Validity and Reliability of a Hand-Held Dynamometer with Two Populations

Transcription

1 Validity and Reliability of a Hand-Held Dynamometer with Two Populations Paul R. Surburg, PhD1 Rory Suomi, PED2 Wendy K. Poppy, MS3 current positive trend is the integration of people with mental retardation into all as-,pects of our society. Persons with mental retardation are being encouraged to engage in sports, physical fitness, and recreational activities. Integration into society has taken place in all phases of living, including the seeking of medical treatment at local hospitals and sports medicine clinics. Evaluation of strength is an integral component in a rehabilitation or conditioning program. Although isokinetic dynamometry has been used widely to quantify muscle performance, there are some inherent limitations, such as generalizability of scores from one type of isokinetic dynamometer to another (1 7), the cost of these dynamometers, the practicality of administering different tests from a temporal perspective, and the portability of isokinetic dynamometers. Alternatives for strength assessment in the clinical setting are manual muscle testing and hand-held dynamometry. A factor that may affect strength assessment of persons with mental retardation is their great variability in performing motor-oriented tasks (2). Hand-held dynamometric studies have been conducted on healthy subjects (3, 16), stroke patients (9). patients with brain damage (1 3), persons with orthopaedic impairments (1 2). and children with muscular dys- With the trend in our society toward mainstreaming persons with mental retardation, there is a need to evaluate the efficacy of rehabilitation procedures for this population. The purpose of this study was to examine the validity and reliability of measurements determined with the Nicholas Manual Muscle Tester (MMT) and with the Cybex I1 isokinetic dynamometer using a sample of 20 adults without mental retardation and 10 adults with mental retardation. Utilizing stabilization techniques, knee extension and elbow flexion were measured with both dynamometers. The Nicholas MMT manifested excellent discriminatory function among subgroups. Correlational analyses between the dynamometers yielded coefficients of.74 and.77 for knee and elbow actions. Significant differences between dynamometers for the subgroups were found in three of the six analyses. lntrarater and interrater reliability coefficients were all greater than.90 for subjects with mental retardation. In conclusion, measurements of muscle strength obtained with the Nicholas MMT from subjects with and without mental retardation were reliable. The validity of this dynamometer assessed by construct and criterion processes was not conclusively established. This dynamometer, which is portable and relatively inexpensive, appears to be suitable as an assessment tool in clinical settings for persons with mental retardation. Key Words: mentally retarded adults, muscle strength, methods ' Physical therapist and professor, Department of Kinesiology, lndiana University, Bloomington, IN ' Doctoral student, Department of Kinesiology, lndiana University, Bloomington, IN ' Physical therapist, sports medicine program, Department of Kinesiology, Indiana University, Bloomington, IN Research supported hy a grant from Nicholas Institute of Sports Medicine and Athletic Trauma. trophy (I 5). The validity issue has been examined by several researchers, including Bohannon (4, 8, 9) and Sullivan (1 6), who noted that hand-held dynamometric scores and a Cybex dynamometer set to measure isometric contractions measured a similar element of strength. The consensus of these studies that addressed the reliability issue, and other studies (1, 6), was that handheld dynamometers yield reliable measurements to access isometric strength in selected muscle groups. A methodological procedure that is critical for this study is found in Byl et al (3). These researchers reported higher intrarater and interrater reli- abilities when a stabilization technique was implemented for measurements of elbow flexion and shoulder abduction. As noted earlier, subjects with mental retardation tend to be more variable in physical and motor skills. Because of these findings, a protocol utilizing stabilization techniques was implemented with the Nicholas Manual Muscle Tester (MMT). The purpose of this study was to determine the validity of the Nicholas MMT with Cybex I1 dynamometry and to assess reliability components of the Nicholas MMT. Utilizing stabilization techniques, upper and lower extremity muscle groups JOSPT Volume 16 Number 5 November

2 of male adults with mental retardation and of males and females without mental retardation were measu red. The null hypotheses were as fol- METHODS Subjects The correlation coefficients of the measurements between the two dynamometers will not be of sufficient magnitude (r =.90) to establish criterion validity. The measurements between the two dynamometers are not from the same population. The Nicholas MMT will not accurately classify subgroup membership to attain construct validity. The measurements taken with the Nicholas MMT lack the consistency between and among raters to establish a p propriate interrater and intrarater reliability. Twenty subjects without mental retardation (1 0 males, 10 females) volunteered to participate in the study. They were recruited from graduate classes at Indiana University. These subjects ranged in age from 21 to 39 years, with a mean age of 28.9 years. Ten male subjects with mild to moderate mental retardation were also participants in this study. These subjects ranged in age from 22 to 39 years, with a mean age of 28.1 years. They were employed at a sheltered workshop. Their I.Q. scores, based upon the Stanford Binet Intelligence Test, ranged from 36 to 69, with a mean of All subjects gave their informed consent, and other procedures pertaining to rights of human subjects were addressed according to university human subject protocols. These 30 subjects were free of neurological or neuromuscular impairments, hypertension, orthopaedic problems, or medication that would affect the results of isometric testing. Instruments Two instruments were used to measure isometric torque force values for knee extension and elbow flexion. Subjects were tested with the Nicholas MMT (Model # , Lafayette Instrument Company, Lafayette, IN, 47402) and the Cykx I1 isokinetic dynamometer (Cybex, Division of Lumex, Inc., 2100 Smithtown Avenue, Ronkonkoma. NY ). Both instruments were calibrated by loading weights of known values prior to the initiation of a Subjects with mental retardation tend to be more variable in physical and motor skills. testing session. The Nicholas MMT's range is kg; its precision is.5 kg. All testing was done in one of the athletic training rooms at Assembly Hall on the Indiana University campus. Testing Position Standardized positions were established with the subjects during an orientation session and were used for strength assessment with the Cybex dynamometer and Nicholas MMT. The positions used for testing are outlined in Table 1. For Cybex testing, a subject grasped the adjustable handle located below the attachment bar. The length of the dynamometer extension arm was noted for each subject, and the precise length was established for subsequent testing of each subject. With the Nicholas MMT, the location of the attachment bar of the Cybex dynamometer was determined, and the handle was then removed. This practice not only kept the angle constant but determined the exact location to situate the pressure distribution plate of the Nicholas MMT. In the case of elbow flexion, the pressure plate was positioned just proximal to the crease of the wrist on the dorsal aspect of the forearm. For subsequent calculations of torque values, the measurement of the moment arm was taken from the tip of the olecranon to the distal edge of the pressure plate in the vicinity of the radial styloid process for elbow flexion and from the knee's medial joint line to the distal edge of the pressure plate slightly above the tip of the medial malleolus for knee extension. As with the Sullivan study (16). no gravity correction adjustments were a p plied to the dynamometric data. Procedures Subjects were involved in three testing sessions: one orientation session, test session one, and test session two. The orientation session consisted of the following components: presenting standardized instructions, watching visual demonstrations, and practicing the appropriate tests. Before each test session, subjects engaged in warm-up activities, which included a 5-minute ride on a bicycle ergometer, hamstring and calf stretching, and elbow range of motion exercises. During one testing session, five trials were administered with the hand-held dynamometer by each rater for both motions. When the Nicholas MMT was used, a 5- minute rest interval preceded the administration of a second set of measurements by the other rater. The other testing session consisted of three isometric trails on the Cybex I1 dynamometer, which had a damp setting of "2". A 5-second isometric contraction constituted a trial with Volume 16 Number 5 November 1992 JOSPT

3 BodV Point of Mde Group Position Limb Position Application Stabilization Elbow flexors Supine Shoulder abducted Proximal to Straps at subject's 2O0, elbow supi- crease of waist, contralatnated and flexed wrist era1 hand at 90' grasps handle on side of table Knee extensors Seated Knee flexed at 60' Proximal to Straps at shoultrunk-thigh angle medial der, waist and 110' malleolar thigh; subject TABLE 1. Jest positions. interpolated rest periods of 1 minute between trials. Sequencing of dynamometer testing and rater measurements was randomly determined. No verbal promoting was used, and the first trial of each testing sequence was considered part of the orientation program. The rater who was not administering the test recorded the score, which was in kg for the Nicholas MMT and ft-lbs for the Cybex I1 dynamometer. One rater was a female physical therapist with 17 years of clinical experience. The other rater was a male doctoral student in the Department of Kinesiolow. This person has had extensive experience evaluating strength parameters with mentally retarded persons and collegiate athletes. Data Analysis Mean force values from the Nicholas MMT were converted to Nm by changing kg into N and multiplying by the force arm. Mean torque values of the two actions measured with Cybex I1 were converted from ft-lbs to Nm. Means and standard deviations for the three subgroups were calculated for both dynamometers. Both criterion and construct analyses were used to address the validity issue. Pearson product moment correlations were calculated to establish criterion validity between the measurements of the Cybex dyna- line grasps handles located on both sides of chair mometer and the Nicholas MMT. A correlation coefficient of.90 was established as the criterion value for this phase of the analysis. Another phase of the criterion validity analysis was to examine if significant differences existed between the measurement of the two instruments. For each action, a student t-test was conducted between the measurements of the two dynamometers. Prior to conducting discriminant function analyses, one-way analysis of variance (ANOVA) was conducted on the scores of the Nicholas MMT for each action to determine if significant differences were found among the subgroups (p =.05). Tukey's orthogonal value was used to determine post hoc pairwise differences. Discriminant function analyses were then computed to ascertain if the Nicholas MMT classified subjects according to subgroup membership. Intraclass correlation coefficients were computed to examine intrarater consistency among the trials administered with the Nicholas MMT (1 4). The nuances of interrater reliability were examined by means of generalizability coefficients (G) (1 1). This coefficient is calculated based upon a number of factors inherent in the experimental design. An inappropriate procedure is to square the G value to ascertain the percent of variance, but a G value of.80 or higher is considered to be of high enough magnitude for significant interpretative purposes (10)- RESULTS Mean torque values and standard deviations for the three subgroups and two dynamometers are found in Table 2. When criterion validity was examined and all subgroups were analyzed as one group, the Pearson product moment correlations between the two dynamometers were.74 and.77 for knee and elbow measurements, respectively. The other phase of this criterion analysis involved determining if For measurements of men with mental retardation and female subjects, no decrease was evident when the number of triais was reduced from four to two, a significant difference was evident between the measurements of the two dynamometers for each motion (Table 2). Significant differences between the measurements of the two dynamometers were found for three of the comparisons. A preliminary step in the construct validity analysis was to determine if significant differences were evident among the subgroups for each joint action with the Nicholas MMT. Significant F values were calculated for knee extension [F (2, 27) = 8.57 (p < 0.05)] and elbow flexion values of the Nicholas MMT [F (2, 27) = 21.0 (p <.05)]. Post hoc analysis revealed significant differences for both motions between the JOSPT Volume 16 Number 5 November 1992

4 Elbow Knee Flexion t Extension t Group N Instrument X/(s) Val ue X/(s) Value Men 10 Cybex 59.7' (12.96) (23.3) 2.37" 2.3; 0.15 Nicholas 69.1 (9.34) (32.6) Women 10 Cybex 30.9 (6.04) (19.79) 2.74" 2.7l 4.60" Nicholas (10.08) (15.9) Men with mental retardation 10 Cybex 32.3 (1 3.8) (33.12) t 0.22 Nicholas (16.08) (39.7) ' Newton-meters. " Critical t value = 2.26, p = TABLE 2. Means, standard deviations of muscle actions and comparisons oldynamometers. men without mental retardation and the other two subgroups. Two discriminant function analyses were conducted on the Nicholas MMT data (Table 3). This was predicated upon the fact that no significant differences were found between the subgroups of women and men with mental retardation. Based upon these discriminant function analyses, the scores of the Nicholas MMT were able to differentiate at 100% iiccuracy rate measurements of men without mental retardation from scores of men with mental retardation or women. Intraclass correlation coefficients were calculated for both raters to ascertain the intrarater reliability of the Nicholas MMT. The consistency of the measurements on the males ' p = with retardation was found to range from.98 for Rater 2 to.99 for Rater 1 (Table 4). With the exception of Rater 2 on knee extension for women subjects, all intrarater coefficients for the Nicholas MMT were 0.90 or higher. Intraclass correlation coefficients were calculated on one set of measurements taken on the Cybex I1 dynamometer; coefficients ranged from.84 to.99 for all three subgroups (Table 4). Generalizability coefficients were calculated to evaluate interrater reliability (Table 5). The coefficients ranged from.90 to.98 for two raters and four trials. It is also noteworthy that the G values associated with cholils MMT. More specifically, for subjects with mental retardation, these results do not completely re- Accurately Nicholas Predicted Group Actual Group Wilks' Lambda Number of Cases Membership Males % 0.38' Males with mental 10 1W/o retardation Males % 0.29' Females % TABLE 3. Discriminant function analvsis of measurements taken with two dynamometers. solve the issue of instrument validity but do show potential and, hopefully, will stimulate additional research in this area. Two facets of reliability were exanlined in this study. Intrarater reliability of strength measurements taken with the Nicholas MMT on subjects with mental retardation was high for both raters (r = ) on elbow flexion. Correlations of similar magnitude with hand-held dynamometers were reported by Riddle (1 3) and Bohannon (5) with braindamaged and neurologically involved subjects. For knee extension, the intrarater consistency was again in the range. Ranges of.93- and -.98 were noted by Riddle and Bohannon. Inspection of Table 4 reveals that with the exception of Rater 2 on knee extension of women subjects (r =.83). all intrarater coefficients were in the range. These results show that handheld dynamometers, including the Nicholas MMT, can provide consistent intrarater measurements. Interrater reliability was examined by the use of G coefficients. Generalizability coefficients are not equivocal to Pearson product moment correlations or intraclass correlations. As noted earlier,.8 is considered the critical value for assessing the appropriate degree of relationship. All G coefficients were.89 or higher, with the measurements by the raters of men with mental retardation attaining the highest rater agreement (-.98). In calculating G coefficients, it is a simple task to determine G values for less than the number of trials measured in a study. Examination of G coefficients for less than four trials provides evidence of very little decrease in interrater reliability. For measurements of men with mental retardation and female subjects, no decrease was evident when the number of trials was reduced from four trials to two. The G coefficient for elbow flexion only decreased from.90 to.89 for men without mental retardation. These Volume I6 Number 5 November 1992 JOSPT

5 U- Wn-rn Men,..~.,.."..~.. Mentally Retarded Motion Rater, Rater, Rater. Rater, Rater. Rater, Nicholas Manual Muscle Tester Knee extension.99' Elbow flexion Men Knee extension.95 Elbow flexion.98 'Criterion reference: r =.90. TABLE 4. lntrarater reliability of dynamometers. Criterion reference: C =.80. " Gneralizability formula (I I): C( pi) = Cybex I1 Men.89' Women Men-Mentally Retarded 2, 2, + 2,oi + 2,,1, TABLE 5. Generalizability coefficients of Nicholas manual muscle tester. findings support the results of Bohannon (7), which showed that three trials or less may be of sufficient number to achieve reliable measurements of dynamometric measurements. Bohannon's research, however, did not evaluate persons with mental retardation, which is an unique element of this study. The results of this study, which were predicted upon certain stabilization techniques, do support future investigations utilizing more clinically oriented protocols using handheld dynamometers but good stabilization is needed. One must exercise caution about generalizing these measurements taken on subjects with mental retardation were or.98. Women Men ~entalmetarded DISCUSSION The issue of validity for handheld dynamometers has been examined previously (8, 9, 1 2, 16). but rarely are both criterion and construct validity addressed in the same study. The results of these analyses provide a mixed picture of the validity issue. Correlation coefficients between Nicholas MMT and the Cybex I I isometric measurements were in the high range (r = ) when compared with similar measurements in a study by Sullivan (1 6) (r = ). The coefficients in this study, however, did not reach the accepted value of.90. This mixed profile is exemplified in the analyses concern- ing significant differences between the means of the two dynamometers for the three subgroups. Significant differences were found between the isometric means in three of the six analyses. Differences, however, were not evident with either action for subjects with mental retardation. Similar nonsignificant differences were reported b$ Bohannon (9) with stroke patients and by Sullivan (16) with healthy subjects. A word of caution is appropriate when proposing a finding based upon nonsignificant analyses. A perusal of Table 2 finds the subjects with mental retardation exhibiting higher measurements with the Nicholas MMT for both actions. For subjects without mental retardation, the scores for the Nicholas MMT were lower on knee extension. One possible explanation is a discomfort factor, which may have precluded a maximal effort on this test. The pressure distribution plate of the Nicholas MMT was applied on the skin of the leg with the lower edge of the plate just slightly above the tip of the medial malleolus. Subjects did not have much padding between the bone and plate. The combination of sparse padding and greater force generated per unit area with knee extension may have caused an inhibitory condition. Precautions may need to be taken when evaluating motions that generate high force values and involve a pressure distribution plate on parts of the body where there is little natural padding afforded by muscle or adipose tissue. The construct validity analysis presented a more favorable assessment of the Nicholas MMT. This dynamometer predicted 100% membership for not only the subjects with mental retardation but for all subgroups. Based upon the results of the entire validity analysis, one must be cautious about any definitive statement pertaining to the validity of measurements taken with the Nifindings to or between different populations. Future studies should in- JOSPT * Volume 16 * Number 5 * November 1992

6 clude different muscle groups, more clinically related protocols, and comparisons of the Nicholas MMT with other hand-held dynamometers in a single experiment. SUMMARY Results of this study indicate that when following manual muscle testing guidelines and adding stabilization, the Nicholas MMT does discriminate the strength performance of elbow flexors and knee extensors of females and males who are mentally retarded with males without retardation based upon construct validity analysis. Criterion validity analyses, however, did not completely confirm the results of the construct analysis. Acceptable levels of intrarater and interrater consistency were found with the use of the Nicholas MMT on three subgroups of subjects. JOSPT REFERENCES 1. Agre IC, Magness ll, Hull SZ, Wright KC, Baxter TL, Patterson R, Stradel L: Strength testing with a portable dynamometer: Reliability for upper and lower extremities. Arch Phys Med Rehabil68: , Berkson C, Baumeister AA: Reaction time variability of mental defectives and normals. Am I Ment Defic 72: , Bly N, Shannon R, Asturias I: lntrarater and interrater reliability of strength measurements of the biceps and deltoid using a hand-held dynamometer. I Orthop Sports Phys Ther 9: , Bohannon RW: Manual muscle test scores and dynamometer test scores of knee extension. Arch Phys Med Rehabil67: , Bohannon RW: Test-retest reliability of hand-held dynamometry during a single session of strength assessment. Phys Ther 66: , Bohannon RW, Andrews AW: Interrater reliability of hand-held dynamometry. Phys Ther 67: , Bohannon RW: Hand-held dynamometry: Stability of muscle strength measurements. Clin Biomech 2:74-77, Bohannon RW: Hand-held compared with isokinetic dynamometry for measurement of static knee extension torque (parallel reliability of dynamometers). Clin Phys Physiol Meas 1 1 : , Bohannon RW: Knee extension torque in stroke patients: Comparison of measurements obtained with a handheld and Cybex dynamometer. Physiother Can 42: , Cardinet 1, Tourneur Y, Allal L: The symmetry of generalizability theory: applications to educational measure- ment. / Educ Meas 13: , Codbout P, Schutz RW: Ceneralizability of ratings of motor performances with reference to various observational designs. Res Q 54:20-27, Marino M, Nicholas /A, Gilbert W, Cleim MS, Rosenthal P, Nicholas I: The efficacy of manual assessment of muscle strength using a new device. Am / Sports Med 10: , Riddle DL, Finucane SD, Rothstein jm, Walker M: lntrasession and intersession reliability of hand-held dynamometer measurements taken on brain-damaged patients. Phys Ther 69: , Safrit MI: Evaluation in Physical Education, Englewood Cliffs, NI: Prentice- Hall, Stuberg WA, Metcalf WK: Reliability of quantitative muscle testing in healthy children and in children with Duchenne muscular dystrophy using a hand-held dynamometer. Phys Ther 68: , Sullivan 51, Chesley A, Hebert C, McFaull S, Scullion D: The validity and reliability of hand-held dynamometry in assessing isometric external rotation performance. I Orthop Sports Phys Ther 10: , Thompson MC, Shingleton LC, Kegerreis ST: Comparison of values generated during testing of the knee using the Cybex I1 Plus and Biodex Model B isokinetic dynamometers. I Orchop Sports Phys Ther 11: , 1989 Volume 16 Number 5 November 1992 JOSPT