If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
To examine the intra- and interrater reliability of the Modified Tardieu Scale (MTS) for lower limb assessment of adults with chronic neurologic injuries.
Design
Single-center intra- and interrater reliability study.
Setting
Outpatient neurorehabilitation unit.
Participants
Adults (N=30; mean age ± SD, 54.1±12.5y) with various chronic neurologic injuries and lower limb spasticity.
Interventions
Two experienced physiotherapists performed slow (R2) and fast (R1) passive movements for lower limb muscles half an hour apart on the same day (interrater reliability), while a third physiotherapist took goniometric measurements only. One physiotherapist repeated the assessment 1 to 3 days earlier or later (intrarater reliability). Assessors qualitatively rated the resistance to fast passive movements.
Main Outcome Measures
Intraclass correlation coefficients (ICCs) and limits of agreement (LOA) were calculated for R1, R2, and R2–R1. Kappa coefficients were calculated for tibialis range of movement and qualitative spasticity ratings.
Results
Intra- and interrater R1 and R2 measurements showed moderate to high reliability for the affected hamstrings, rectus femoris, gastrocnemius, soleus (mean ICC ± SD, .79±.08), and tibialis anterior (mean κ ± SD, .58±.10). Only intrarater measurements of the affected tibialis posterior were moderately reliable (R1=.57, R2=.77). Seven of 16 spasticity angle measurements of the affected muscles were moderately reliable. LOA were mostly unacceptably wide. Qualitative spasticity ratings were moderately reliable for affected hamstrings, gastrocnemius, and tibialis muscles (mean κ ± SD, .52±.10).
Conclusions
The MTS is reliable for assessing spasticity in most lower limb muscles of adults with chronic neurologic injuries. Repeated MTS measurements of spasticity are best based on R1 measurements rather than spasticity angle or qualitative ratings of spasticity. Optimally, MTS measurements should be undertaken by the same clinician.
Individuals affected by spasticity have a greater degree of motor impairment, more activity limitation, and a lower quality of life compared with those who are not affected by spasticity.
Feasibility, test-retest reliability, and interrater reliability of the Modified Ashworth Scale and Modified Tardieu Scale in persons with profound intellectual and multiple disabilities.
The Tardieu Scale includes quantitative joint angle measures taken at 3 speeds of passive movements: as fast as possible, at the speed of a limb falling under the force of gravity, and as slow as possible.
In the clinical setup, it may be difficult to standardize the speed of a limb falling under the force of gravity. Hence, a clinically useful variation of the Tardieu Scale is the Modified Tardieu Scale (MTS),
where passive movements are tested at 2 speeds, as slow and as fast as possible. Clinicians compare the angle of full slow passive range of movement (R2) and the angle of resistance first felt during fast passive movements (R1) to identify the physiological and neurologic mechanisms affecting resistance to passive movements, respectively. For this reason, the MTS is designed specifically to identify an increase in muscle resistance caused by spasticity.
: 0 - No resistance throughout the course of the passive movement; 1 - Slight resistance throughout the course of passive movement; no clear “catch” at a precise angle; 2 - Clear catch at a precise angle, interrupting the passive movement, followed by release; 3 - Fatigable clonus (<10s when maintaining the pressure) appearing at a precise angle; 4 - Unfatigable clonus (>10s when maintaining the pressure) at a precise angle; 5 - Joint immovable
Feasibility, test-retest reliability, and interrater reliability of the Modified Ashworth Scale and Modified Tardieu Scale in persons with profound intellectual and multiple disabilities.
Feasibility, test-retest reliability, and interrater reliability of the Modified Ashworth Scale and Modified Tardieu Scale in persons with profound intellectual and multiple disabilities.
Clinical assessment of ankle plantarflexor spasticity in adult patients after stroke: inter- and intra-rater reliability of the Modified Tardieu Scale.
studied the hip and knee flexors and extensors and the ankle plantar flexors and found moderate to high intrarater reliability and poor to moderate interrater reliability. Singh et al
Clinical assessment of ankle plantarflexor spasticity in adult patients after stroke: inter- and intra-rater reliability of the Modified Tardieu Scale.
Feasibility, test-retest reliability, and interrater reliability of the Modified Ashworth Scale and Modified Tardieu Scale in persons with profound intellectual and multiple disabilities.
studied a single unspecified knee muscle and found moderate to high intra- and interrater reliability. Despite clinically observed spasticity, the reliability of the MTS has not been studied for the hip adductors in adults or for the tibialis anterior and posterior muscles in either adults or children.
Previous study designs varied considerably with respect to standardization of the speed of movements, testing positions, angle measurement methods, and blinding. Furthermore, studied populations were restricted to severely brain-injured adults with impaired consciousness,
Feasibility, test-retest reliability, and interrater reliability of the Modified Ashworth Scale and Modified Tardieu Scale in persons with profound intellectual and multiple disabilities.
Clinical assessment of ankle plantarflexor spasticity in adult patients after stroke: inter- and intra-rater reliability of the Modified Tardieu Scale.
We aimed to examine the intra- and interrater reliability of the MTS across a range of lower limb muscles, in adults with heterogeneous chronic neurologic conditions.
Methods
A convenience sample of 32 adults was recruited from the outpatient neurorehabilitation unit at Caulfield Hospital. Participants were referred to the study by their treating physiotherapist. Inclusion criteria were as follows: (1) adults 18 years or older; (2) chronic (≥6mo) and stable brain or spinal cord injury; (3) increased lower limb muscle tone as reported by their treating physiotherapist; (4) English speaking and cognitively able to give consent and participate in the study; and (5) living within 15km of the hospital because of the need for frequent and nonremunerated hospital attendance. The study was approved by the Alfred Health Ethics Committee and adhered to the Declaration of Helsinki standards for ethical research.
The assessors were 2 physiotherapists with at least 7 years' experience treating neurologically impaired adults and using the MTS. Experienced physiotherapists were chosen because they perform the MTS in the chosen setting. To ensure optimal standardization of the MTS assessment, a detailed protocol was followed (see table 1), and assessors undertook two 1-hour training sessions.
In 1 session, participants were examined by N.F. and C.B. (interrater reliability), with a 30-minute break between assessments to minimize the effect of one assessment on the other. The other session took place 1 to 3 days earlier or later, and participants were examined by N.F. only (intrarater reliability). Participants' and assessors' availability determined randomization in a Latin squares design. Testing times and conditions, however, were consistent between assessment sessions to minimize daily variability in spasticity.
A third experienced physiotherapist performed and recorded goniometric measurements (see goniometer alignment details in table 1). The assessors were blind to goniometer measurements. Each assessment was performed independently, and the 3 therapists were not permitted to observe previous assessment forms.
In supine, the hip adductors, hamstrings, gastrocnemius, and tibialis anterior were tested first on the right leg and then the left. Then in prone, the rectus femoris, soleus, and tibialis posterior were tested on the right side followed by the left. Finally, participants sat on the edge of the bed for testing of the right and then the left quadriceps muscles.
The testing protocol conformed to the principles of the MTS testing. Assessors tested muscles from their shortest to longest lengths.
Spasticity angles were calculated during data analysis as R2–R1. Assessors used a standard 0 to 5 scale (see table 1) for qualitative rating of muscle resistance to fast movements.
The testing protocol varied from other protocols in a few respects. First, goniometer positions were standardized to bony landmarks (see table 1), and angles of end positions were read from the goniometer. Second, a novel testing protocol was used for the tibialis muscles. The tibialis muscles produce triplanar foot movements that cannot be measured with a 2-dimensional goniometer. Hence, assessors estimated R1 and R2 as fractions of eighths of the expected normal range of movement. Fractions of eighths were the highest resolution raters were confident using. Third, the right and left hip adductors were tested separately, and not bilaterally as tested in children.
Fourth, unlike other MTS testing protocols, a 1-minute stretch for each of the gastrocnemius and soleus muscles was applied before testing. The stretch was designed to reduce the physiological effect of thixotropic muscle changes,
Fifth, the quadriceps muscle was tested in sitting to allow for differentiating spasticity of muscle heads that cross 1 joint (vastus medialis, lateralis, and intermedius) from that of the muscle head that crosses 2 joints (rectus femoris). Sixth, the rectus femoris muscle was tested in prone, rather than supine,
Finally, for each of the 8 muscles tested, the protocol specified participants' position, start and end positions of the tested limb, and the bony landmarks used for goniometer measurements.
Statistical analysis
Intra- and interrater reliabilities were calculated for the affected and unaffected muscles. Consistency between measurements was assessed with intraclass correlation coefficients (ICCs) and 95% confidence intervals. The ICC calculations were performed for the R1, R2, and spasticity angle measurements of all but the tibialis muscles. Analysis was performed in SPSSa (version 18) with a 2-way, random single-measure model and an absolute agreement type. A priori significance levels were ICCs ≥0.8 for good consistency and ≥0.6 for moderate consistency.
were generated using Excel software.b A range of 20° or more between the upper and lower LOA was deemed clinically unacceptable.
Raters used fraction of eighths for estimating the ranges of movement of the tibialis muscles. The scale was considered an ordinal scale because raters were forced to choose from a 9-point scale. Therefore, reliability was assessed with symmetrical linear-weighted kappa calculations. Interpretation of the kappa coefficients was as follows: .41 to .60, moderate; .61 to .80, substantial; and .81 to 1.00, almost perfect reliability.
Percentage agreements were calculated for each measurement as the sum of agreement observations divided by the total number of observations. Percentage agreement of 0.7 or more was considered high. Unlike kappa calculations, percentage agreement is unaffected by data aggregated on 1 end of the scale. For this reason, high percentage agreement (≥.70) with a low kappa coefficient served to identify skewed data distribution associated with a low kappa coefficient.
Reliability of the qualitative ratings of spasticity was also assessed using the symmetrical linear-weighted kappa and percentage agreement calculations.
Results
Thirty-two participants were recruited, and 2 participants were unable to attend repeat assessments. The mean age ± SD of the 30 included participants was 54.1±12.5 years (range, 25–86y), and 12 were women. On average ± SD, participants sustained a neurologic injury 106±144 months before testing (range, 10–625mo). Participants' diagnoses included ischemic stroke (n=12), hemorrhagic stroke (n=8), traumatic brain injury (n=4), incomplete spinal cord injury (American Spinal Injury Association grade D, n=3), cerebral palsy (n=1), excised meningioma (stable at the time of testing, n=1), and primary progressive multiple sclerosis (stable at the time of testing, n=1). Participants' heterogeneity was representative of the range of neurologic conditions seen in the outpatient neurorehabilitation unit at Caulfield Hospital and similar services. The neurologic mechanisms underlying the observed spasticity in any of the participants were expected to be similar—that is, upper motor neuron lesions associated with hypersensitive stretch reflexes.
Ten participants were affected in the right lower limb, 15 in the left lower limb, and 5 bilaterally. Reliability was assessed for 35 affected sides and 25 unaffected sides. Eight muscles were assessed for each side; thus, 16 muscles were assessed on each of the 3 assessments. Each assessment took an average ± SD of 23±3.00 minutes (range, 18–37min).
Intrarater reliability
ICCs were moderately to highly reliable for the R1 and R2 measurements of the affected hamstrings, gastrocnemius, rectus femoris, and soleus (mean ± SD, .83±.08) (fig 1, table 2). Poor reliability was found for all measurements of the affected adductors and quadriceps. Moderate reliability was found for the spasticity angle measurements of the affected hamstrings, rectus femoris, and soleus. LOA were mostly unacceptably wide (fig 2, table 3). Kappa coefficients indicated moderate to substantial reliability for all measurements of the affected tibialis anterior and posterior (table 4).
Fig 1ICC calculations with 95% confidence intervals for intra- and interrater quantitative measurements. Abbreviations: Add, hip adductors; Gast, gastrocnemius; Hams, hamstrings; Quads, quadriceps; RF, rectus femoris; R2–R1, spasticity angle.
Fig 2Bland and Altman plot for intrarater measurements of slow movements of the affected rectus femoris. Each data point represents a participant. On the x axis is the mean of the 2 measurements taken by the same rater, while on the y axis the difference between the 2 measurements is plotted. The bold horizontal lines represent the group upper LOA of 9.88°, bias of −.11°, and lower LOA of −10.11°, respectively. The range between the upper and lower LOA is 19.99°.
Kappa coefficients for qualitative spasticity ratings were moderately to substantially reliable for the affected hamstrings, gastrocnemius, soleus, tibialis anterior, and tibialis posterior (table 5). Low kappa coefficients were found for the affected adductors, rectus femoris, and quadriceps.
Table 5Weighted kappa and percentage agreement calculations for qualitative spasticity measurements
ICCs were moderately to highly reliable for the R1 and R2 measurements of the affected adductors, hamstrings, gastrocnemius, rectus femoris, and soleus (mean ± SD, .73±.07) (see fig 1, table 6). Poor reliability was found for all measurements of the affected quadriceps. Spasticity angle measurements were moderately reliable for the affected hamstrings and rectus femoris. LOA were mostly unacceptably wide (table 7). Kappa coefficients for the tibialis muscles showed moderate reliability for the R1 and R2 of the affected tibialis anterior, and R2 of the affected tibialis posterior (see table 4). Low reliability was found for spasticity angle measurements of the affected tibialis anterior and posterior muscles.
Kappa coefficients for the qualitative spasticity ratings showed moderate reliability for most affected muscles (see table 5). Low reliability was observed for the affected adductors, soleus, and quadriceps.
Discussion
To our knowledge, this is the first report of MTS reliability where a wide range of lower limb muscles was tested in a heterogeneous population of adults with chronic neurologic injuries. With the use of a stringent and standardized protocol, moderate to high reliability was found for intra- and interrater R1 and R2 measurements of the affected hamstrings, gastrocnemius, rectus femoris, soleus, and tibialis anterior. The standardized protocol was one of the strengths of this study. A previous study
has shown that using a standard protocol improved the reliability of the MTS performed by experienced examiners. This is because the standard protocol reduces measurement variability attributed to the use of various participants' positions, raters' handling methods, speeds of movement, and goniometer placements. Both our and previous
findings suggest that to achieve an acceptable level of reliability, examiners should use a standardized protocol.
Are MTS measurements reliable for all muscles?
Results suggest that the reliability of R1 and R2 measurements of the MTS is (1) acceptable for both intra- and interrater measurements of the affected hamstrings, gastrocnemius, rectus femoris, soleus, and tibialis anterior muscles; (2) inconsistent for measurements of the affected adductors and tibialis posterior muscles; and (3) unacceptable for measurements of the affected quadriceps muscle.
Acceptable intra- and interrater reliability of R1 and R2 measurements for the affected hamstrings was reported in studies of children with cerebral palsy
Our findings extend this knowledge and show acceptable reliability of hamstrings measurements in conscious adults with a variety of chronic neurologic injuries.
of the gastrocnemius and soleus muscles in both adults and children reported acceptable intrarater reliability of R1 and R2 measurements. Interrater reliability, however, was found to be acceptable in some studies
We found acceptable reliability for both intra- and interrater gastrocnemius and soleus muscle measurements. The 1-minute stretch before MTS measurements was the main difference between previous studies and the current one, and may have contributed to the acceptable reliability reported. The stretch was likely to reduce the thixotropic resistance to passive movements.
Because of the larger limbs of adults, we tested the hip adductors unilaterally. Measurements performed on the same day (interrater) were reliable, but measurements performed on different days (intrarater) were not. Daily variations in the degree of muscle spasticity may explain these findings. Another explanation may be the lack of muscle isolation in testing this muscle group. The clinical implication of our findings is that MTS measurements of the adductor muscles may not be useful for assessing changes in spasticity or treatment efficacy.
To the best knowledge of the authors this is the first time that the reliability of MTS measurements for the tibialis muscles was studied. The acceptable intrarater reliability suggests that fraction of movement estimations for assessing the tibialis muscles can be performed reliably by the same rater. Estimations of different raters, however, were consistent for the tibialis anterior but not for the tibialis posterior. Whereas the tibialis anterior was tested in an anatomic position in supine, the tibialis posterior was tested in prone with the knee bent. The positioning of the limb for tibialis posterior testing may have presented a challenge for both the handling of the foot and the visual estimation of fractions of movements.
Knee extensors were studied by 1 group, and acceptable reliability was reported.
However, a meaningful comparison with the current study is not possible because of methodological differences. We found acceptable reliability for rectus femoris measurements but not for quadriceps measurements. Lack of reliability of the quadriceps measurements may have been due to participants' position in sitting rather than supine (which could have affected limb relaxation), difficulty in controlling the lower leg during the tested movement, or both.
Which MTS measurements are best used?
The quantitative measurement of R1 seems to be the most reliable spasticity measurement. Reliability of spasticity angle measurements was less consistent in accordance with previous reports in children.
Spasticity angles are produced by subtracting R1 from R2. The compounding of R1 and R2 errors contributed to the lower reliability of spasticity angles. For this reason, clinicians would be more confident with comparisons of R1 rather than spasticity angle measurements. Qualitative ratings of spasticity were inconsistent and moderately reliable. Furthermore, the reliability of qualitative ratings was not always consistent with the reliability of quantitative measurements. Our findings are supported by a previous study
in showing that the qualitative assessment of spasticity is not as reliable as the quantitative assessment. The likely reason is the subjective nature of the qualitative assessment.
Study limitations
As with most studies, inclusion criteria pose a restriction on the studied sample; hence, the study results could only be generalized to populations with similar characteristics to those of the studied sample. The protocol for the current study did not include measurements of the speed of passive movements, the force applied to the limb, and muscle electromyographic recordings. Lack of these measurements may be considered as a limitation; however, the study was designed to reflect clinical use of the MTS.
The raters were trained and experienced physiotherapists. Repeating the study with less experienced raters may reveal lower reliability. However, a recent study
has shown that the reliability of inexperienced raters was compatible with that of experienced raters provided all raters underwent MTS training.
Conclusions
The R1 and R2 MTS measurements are reliable for assessing spasticity in the hamstrings, rectus femoris, gastrocnemius, soleus, and tibialis anterior of adults with a wide range of chronic neurologic injuries. Both intra- and interrater measurements are reliable. The clinical implications are that MTS measurements could be used reliably by the same or different clinicians for assessing changes in spasticity of these muscles, treatment efficacy, or both. In contrast, R1 measurements of the tibialis posterior are reliable only when performed by the same clinician. For this reason, it is preferable for the same clinician to perform repeated MTS measurements. The R1 and R2 measurements of the hip adductors muscles were reliable only when performed on the same day, and therefore are not useful for assessing changes in spasticity and treatment efficacy. None of the MTS measurements of the quadriceps were reliable, which casts doubt on their clinical utility. The R1 and R2 MTS measurements were more reliable than the spasticity angle measurements or the qualitative ratings of spasticity. Clinicians should therefore use the R1 rather than spasticity angle measurements for reliable spasticity assessment.
Suppliers
a.
SPSS Inc, 233 S Wacker Dr, 11th Fl, Chicago, IL 60606.
b.
Microsoft Excel; Microsoft Corp, One Microsoft Way, Redmond, WA 98052.
Acknowledgments
We thank J. Olver, MBBS, MD, FAFRM(RACP), N.A. Lannin, PhD, Mr. Rod Sturt, Ms. Ashleigh Mooney, and Ms. Genevieve Tole for comments on the manuscript.
References
Lance J.W.
Symposium synopsis.
in: Feldman R.G. Young R.R. Koella W.P. Spasticity: disordered motor control. Year Book Medical,
Chicago1980: 485-494
Feasibility, test-retest reliability, and interrater reliability of the Modified Ashworth Scale and Modified Tardieu Scale in persons with profound intellectual and multiple disabilities.
Clinical assessment of ankle plantarflexor spasticity in adult patients after stroke: inter- and intra-rater reliability of the Modified Tardieu Scale.
Supported by Alfred Health : “The Caulfield Hospital Research Grants Program—Major Project Grant 2010–2011” and the “Senior Clinician Physiotherapist Research Fellowship 2011–2012.”
No commercial party having a direct financial interest in the results of the research supporting this article has conferred or will confer a benefit on the authors or on any organization with which the authors are associated.
00522-4&id=gr1.jpg){kind=link}
00522-4&id=gr2.jpg){kind=link}