3D-MAa / Difference plotsMean difference Limits of agreement3D-MA Within Examiner 1-0.02 [-0.05; -0.01]ULc 0.23 [0.23;0.24]LLb -0.28 [-0.28; -0.27]3D-MA Within Examiner 20.03 [0.01; 0.06]UL 0.23 [0.19;0.27]LL -0.16 [-0.20; -0.12]3D-MA Between Examiner 1 and 2-0.04 [-0.09; -0.01]UL 0.32 [0.25;0.40]LL -0.40 [-0.48; -0.32]a. 3D-MA: Three-dimensional motion-analysisb. LL; Lower limit of agreement, mean difference – 1.96 x standard deviationsc. UL: Upper limit of agreement, mean difference + 1.96 x standard deviations Agreement for measurements analysed with difference plots showing mean difference and upper and lower limits of agreement. All estimates given with 95% confidence intervals. The 3D-MAa analysis comprise 66 measurements based on 24 individuals ULc 0.23 [0.23;0.24] LLb -0.28 [-0.28; -0.27] UL 0.23 [0.19;0.27] LL -0.16 [-0.20; -0.12] UL 0.32 [0.25;0.40] LL -0.40 [-0.48; -0.32] a. 3D-MA: Three-dimensional motion-analysis b. LL; Lower limit of agreement, mean difference – 1.96 x standard deviations c. UL: Upper limit of agreement, mean difference + 1.96 x standard deviations Both methods, FST and categorized 3D-MA, exhibited an increase of average foot movement during the follow-up period, regardless of examiner (Fig. 5). Fig. 5Line chart of means per week, FSTb and categorised 3D-MAa scores for week 2, 4, 8, 10 and 12 of follow-up. The values indicate the direction of the general development of the foot movement over time. a. 3D-MA: Three-dimensional motion-analysisb. FST: Four-point scale of Thompson´s test Line chart of means per week, FSTb and categorised 3D-MAa scores for week 2, 4, 8, 10 and 12 of follow-up. The values indicate the direction of the general development of the foot movement over time. a. 3D-MA: Three-dimensional motion-analysisb. FST: Four-point scale of Thompson´s test Discussion This agreement study examined if eventual movement of the injured foot compared to movement of the non-injured foot could reliably be assessed with FST during early rehabilitation after ATR using a 3D-MA motion analysis system as a comparison. FST is a new measurement procedure based on dynamic foot movement. We wanted to test this measurement method in this pilot study, since a reliable method to assess healing throughout the first 12 weeks of rehabilitation was needed in order to individualize rehabilitation after ATR. The current study evaluated FST by examining the agreement in various ways, including comparing it to a 3D-MA system which simultaneously capture the movement of the foot. The first step in the agreement analyses was to find the agreement of FST and 3D-MA, both within and between individual examiners. The second step was to find out if the judgements in the performed FST procedure agreed when registering the same movement using the 3D-MA system. Finally, we looked at the average development of feet movement per examination week to allow for a visual inspection of the trends for FST and categorised 3D-MA per examiner. It was suspected that repeated compressions at each visit could affect the outcome. To examine this, four compressions were performed by each examiner, and these data showed no substantial differences with these added compressions. Furthermore, a reduced activity after injury may have generated changes in muscle mass and tendon elasticity, which in turn could influence the movement of the foot in the compression [38]. High values of agreement were found intra examiner for both FST and 3D-MA, suggesting repeatability. Inter examiner agreement were high for FST and somewhat lower for 3D-MA which could have been influenced by how the camera system was set up. The difference plot for 3D-MA comparing examiners gave further insight, as the distribution of the values along the mean difference showed increasing variation as the ratios approached one. This strengthened our suspicion that variation could be related to the camera system. We observed that in the 3D-MA data, the ratio sometimes exceeded the value one, meaning that the 3D-MA measurements suggested the injured foot showed a larger movement than the non-injured reference foot, which was not plausible. Analysing agreement between the two methods, results were considered to be moderate. These lower agreements were possibly due to the increased variation observed in 3D-MA. Yet, both FST and 3D-MA showed increased average values and similar development of movement improvement in the injured foot during the twelve-week follow-up, which corroborates with our principal findings. The alternative measurement methods ATRA and ATLM are static metrics, while FST is dynamic, in the sense that it is a passive movement, as a response to calf muscle compression. Therefore, by using dynamic foot movement in the ankle joint, the FST could provide a method of assessing rehabilitation progress after ATR while minimising the risk of re-injury of the healing tendon [27–29]. 3D-MA is widely used in clinical and research purpose, nonetheless there are challenges. In our study, a standardized placement protocol was used [39], nevertheless variation in the difference plots became apparent in this study. One possible explanation could be the variation in marker placement. Another explanation could be placement of the markers on the relatively small foot and ankle area which challenge the camera registration causing errors in the calculation of the joint kinematics [40]. Furthermore, in our study participants were examined in a prone position, where the camera system in the gait analysis laboratory captured the movement which to our knowledge have not been attempted before and the camera system is primarily designed to capture gait. The methods used to measure foot movement in this study generated both continuous and categorical data, and therefore, different approaches in the statistical processing were used. The agreement statistic AC2 was chosen to analyse categorical data since this allows for custom weights and also handles the correction of chance agreement [41, 42]. The weights were chosen with the aim to emphasise the most important differences with respects to examiner performance focusing on disagreements in the lower end of the scale. During the healing process the examiners obviously do not agree in every instance. Furthermore, disagreements were considered more serious when examiners differed with two scale steps or more in the FST scale. As a consequence of this the AC2 values will be somewhat inflated, since not every case of disagreement will influence the Kappa value in the same way. To enable method comparisons, we categorized 3D-MA ratios as close as possible to the definition used when the values for FST were selected. The continuous 3D-MA data were also analysed using difference plots, providing information about agreement and the variation of the measurement. The new FST and the reference method showed a high agreement within examiners which suggests repeatability. An obvious limitation of this pilot study is the relatively small sample size where one is the relatively small sample size where a total of 93 foot-movement investigations were performed. In 27 of these instances the 3D-MA system didn’t capture the movements which meant that any analyses involving 3D-MA had to settle with 66 observations. We examined these 27 subjects and could not find any systematic differences compared to the 66 remaining subjects. However, the absence of a systematic error cannot be guaranteed. Another limitation is the blinding procedure which was challenging since it was difficult to blind both participants and involved personnel. It is unclear whether it is possible to perform a complete blinding in a study like this one. A further limitation is that the outcomes from the 3D-MA were not consistent enough to be regarded as a gold standard. This was reflected in our statistical approach to the difference plots, where we used the average of the methods as the best estimate of the true values [37]. Yet we saw a value of reporting our findings and problems since this could provide guidance for future studies. One final limitation is that our pilot study only involved two examiners. The two examiners showed a high level of agreement despite that they differed substantially in clinical experience. However, to examine this in more detail future studies should include a larger number of examiners. Conclusions Our findings suggests: That the inter- and intra-examiner agreements were high indicating that the FST scale is reliable within and between raters.That the agreement between FST and 3D-MA was moderate indicating that criterion validity needs further studies against a modified marker-based 3D motion analysis. That the inter- and intra-examiner agreements were high indicating that the FST scale is reliable within and between raters. That the inter- and intra-examiner agreements were high indicating that the FST scale is reliable within and between raters. That the agreement between FST and 3D-MA was moderate indicating that criterion validity needs further studies against a modified marker-based 3D motion analysis. That the agreement between FST and 3D-MA was moderate indicating that criterion validity needs further studies against a modified marker-based 3D motion analysis. This agreement study is a pilot-study and therefore the results should be interpreted with some caution. However, since FST is based on the established Thompson´s test we believe that these findings suggests that FST could be a promising assessment method in early rehabilitation after ATR. Supplementary Information Supplementary Material 1: Appendix 1 AC2 [41, 42]. Application of custom weights. Appendix Table 1, custom weights AC2. Appendix 2 Difference plots [37]. Appendix 3 Sample size justification [41, 42, 43]. Supplementary Material 1: Appendix 1 AC2 [41, 42]. Application of custom weights. Appendix Table 1, custom weights AC2. Appendix 2 Difference plots [37]. Appendix 3 Sample size justification [41, 42, 43].