Sensibility and measurement properties of the Tampa Scale of Kinesiophobia to measure fear of movement in children and adults in surgical settings

Abstract Purpose: Fear of movement, or kinesiophobia, is a risk factor for developing chronic post-surgical pain (CPSP) and may impede recovery. Identifying people with kinesiophobia peri-operatively is potentially valuable to intervene to optimize rehabilitation and prevent CPSP. This narrative review aims to describe and critically appraise the sensibility and measurement properties of the Tampa Scale of Kinesiophobia (TSK) in the surgical setting in both pediatric and adult populations. Material and methods: PubMed was searched for relevant articles using search terms related to the TSK and measurement properties; the search was restricted to articles published in English. COSMIN guidelines were used to rate measurement property sufficiency and study quality. Results: Four articles examined the measurement properties of the TSK-17 in the surgical setting. Included studies demonstrated sufficient internal consistency, structural validity, construct validity, but insufficient predictive validity. Study quality was variable. Although the TSK was not originally intended for the surgical setting, with minor modification, it appears sensible to use in this population. Conclusions: The TSK is a sensible tool to measure fear of movement in children and adults undergoing, or who underwent, surgery. Future studies are needed to test content validity, test-retest reliability, measurement error, and responsiveness in the surgical setting. IMPLICATIONS FOR REHABILITATION Fear of movement is a predictor of developing chronic post-surgical pain in children and adults. Rehabilitation interventions can address fear of movement in hopes to optimize surgical outcomes and prevent chronic post-surgical pain. The Tampa Scale of Kinesiophobia (TSK), with minor modification, is a sensible tool to measure fear of movement in surgical settings. There is some evidence that the TSK is reliable and valid to use with older children, adolescents, and adults who are undergoing or underwent surgery.


Introduction
Chronic pain is a common complication of surgery in pediatric [1] and adult [2] populations. The prevalence of chronic post-surgical pain (CPSP) ranges from 5% to 85% in adults [3] and 11À 54% in pediatrics [1,4]. Prevalence of CPSP varies across types of surgery (e.g., lowest prevalence in adults for cesarian section and highest for amputation) and a number of risk factors have been purported (demographic, psychosocial, type of surgery, genetic predisposition) [3,5]. CPSP can have devastating impacts on patients (e.g., increased disability [6], decreased quality of life [7]) and societies (increased health care utilization [8,9]). Fear of movement is one risk factor for developing CPSP [10,11]. The Tampa Scale of Kinesiophobia (TSK) is the only self-report tool that specifically measures fear of movement, contributing to its widespread use in multiple languages and settings [12]. Other conceptually related measures include the Fear of Pain Questionnaire and the Fear Avoidance Beliefs Questionnaire [13,14]. These measures were not chosen as both measure the concept fear of pain, which is a broader concept, and not specific to the target concept fear of movement. In the surgical setting, the TSK is a desirable tool to administer to patients peri-operatively to identify patients with fear of movement and provide tailored interventions to prevent CPSP and optimize surgical outcomes. However, it is unclear whether the TSK is suitable, reliable, or valid for use in the surgical setting.
The term "kinesiophobia" was first introduced by Kori et al. and was defined as a "fear of movement resulting from a feeling of vulnerability to painful injury or reinjury" [15]. Fear of movement has been associated with higher levels of pain and disability [16,17]. Fear of movement is a component of the fear-avoidance model that describes how acute pain can become chronic when fear of pain following a painful experience contributes to activity avoidance, which in turn contributes to disability and increased pain [16].
The TSK, developed by Miller et al., is a self-report measure with 17-items intended to assess fear of pain related to movement and physical activities [18]. Responses are rated on a 4-point Likert scale ranging from strongly agree to strongly disagree [18]. The TSK-17 was initially developed in English and was intended to be administered to adults with chronic pain [18]. The purpose of the tool is to identify people with kinesiophobia to better provide treatment to address kinesiophobia including graduated physical exercise to reduce fear of painful re-injury [15]. In the surgical setting, people with kinesiophobia may benefit from pain education, physiotherapy for early mobilization, psychological therapy, and/or specialized transitional pain clinics that can manage opioid utilization and/or provide multidisciplinary interventions [19].
Given the potential utility of the TSK in surgical settings, a critical appraisal of the tool including its development and measurement properties is warranted. This can help inform clinicians in identifying patients with fear of movement peri-operatively to provide interventions in hopes of preventing CPSP and optimizing surgical outcomes by improving function and overall quality of life. Additionally, identifying gaps in knowledge can be useful in understanding the limitations of a tool in clinical and research settings and to design future studies to test measurement properties. The objectives of this narrative review were to summarize and critically appraise the: (1) sensibility and (2) measurement properties of the TSK in the surgical setting in both pediatric and adult populations.

COSMIN
COSMIN (COnsensus-based Standards for the selection of health Measurement INstruments) is an international group of experts in the field of outcome measures who have developed several tools to aid in selection and use of outcome measures (available online at www.cosmin.nl). This paper references a number of COSMIN tools to aid in identifying relevant studies, define measurement properties, assess study quality, and assess measurement property sufficiency.

Search strategy
PubMed was searched (from database inception to January 2021 and updated in April 2022) for relevant articles including key terms related to the tool (i.e., Tampa Scale Kinesiophobia, kinesiophobia, fear of movement, TSK) and the COSMIN PubMed search filter for measurement properties [20]. This filter was developed to aid in identifying relevant studies for reviews of outcome measures and reports a sensitivity of 97% [20]. The search strategy is available as supplementary material (Supplementary File 1). The search was restricted to articles published in the English language.

Study selection
Articles were selected that used the English language version of the TSK. To address the first objective (to review the sensibility of the TSK), articles were selected that described the: (1) development, (2) feasibility or (3) content validity of the TSK. To address the second objective, articles were selected that (1) reviewed additional measurement properties of the TSK and (2) included patients in the surgical setting (pediatric or adult populations). Study selection was conducted by the first author (GM).

Data collection
Bibliographical data was collected for each relevant article including publication year, first author, study design, patient population (number, % female, age range). Additional data were collected on sensibility and measurement properties described below. Data was extracted from relevant articles by the first author (GM).

Sensibility:
Sensibility in the context of measurement tools refers to the "common sense" aspects of the tool including face and content validity [21]. Content validity refers to the degree to which the content of the tool reflects the construct (kinesiophobia), and face validity refers to the degree to which the items in the tool look at face value to adequately reflect the construct [22]. To critically appraise the sensibility of the TSK, Bombardier's Sensibility Assessment worksheet was used (Supplementary File 2) [23]. This worksheet is based on prior work published by Feinstein (1983) and Rowe and Oxman (1993) and includes 14 items divided into 4 categories: (1) purpose, population, setting, (2) content validity, (3) face validity and (4) feasibility. Data was collected pertaining to each of the 14 items from relevant studies.

Measurement properties:
Measurement properties are indicators of the quality of a selfreport tool and can be categorized into three domains: reliability, validity and responsiveness. Interpretability, although not a measurement property, is an important characteristic of measurement tools and was also included in this review. This paper used the measurement property taxonomy, ratings of study quality, and ratings for measurement property sufficiency by COSMIN [24]. Refer to Table 1 for a list of measurement properties, corresponding statistical tests, and guidelines for measurement property sufficiency adapted from the COSMIN guidelines [24].

Results
The database searches identified 317 articles; titles and abstracts were screened, and 116 full-text articles were reviewed for relevance (see Figure 1). In total, 9 articles were identified as eligible and included. Of these, 2 articles described the process of developing the TSK [15,18], 1 article described content validity [25], 1 article described feasibility [12], and 1 interpretability [26]. Table 2 describes the characteristics of the TSK. Four articles were identified that studied the measurement properties of the TSK (English language version) in the surgical setting [27][28][29][30]. Table 3 describes the characteristics of these four included studies including the type of surgery and timing of administration relative to surgery. These articles all used the original 17-item version of the tool (TSK-17). These articles measured internal consistency, structural validity, construct validity and predictive validity. Table 4 summarizes the quality of the studies and Table 5 summarizes study results.

Sensibility
The tool was originally developed with the intended purpose to discriminate between people with and without fear of movement [15]. The developers of the TSK described the intended population as people with chronic pain, presumably adults [15]. Chronic pain is a broad diagnosis encompassing many causes and bodily locations. Item #6 in the tool ("my accident has put my body at risk … ") implies this tool targets post-traumatic pain and was not necessarily intended for post-surgical pain. The developers did not define any specific setting or context of use [15].

Face and content validity:
With respect to item selection, the authors describe selecting a pool of "rationally generated items"; there are no details regarding whether experts or patients have provided input into whether the items are relevant or if the tool is comprehensive [18]. With respect to face validity, each item in the TSK-17 appears to be suitably phrased, concise, and most items are relatively easy to understand. With respect to comprehensibility, this has only been evaluated in one study with a specific patient population: adults with chronic neck pain [25]. Additionally, several studies have postulated concerns with the four reversely scored items (items 4, 8, 12 and 16) both in chronic pain and surgical settings. Future studies are needed to determine whether there is an issue with the comprehensibility of these items or the relevance of the content [27].

Feasibility:
The TSK-17 is simple to administer and requires no special tools, training, or expertise. The tool was developed in paper format but has been adapted for electronic use. Estimated completion time is short (4-8 min) [12]. The original version of the tool includes clear instructions for the respondent. The tool is easy to score, each item is scored according to the response selected on the Likert scale (strongly disagree ¼ 1 point to strongly agree ¼ 4 points). Items 4, 8, 12 and 16 are reversely scored. A total score is calculated by summing the scores for items 1-17, and therefore the range of scores is 17-68 with higher scores indicating higher degrees of kinesiophobia. Interpreting the score is cumbersome as there is no well-defined cut-off score. Roelofs et al. published normative scores for the TSK-17 and its subscales in a large cohort    (n ¼ 3082) of patients with chronic pain from Canada, Sweden, and the Netherlands [26]. As an appendix to the publication, they include tables of normative scores with z-intervals [26]. Administrators of the TSK-17 would then have to refer to this table to determine where their patient's raw score lies within a zinterval in the table, indicating how extreme the patient's score is from "normal". The age range is not reported, and it is unclear if it includes children or people with CPSP.  [28]. They reported Cronbach's alpha for each subscale, activity avoidance (AA) and somatic focus (SF) at two timepoints (preoperative and 6-weeks post-operative): TSK-AA (a ¼ 0.76 and 0.70; sufficient �0.70) and TSK-SF (a ¼ 0.11 and 0.56; insufficient <0.70) [28].

Test-retest reliability
This has not been tested in the surgical setting.

Measurement error
This has not been tested in the surgical setting.

Structural validity
Structural validity of the TSK has not been tested in adults in the surgical setting. Structural validity of the TSK was tested in the pediatric surgical setting by Rosenbloom et al. [27] and Ye et al. [28]. Rosenbloom et al. conducted an exploratory factor analysis (EFA) of the TSK-17 in children who underwent major surgery using data obtained from the TSK-17 administered pre-operatively [27]. They reported a two-factor model fit the data best, with a standardized root mean residuals (SRMR) of 0.063 (sufficient) [27]. The two factor model separated the positively and reversely scored items. Rosenbloom et al. [27] additionally did an EFA analysis of the TSK-13 in this sample and found a three-factor model (fear of injury, bodily vulnerability, and activity avoidance) with a SRMR of 0.04 (sufficient); indicating an even better fit than the twofactor model of the TSK-17. Ye et al. [28] conducted EFA in a cohort of 55 adolescents who underwent spinal fusion surgery, with the TSK-17 administered pre and post-operatively. Indices of model fit were not reported for the EFA. Additionally, confirmatory factor analysis (CFA) was performed exclusively with the TSK-AA subscale with indices of fit indicating sufficiency at both timepoints, pre-and post-surgery (CFI 0.96 and 0.94; RMSEA of 0.06 and 0.06) [28]. However, the sample size in this study (n ¼ 55) was insufficient. Rosenbloom et al. [27] tested the convergent validity of the TSK-17 and TSK-13 measured pre-operatively against a battery of outcome measures in pediatric patients undergoing major surgery. TSK-17 was significantly (p < 0.05) correlated with pain intensity on the numerical rating scale (NRS) (r ¼ 0.24; weak), pain unpleasantness on the NRS (r ¼ 0.43; moderate), avoidance subscale of  the Child Pain Anxiety Symptoms Scale (r ¼ 0.54; moderate), and the child version of Pain Catastrophizing Scale (r ¼ 0.55; moderate) [27]. TSK-17 was not significantly correlated with movement on post-operative day 2 measured using actigraphy (r ¼ À 0.08) [27]. Norte et al. [30] tested divergent validity by administering the TSK-17 6-months post-operatively in patients who underwent knee surgery (anterior cruciate ligament repair) and tested for correlation with a battery of 23 measures administered concurrently: performance-based measures (strength, hop performance), selfreport measure of physical activity (Godin Leisure Time Exercise Questionnaire), global health, and knee function (Knee Injury Outcome and Osteoarthritis Score (KOOS) and International Knee Documentation Committee form (IKDC)). Significant (p < 0.05) negative correlations of moderate strength were found with the TSK-17 and physical activity levels (GLTEQ) (r ¼ À .312), IKDC, triple hop distance (r ¼ À 0.323) and crossover hop distance (r ¼ À 0.331) [30]. Remaining measures of knee function, strength, and hop performance were only weakly correlated [30].

Construct validity
Norte et al. additionally tested discriminative validity by doing a sub-group analysis in their cohort of 77 patients who underwent knee surgery [30]. Sub-group analysis was done between patients with low and high levels of physical activity measured using the Godin Leisure Time Exercise Questionnaire. The authors provided an adequate description of characteristics of the subgroups including demographic variables, surgical variables, knee function and strength. There was a significant difference (p < 0.05) in TSK-17 scores between patients with low and high levels of physical activity, with higher scores (higher degree of kinesiophobia) in patients with low levels of physical activity in accordance with the authors hypothesis [30]. Rosenbloom et al. [28] tested the predictive validity of the TSK-17 measured pre-operatively in relation to post-operative function. Correlations with the Functional Disability Inventory (r¼.12) and PROMIS Pediatric Pain Interference Scale (r ¼ 0.22) measured 12months pre-operatively were "weak" and not statistically significant (p > 0.05) [27].

Responsiveness
This has not been tested in the surgical setting in adults or children.

Interpretability
This has not been tested in the surgical setting in adults or children.

Discussion
Fear of movement in people who are undergoing or underwent surgery can hinder recovery and put them at risk for developing CPSP [10,11]. Fear of movement can be addressed peri-operatively with interventions such as pain neuroscience education, physiotherapy, psychological therapy, and specialised transitional pain clinics [19]. The goal of intervening is to prevent CPSP and optimize surgical recovery by improving physical function and patient quality of life. The TSK is one of the only known self-report tools to measure fear of movement, although not originally intended for the surgical setting.
The original TSK is a 17-item questionnaire developed to measure fear of movement in people living with chronic pain. Since its creation in 1991, it has been translated into 10 languages, adapted to specific populations (temporomandibular joint dysfunction and fatigue) and used in a wide array of settings [12]. There are four studies that have tested the measurement properties of the TSK-17 in surgical settings (major surgery, spinal and knee surgery) including both pediatric (> 8 years) and adult populations.

Measurement properties that were unstudied in the surgical setting
A number of measurement properties of the TSK have not been studied in the surgical setting, including content validity, testretest reliability, measurement error, and responsiveness. Clinicians and researchers may reference studies of measurement properties of the TSK in similar patient populations (e.g., chronic low back pain). Of note, many studies of the measurement properties of the TSK used the Dutch language version; our discussion is limited to studies that used the English language version. Woby et al. [31] measured test-retest reliability, measurement error and responsiveness of the TSK-17 in a cohort of 111 adults (mean age 43.4 years) with chronic low back pain referred to an outpatient physiotherapy department in the UK. The authors report test-retest reliability using intraclass correlation coefficient (95% confidence interval) of 0.82 (0.72-0.88), which is considered sufficient [31]. Measurement error was reported using the standard error of measurement (95% confidence interval) as 3.16 points (2.71À 3.78) on the TSK-17 (score ranges from 17 to 68 points) [31]. Woby et al. [31] tested the responsiveness of the TSK-11 and À 17 against the Global Rating Scale-13, a widely used scale to test responsiveness of self-report measures in patients with low back pain. Receiver operating characteristic curves were produced separately for the TSK-11 and À 17, both with area under the curve (AUC) values >0.70 (sufficient) and the authors suggested a reduction in 4 points on the scale indicates an important reduction in fear of movement [31]. Although these measurement properties are untested in CPSP, these results in related populations are promising (acceptable test-retest reliability and responsiveness) and can inform the interpretation of results in CPSP (measurement error).
Although structural validity of the TSK was tested in the pediatric surgical setting, this is unstudied in adults. Cook et al. (2006) conducted a CFA of the TSK-17 and TSK-13 (English language version) in a large sample of patients with diverse types of chronic pain and diverse age group (15À 82 years). This study conducted an important analysis using CFA on 4 models that were previously published and widely referenced [32][33][34][35]. Cook et al. (2006) report a two-factor model of the TSK-13 (AA and SF) fits the data best with confirmatory factor index (CFI) 0.87 (insufficient) and root mean square of approximation (RMSEA) of 0.085 (insufficient). In congruence with Rosenbloom et al. [27], it appears that the structural validity of the TSK is stronger with the 13 item version, indicating perhaps the TSK-13 is a more appropriate version to administer.

Implications for clinicians
The TSK appears to be a sensible tool to administer to children (> 8 years) and adults in surgical settings to measure fear of movement peri-operatively. In this setting, item #6 could be revised to read "my surgery..." or "my condition has put my body at risk … " instead of "my accident … ". Attention must be paid to the timing of questionnaire administration as fear of movement may vary in the pre-operative, early post-operative, and late postoperative phases. It is not known which time point is most suitable; multiple time points of administration may be warranted as it is unclear whether pre-operative questionnaires scores correlate to post-operative scores. Clinicians must use this tool with caution, as there are several notable limitations, particularly with respect to the interpretability of the tool as there is no welldefined cut-off score to determine which patients have fear of movement and may benefit from tailored interventions to address fear of movement. Clinicians must pair test results with clinical judgement when deciding which patients may require additional interventions to address fear of movement peri-operatively. Additionally, clinicians may consider administering additional tools in parallel that measure similar concepts, including the Fear of Pain Questionnaire [13] and Fear Avoidance Beliefs Questionnaire [14]. In the pediatric surgical setting, the TSK-13 appears to be the preferable tool version with the 4 reversely scored items removed.

Implications for researchers
In research settings, caution must be taken when applying this tool in studies pertaining to patients undergoing surgery. The TSK may be used in research to describe the degree of fear of movement of participants. It is difficult to use this tool to discriminate between people with and without fear of movement, as there is no well-defined cut-off score to discriminate between groups. For researchers who wish to use this tool to measure change over time or predict a future state, extreme caution must be taken as the ability of the tool to detect change over time is unknown in this setting and preliminary studies of predictive validity in children indicate insufficiency.

Areas for future research
There are several opportunities for future research to test the measurement properties of the TSK in the surgical setting. First, measuring the content validity in both children and adults in the peri-operative setting is needed to understand fully if the tool is comprehensive, if respondents can comprehend the items, and if the items are relevant. This may best be done using qualitative interview methods (e.g., think aloud) especially to explore patient comprehension of the reversely scored items [36]. Second, studies are needed to measure test-retest reliability, measurement error, and responsiveness in the surgical setting. Third, future research is needed to determine a cut-off score to assist clinicians and researchers in interpreting test results. Future studies in surgical settings should administer the tool at multiple time points (e.g., pre-operative, 7 days post-operative, and 4-6 weeks post-operative) as the ideal time of administration is unknown and patient fear of movement may change over time in response surgery and/or intervention (e.g., changes in fear of movement in response to acute post-operative pain).

Conclusions
The TSK is a sensible tool to measure fear of movement in children (> 8 years) and adults undergoing, or who underwent, surgery. Studies to date demonstrate the TSK-17 has sufficient internal consistency, structural validity, construct validity and insufficient predictive validity, although the quality of study methods is variable. Clinicians may use the TSK to measure fear of movement and provide interventions to patients with high degrees of fear of movement in an effort to prevent or manage CPSP and optimize post-operative rehabilitation. Clinicians are limited in that it is challenging to interpret what constitutes high and low degrees of kinesiophobia based on the results of the tool, and therefore, clinical judgement must be used in parallel with test results. Clinicians and researchers may wish to administer this tool at multiple time points in relation to surgery. Future research is needed to test additional measurement properties in the surgical setting, including content validity, test-retest reliability, measurement error, and responsiveness. Clinicians and researchers should take caution if attempting to use the tool to predict future disease states or measure change over time.
Researchers who wish to test the measurement properties of the TSK can consult the COSMIN guidelines to develop rigorous study design [24].