Intratester Reliability and Validity of Concentric Measurements Using a New Hand-Held Dynamometer

Article Outline

• Abstract
• Methods
  • Instrumentation
    • The mechanical arm
    • Dynamometers
  • Protocol
    • Clinicians
    • Design
  • Statistical Analysis
    • Descriptive statistics
    • Intratester reliability
    • Validity
• Results
  • Descriptive Statistics
  • Intratester Reliability
  • Validity
• Discussion
  • End Range of Motion and Angle of Peak Torque
  • Evaluation of Hand-held Dynamometer in the Future
• Conclusions
• Acknowledgments
• References
• Copyright

Abstract 

Janssen JC, Le-Ngoc L. Intratester reliability and validity of concentric measurements using a new hand-held dynamometer.

Objective

To assess the reliability of a new hand-held dynamometer (HHD) to perform concentric measurements, and to determine the agreement between the HHD and the criterion standard isokinetic dynamometer.

Design

Elbow flexion concentric measurements were performed on a mechanical arm using the HHD and the isokinetic dynamometer.

Setting

Engineering laboratory and university strength-testing facility.

Participants

Three patient profiles, differing in range of motion (ROM) and strength, were simulated by a mechanical arm.

Interventions

Not applicable.

Main Outcome Measures

Peak torque and ROM obtained from concentric elbow flexion profiles.

Results

Intratester reliabilities, measured with the intraclass correlation coefficient (ICC_1,1), of the peak torque and start and end ROM are excellent for both the HHD (.99, .98, and .99, respectively) and the isokinetic dynamometer (.99 for all 3 variables). The angle of peak torque was rated fair to good in intrareliability for both devices, at .64 (HHD) and .69 (isokinetic dynamometer). Validity, measured within the limits of agreement (LOA) between the 2 devices, was clinically acceptable for peak torque and start ROM, although not for end ROM and angle of peak torque.

Conclusions

It is possible to use the new HHD to obtain dynamic measurements of joint motion. Intratester reliability of the HHD is excellent and is in clinical acceptable agreement with the isokinetic dynamometer for peak torque and start ROM. End ROM was, however, not in agreement because of a systematic error in the isokinetic dynamometer measurement for 1 of the 3 tested profiles. Intratester reliabilities of the angle of peak torque were fair to good for both the HHD and isokinetic dynamometer, but the LOA were not clinically acceptable. Stability of the arm and speed of measurement might be confounding factors in this study.

Key Words: Muscle strength dynamometer, Rehabilitation, Reliability and validity

List of Abbreviations: ANOVA, analysis of variance, HHD, hand-held dynamometer, ICC, intraclass correlation coefficient, LOA, limits of agreement, ROM, range of motion

MEASUREMENT OF STRENGTH and ROM is a common way to assess the status of patients. It is generally accepted that manual muscle strength measurements, using the modified Oxford scale, are subjective and cannot reliably distinguish subtle differences in strength.¹ More objective results can be obtained by using an isometric HHD. Such dynamometers have been proven to have good to excellent reliability in different populations.², ³, ⁴ In a single test, however, they can assess the strength of a patient at only one joint angle, and not through the patient's entire ROM.²

Stationary isokinetic dynamometers, can provide highly accurate isokinetic measurements. The isokinetic dynamometer is considered to be the standard in simultaneous strength and angle measurements.⁵, ⁶, ⁷ The advantages of the isokinetic dynamometer over the current HHDs are that tester's strength is not an issue, the patient is stabilized during testing, and the joint angle and strength are measured simultaneously during testing.³, ⁶, ⁸, ⁹ The latter is important because strength is dependent on joint angle.⁵ Disadvantages of devices such as the isokinetic dynamometer are their size and cost, which make them impractical for routine clinical examinations.¹, ⁷, ¹⁰ Li et al¹ suggested a cheaper and less bulky alternative to stationary isokinetic dynamometers: a manual muscle tester system that comprises a hand-held force transducer, 2 motion sensor packs, and a pocket personal computer. Unfortunately, the authors did not perform any dynamic measurements using their system.

A new HHD combines all 3 components suggested by Li into a single hand-held unit. This new device can measure force and angle simultaneously and therefore can measure the joint positions during testing. This can be used in an isometric setting to provide feedback on the tested angle, or in a dynamic setting to obtain data on muscle dynamics that are more functionally relevant than isometric measurements.¹⁰ In addition, the device can provide feedback on the applied force, which allows standardization of passive ROM measurements.

In this study, the intratester reliability of concentric assessments using the new HHD was tested with repeated measurements, and the validity was determined by comparing results with the isokinetic dynamometer. If proven reliable and valid, this device might help clinicians to assess patients more accurately and objectively.

Back to Article Outline

Methods 

Instrumentation 

New protocols and signal processing algorithms are required to allow clinicians to perform dynamic measurement reliably and accurately using the HHD. Confounding human factors, such as the influence of prestretching before measurement¹¹ and the uncertainty of maximal effort of the patient,⁹ would hinder the development of this process. In this research, a mechanical arm has been built to eliminate the human factors, while retaining all other aspects of using the device to capture dynamic data. Those aspects include palpation of bony landmarks, alignment of the device by eye, handling of the device, and the physiotherapist's skills in stabilizing and controlling the joint movement.

The mechanical arm, which was developed from shoulder to fingertips, is shown in figure 1. It is driven by a pneumatic cylinder that provides a force, dependent on the joint angle, so that typical flexion and extension of the right elbow can be mimicked. The artificial shoulder joint of the mechanical arm was designed to allow motion in the frontal and sagittal planes, while the artificial elbow may move only in the sagittal plane. The elbow joint is constrained to move between 10° hyperextension (–10° flexion) to 135° flexion. The maximum force at the wrist is limited to 120 N. This strength was chosen because Wikholm and Bohannon¹² stated that forces below this threshold were less likely to be affected by limited tester strength. This mechanical arm was configured to simulate 3 strength profiles, each differing in ROM and strength, reflecting 3 different patients. Start ROM varied from hyperextension to extension contracture, end ROM varied from flexion contracture to normal flexion, and force varied from 3+ to 5 on the modified Oxford scale. Several important landmarks such as the acromion, lateral epicondyle of the humerus, olecranon, and radial styloid were placed on the arm for identification purposes.

• 
  • View Large Image
  • Download to PowerPoint

• Fig 1. 

  The mechanical arm.

Two dynamometers, the isokinetic dynamometer (Biodex^a) and the HHD,^b were used to measure strength and ROM of different profiles of the mechanical arm. The new HHD was designed to measure force and ROM simultaneously (fig 2). In this study, the angle measurement was limited to motion in the vertical plane. In a laboratory setting, it has a known accuracy of ±1° for angle and ±1N for force in the ranges of 0° to 359° and 0 to 750N, respectively. A 3-step procedure was followed to capture strength versus joint angle data using the HHD: (1) defining the zero angle of the joint; (2) starting the measurement; and (3) stopping the measurement. In this study, the Biodex measurements were corrected for the effect of gravity caused by the Biodex lever arm.

• 
  • View Large Image
  • Download to PowerPoint

• Fig 2. 

  The new HHD with simultaneous force and angle measurement.

Protocol 

One registered physiotherapist was recruited and trained to assess the strength and ROM using the HHD and the Biodex.

Three profiles of the mechanical arm were tested. Each profile was tested 3 times with the Biodex and 5 times with the HHD using a maximal concentric flexion test, in which the examiner holds the dynamometer while the subject exerts maximal force against the dynamometer and the examiner through a ROM. The test was first performed with the Biodex, then with the HHD. The testing order of the 3 profiles was randomized, and the physiotherapist performing the tests was unable to see the outcome measures. Both dynamometers were calibrated before the test. Start ROM, end ROM, peak torque, and angle of peak torque were obtained from the torque versus joint angle curves captured by the dynamometers.

In the Biodex protocol, the rotational axis of the mechanical arm was aligned with the rotational axis of the Biodex, and the Biodex lever arm was in contact with the wrist of the mechanical arm .23m from the rotational axis. The isokinetic mode of the Biodex was used for testing with a maximum speed of 60°/s. In this mode the total ROM had to be set before starting the test.

In the HHD protocol, the physiotherapist stabilized the arm with the left hand and used the device with the right hand (fig 3). To measure ROM and torque during elbow flexion, the zero position of the elbow was identified by placing the device lengthwise on the reference line between the acromion and the lateral epicondyle of the humerus and clicking a button on the HHD (defining the zero position). Then the device was placed with the force pad 2cm proximal of the wrist (.23m from rotational axis of the elbow), and the button was clicked a second time to commence the measurement. The physiotherapist gave a cue that the mechanical arm could be flexed, and as it flexed, the physiotherapist resisted the force in such a manner that the speed remained as constant as possible and with an average below 60°/s.¹³ The third click ended the measurement, and the device could then be lifted off the mechanical arm. Each measurement was followed by a 30-second rest period.

• 
  • View Large Image
  • Download to PowerPoint

• Fig 3. 

  Photos of (A) the Biodex measurement and (B) HHD measurement.

Statistical Analysis 

Descriptive statistics of muscle torques and angles are presented in newton meters and degrees, respectively. Torque is calculated from the measured peak force times the length from the center of the force pad to the rotational axis of the elbow (.23m). Mean and SDs are reported. All analysis was performed using the Matlab software package.^c

The degree of correlation between the 5 repetitions of all profiles of the HHD test was calculated with the ICC_1,1 defined by Shrout and Fleiss.¹⁴ The same test was performed on the 3 repetitions of the Biodex. The most critical reliability assessment is the ICC_1,1, which assumes that every individual measurement is independent and the error of measurement is assumed to be normally distributed.¹⁴ Other authors have used ICC_2,1 for their reliability measurement, which tends to give more optimistic values than ICC_1,1. In this article, all ICC_1,1 results are almost equal to the ICC_2,1 values. According to Fleiss,¹⁵ the reliability of an ICC over .75 is considered to be excellent, and between 0.4 and .75 as fair to good.

The agreement between the 2 devices can be quantified using the Bland-Altman 95% LOA method,¹⁶, ¹⁷, ¹⁸, ¹⁹ where the mean and SD of the differences between the measurements by the 2 devices are calculated. For repeated measurements, a 1-way ANOVA is performed for each device separately. Outcomes of the 1-way ANOVA are then used to calculate the lower and upper LOA (mean ± 1.96SD), which are then compared with predefined clinically acceptable limits. In this study, clinical acceptability of the HHD is set to 10° in angle and 5Nm in torque; that is, the difference between the upper and lower LOA should not exceed these values.

Back to Article Outline

Results 

Descriptive Statistics 

Table 1 shows the means and SDs of peak torque, angle of peak torque, start ROM, and end ROM for the HHD and the Biodex measurements. In both devices, profile B shows the highest peak torque, while profile A shows the lowest peak torque. The angle of peak torque was measured between 60° and 72° of flexion in all profiles. Both devices measured a hyperextension in the A and B profiles and a flexion contracture of 14° to 15° in the C profile. End ROM was highest in the C profile.

Table 1. Peak Torque, Angle of Peak Torque, Start ROM, and End ROM

NOTE. Values are mean ± SD.

Abbreviation: PT, peak torque.

Figure 4 shows the 3 different profiles measured with the HHD and the Biodex. The HHD graph shows a similar curve to that of the Biodex, identifying a start and end ROM and a peak torque.

• 
  • View Large Image
  • Download to PowerPoint

• Fig 4. 

  Three profiles of the mechanical arm measured with (A) the HHD and (B) the Biodex.

Figure 5 shows the angle versus time graphs of 5 repeated tests for each of the profiles using the HHD (see fig 5A–C), and the Biodex tests at 60°/s (see fig 5D). Although it is not possible to obtain perfect isokinetic results with the HHD, these graphs show that the manual tests with the HHD can produce reasonably constant speeds (linear graphs in the midrange for at least 70% of the ROM). It is also evident from these graphs that the manual tests were very consistent.

• 
  • View Large Image
  • Download to PowerPoint

• Fig 5. 

  Angle versus time graphs showing repeated measurements for (A) the HHD on profile A, (B) on profile B, (C) on profile C, and (D) the Biodex measurement with speed setting at 60°/s on profile B.

Intratester Reliability 

Five repetitions measured with the HHD and 3 repetitions with the Biodex were used to calculate the ICC and their 95% confidence intervals. The results are shown in table 2. The ICC_1,1 values of both devices indicate excellent intratester reliability in the peak torque and start and end ROM. Repeatability of the angle of peak torque by both devices is rated fair to good.

Table 2. ICC Between 5 Repetitions for the HHD Test and 3 Repetitions for the Biodex

Measurements	HHD	Biodex
PT	.99(.97–.99)	.99(.98–.99)
Angle of PT	.64(.15–.99)	.69(.15–.99)
Start ROM	.98(.91–.99)	.99(.99–.99)
End ROM	.99(.94–.99)	.99(.98–.99)

NOTE. 95% confidence intervals shown in parenthesis.

Abbreviation: PT, peak torque.

Validity 

LOA, calculated for peak torque, angle of peak torque, start ROM, and end ROM, are shown in figure 6. LOA for the peak torque are –0.2 and 2.6Nm, with a mean difference of 1.2Nm between the Biodex and the HHD. This means that the Biodex measures peak torque 1.2Nm higher on average than the HHD. LOA of the start ROM are –3.2° and 6.0°, with a mean difference of 1.4°. The end ROM gives considerably larger LOA, –11.8° and 8.8°, with a mean difference of –1.5°. The LOA for the angle of peak torque are –12.1° and 13.6°, with a mean difference of 0.7°.

• 
  • View Large Image
  • Download to PowerPoint

• Fig 6. 

  LOA plots for (A) peak torque, (B) angle of peak torque, (C) start ROM, and (D) end ROM. Circles represent the measured difference between the 2 devices for the 3 profiles. The solid line represents the mean difference between the 2 devices; the dashed lines represent the lower and upper LOA.

To investigate the end ROM disagreement, 5 repeated measurements of the start and end ROM of the 3 profiles were carried out with a plastic goniometer.^d The mean outcomes (with SD) for start ROM and end ROM were –7.8° (0.8°), 115° (0.0°) for profile A; –7.2° (1.1°), 120° (0.7°) for profile B; and 15.0° (0.7°), 135.6° (0.5°) for profile C, respectively. These measurements are comparable to the Biodex angle measurements except for the end ROM of the C profile, where the Biodex measured 6° less than that measured by the goniometer.

Back to Article Outline

Discussion 

The results from this study indicate that the measurements obtained with the HHD have excellent reliability when used by the same tester and are in agreement with the Biodex for peak torque and start ROM. However, end ROM and the angle of peak torque are not within the clinically acceptable LOA.

End Range of Motion and Angle of Peak Torque 

One explanation for the end ROM disagreement is that the mechanical arm in profile C was very weak near the end range, and the weight of the Biodex lever arm may have prevented the mechanical arm from reaching its end ROM. This was not the case in the other 2 profiles because there was sufficient power to overcome the weight of the lever arm. This effect can be seen in figure 4, where the slope of the C profile in the end ROM is less steep than in the other 2 profiles. This reduction in the end ROM of profile C as measured by the Biodex results in a larger variance of the difference between the Biodex and the HHD, hence producing LOA that are larger than the clinically acceptable value.¹⁹ In this situation, a better measure of validity of the HHD in measuring ROM is to calculate the LOA in ROM between the goniometer and the HHD. They are –3.4°, 7.4° and –4.5°, 8.2° for start ROM and end ROM, respectively. For the end ROM, the difference of the LOA between the Biodex and the HHD is greater than the difference of the LOA between the goniometer and the HHD, but it is still not within the acceptable limit. However, if total ROM were to be considered, the LOA decrease from –11.6 and 5.8, between the Biodex and the HHD, to –5.3 and 5.0 between the goniometer and the HHD, then it is marginally acceptable.

The disagreement of the angle of peak torque in this study may be attributed to speed variation in the HHD measurements and the method of stabilizing the arm. It has been observed²⁰, ²¹ that speed and its variation during measurement play a significant role in assessing the angle of peak torque. When a physiotherapist or other clinician performs a dynamic measurement, it can only be an approximation of a true isokinetic measurement. This is because the speed and its variation differ within and between measurements by the same tester. Further research on the effect of the variation in speed of the HHD is needed to improve the accuracy of the angle of peak torque measurements. Also, if the arm is not stabilized properly, then the angle of peak torque may vary and so reduce the reliability of this measurement.

Evaluation of Hand-held Dynamometer in the Future 

Four factors that are detrimental to the evaluation of the HHD have been identified. First, this study was performed using only 3 profiles (resembling 3 patients). Although 3 to 5 repetitions were made with each device, this does not reduce the LOA. Second, the zero position in this study was recorded by aligning the HHD with the reference line between the acromion and the lateral epicondyle of the humerus. The HHD was then moved on to the forearm, without identifying the reference line for the forearm (between the lateral epicondyle of the humerus and the radial styloid). This position of the HHD was not parallel to the reference line of the forearm. The mean angle of misalignment was found to be 7.8° with 1.5° SD for 30 repeated measurements. In all the measurements, this angle was added to the start ROM, end ROM, and angle at peak torque. This creates an extra variation in the collected data. Third, Martin et al⁸ investigated the accuracy of quadriceps strength measurements in elderly subjects and found an underestimation of the peak torque, with a mean of 14Nm, when it was measured with an HHD, compared with the Biodex. They also noted that there was a positive trend between measured torque and underestimation of the torque, meaning that the stronger the patient, the greater the underestimation of that force. In this study, the force of the mechanical arm was relatively low, so it may be assumed that this effect was unimportant. However, were the HHD used on bigger muscles, the observation of Martin would become significant. Dynamometers such as the Biodex do not have this problem because the strength of the tester is replaced by a motor.

The application of the HHD still needs to be proven in a clinical setting with human subjects both with and without injuries or disabilities. Because of the motivational and physiologic differences in humans, measurements in patients may demonstrate more variability. Further, an intertester reliability study with several clinicians is needed to get a broader view of the reliability of measurements obtained with the HHD. If proven reliable, the device would be expected to produce consistent results when used in different clinics.

Back to Article Outline

Conclusions 

We have shown that the new HHD can be used to obtain concentric measurements of joint motion. It has excellent intratester reliability and is in clinically acceptable agreement with the Biodex for peak torque and start ROM. End ROM was not in agreement because of a systematic error in the Biodex measurement for 1 of the 3 tested profiles. Intratester reliabilities of the angle of peak torque were fair to good for both the HHD and Biodex, but they were not within the clinical agreement limits. Stability of the arm and speed of measurement might be confounding factors for the measurement of angle of peak torque.

Further research is now required to assess the intertester reliability, and the application of the device on human subjects, before the HHD can be introduced to clinical settings to obtain objective and reliable dynamic measurements.

Suppliers

Back to Article Outline

Acknowledgments 

We thank Burwood Academy of Independent Living and the School of Physiotherapy, Otago University for their assistance in this research.

Back to Article Outline

References 

1. Li RC, Jasiewicz JM, Middleton J, et al. The development, validity, and reliability of a manual muscle testing device with integrated limb position sensors. Arch Phys Med Rehabil. 2006;87:411–417
   • View In Article
   • Abstract
   • Full Text
   • Full-Text PDF (377 KB)
   • CrossRef
2. Stratford PW, Balsor BE. A comparison of make and break tests using a hand-held dynamometer and the Kin-Com. J Orthop Sports Phys Ther. 1994;19:28–32
   • View In Article
   • MEDLINE
3. Andrews AW. Hand-held dynamometry for measuring muscle strength. J Hum Muscle Perform. 1991;1:35–50
   • View In Article
4. Bohannon RW, Andrews AW. Interrater reliability of hand-held dynamometry. Phys Ther. 1987;67:931–933
   • View In Article
   • MEDLINE
5. Ortqvist M, Gutierrez-Farewik EM, Farewik M, Jansson A, Bartonek A, Broström E. Reliability of a new instrument for measuring plantarflexor muscle strength. Arch Phys Med Rehabil. 2007;88:1164–1170
   • View In Article
   • Abstract
   • Full Text
   • Full-Text PDF (641 KB)
   • CrossRef
6. Lund H, Søndergaard K, Zachariassen T, et al. Learning effect of isokinetic measurements in healthy subjects, and reliability and comparability of Biodex and Lido dynamometers. Clin Physiol Funct Imaging. 2005;25:75–82
   • View In Article
   • MEDLINE
7. Fosang A, Baker R. A method for comparing manual muscle strength measurements with joint moments during walking. Gait Posture. 2006;24:406–411
   • View In Article
   • Abstract
   • Full Text
   • Full-Text PDF (289 KB)
   • CrossRef
8. Martin HJ, Yule V, Syddall HE, Dennison EM, Cooper C, Aihie Sayer A. Is hand-held dynamometry useful for the measurement of quadriceps strength in older people? (A comparison with the gold standard Biodex dynamometry). Gerontology. 2006;52:154–159
   • View In Article
   • MEDLINE
   • CrossRef
9. Harlaar J, Roebroeck ME, Lankhorst GJ. Computer-assisted hand-held dynamometer: low-cost instrument for muscle function assessment in rehabilitation medicine. Med Biol Eng Comput. 1996;34:329–335
   • View In Article
   • MEDLINE
   • CrossRef
10. Mital A, Kopardekar P, Motorwala A. Isokinetic pull strengths in the vertical plane: effects of speed and arm angle. Clin Biomech. 1995;10:110–112
    • View In Article
11. Dunn JC, Iversen MD. Interrater reliability of knee muscle forces obtained by hand-held dynamometer from elderly subjects with degenerative back pain. J Geriatr Phys Ther. 2003;26:23–29
    • View In Article
    • CrossRef
12. Wikholm JB, Bohannon RW. Hand-held dynamometer measurements: tester strength makes a difference. J Orthop Sports Phys Ther. 1991;13:191–197
    • View In Article
13. Baltzopoulos V, Brodie DA. Isokinetic dynamometry (Applications and limitations). Sports Med. 1989;8:101–116
    • View In Article
    • MEDLINE
    • CrossRef
14. Shrout PE, Fleiss JL. Intraclass correlations: uses in assessing rater reliability. Psychol Bull. 1979;86:420–428
    • View In Article
    • CrossRef
15. Fleiss JL. The design and analysis of clinical experiments. New York: John Wiley & Sons; 1986;
    • View In Article
16. Bland JM, Altman DG. Agreement between methods of measurement with multiple observations per individual. J Biopharm Stat. 2007;17:571–582
    • View In Article
    • CrossRef
17. Bland JM, Altman DG. Measuring agreement in method comparison studies. Stat Methods Med Res. 1999;8:135–160
    • View In Article
    • MEDLINE
    • CrossRef
18. Bland JM, Altman DG. A note on the use of the intraclass correlation coefficient in the evaluation of agreement between two methods of measurement. Comput Biol Med. 1990;20:337–340
    • View In Article
    • MEDLINE
    • CrossRef
19. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet. 1986;1:307–310
    • View In Article
    • CrossRef
20. Gore CJ. Physiological tests for elite athletes: Australian Sports Commission. In: Champaign: Human Kinetics; 2000;p. 155–177
    • View In Article
21. Kannus P. Isokinetic evaluation of muscular performance: implications for muscle testing and rehabilitation. Int J Sports Med. 1994;15:S11–S18
    • View In Article
    • CrossRef

• a Biodex; Biodex Medical Systems Inc, 20 Ramsay Rd, Shirley, NY 11967-4704.
• b Hand-held Dynamometer; Industrial Research Ltd, PO Box 20-028, Bishopdale, Christchurch 8053, New Zealand.
• c Matlab software package; The MathWorks Inc, 3 Apple Hill Dr, Natick, MA 01760-2098.
• d Goniometer G-300; Whitehall Manufacturing, PO Box 3527, City of Industry, CA 91746.