Reliability of Lower Extremity Functional Performance Tests

Transcription

1 Reliability of Lower Extremity Functional Performance Tests Lori A. Bolgla, PT' Douglas R. Keskula, Ph D, PT, ATC2 ports medicine clinicians commonly see athletes who have sustained a lower extremity injury or surgical intervention. The ultimate goal of rehabilitation is to return the athlete to the highest functional level in the most efficient manner. Athletes must demonstrate normal range of motion, strength, proprioception, and coordination before safely returning to the playing field. Often, clinicians use impairments such as range of motion goniometric measurements, manual muscle testing, girth measurements, and isokinetic testing to assess when the athlete can safely resume unrestricted sporting activities. Additional information on functional performance would enhance the clinical decision-making process. Instruments used in assessing function include a subjective knee score questionnaire (10,15) and lower extremity functional performance testing. In an attempt to quantify "function," researchers have designed various functional performance tests which simulate the stresses about the knee encountered during athletic activities (9). Functional performance tests cannot detect specific abnormalities; however, they are useful in assessing lower limb function. According to Barber et a1 (2), their importance resulted from the fact that lower extremity function encompasses many variables, such as pain, swelling, crepitus, neuromuscular coordination, muscular strength, and joint stability. Furthermore, func- Clinicians routinely have used functional performance tests as an evaluation tool in deciding when an athlete can safely return to unrestricted sporting activities. These practitioners assumed that these tests provide a reliable measure of lower extremity performance; however, little research has been reported on the reliability of these measures. The purpose of this investigation was to determine the reliability of lower extremity functional performance tests. Five male and 15 female volunteers were evaluated using the single hop for distance, triple hop for distance, 6-m timed hop, and cross-over hop for distance as described by Noyes (10). One clinician measured each subject's performance using a standardized protocol and retested subjects in the same manner approximately 48 hours later. The order of testing was randomly determined. Subjects' average and individual scores on each functional performance test were used for statistical analysis. lntraclass correlation coefficients (ICCs) and standard error of measurement (SEM) values based on average day 1 and day 2 scores were used to estimate the reliability of each functional performance test. lntraclass correlation coefficients were.96,.95, and.96, and SEMs were 4.56 cm, cm, and cm, respectively, for the single hop, triple hop, and cross-over hop for distance tests. An ICC of.66 and SEM of. 13 seconds for the 6-m timed hop resulted from limited variability between measurements; however, its small SEM value inferred that the inconsistency of measurement would occur in an acceptably small range. A repeated measures analysis of variance revealed no significant difference (p >.05) between individual trial scores except for the single hop for distance. We concluded that this difference represented a learning effect not found with the other tests. The results of this investigation demonstrate that clinicians can use functional performance testing to obtain reliable measures of lower extremity performance when using a standardized protocol. Key Wonls: reliability, functional performance tests, lower extremity dysfunction ' Senior Physical Therapist, Center for Sports Medicine, Medical College of Georgia Hospital and Clinics, Augusta, GA At the time of this study, Ms. Bolgla was a staff physical therapist, Hitchcock Rehabilitation Center, Aiken, SC. Assistant Professor of Physical Therapy, Department of Physical Therapy, School of Allied Health Sciences, Medical College of Georgia, Augusta, GA The authors received no grant monb for this research investigation. tional performance tests are fast, simple to perform, require minimal staff training, and can be conducted in any clinical setting. Clinicians use functional performance testing as an assessment tool; therefore, the reliability of functional performance tests is essential for a p propriate data analysis. Reliability is the extent to which measurements are consistent, dependable, and free from error (5,12). Reliability also refers to the stability of measures with respect to time and evaluator; use of a reliable assessment tool means that variations between measurements are attributed to changes in the variable being measured (7). Measurement error greatly reduces reliability and may result from inconsistent measurements by individual clinicians, inaccuracies in the Volume 26 Nr~mber 3 September 1997 JOSPT

2 RESEARCH STUDY measuring instrument, or variability of the characteristic being measured. Use of a standardized protocol with exact instructions, clear operational definitions, and appropriate measuring instrument can minimize such measurement error (12). Functional performance tests cannot detect specific abnormalities. Many authors have reported on various functional performance tests (1,8,11,13). Although these researchers believed that functional testing can reasonably identify lower extremity limitations, they did not address the reliability of the measures. Conversely, Greenberger et al (6) developed three functional performance tests and also estimated the test-retest reliability of the measurements using intraclass correlation coefficients (ICCs). They reported ICCs ranging from.73 to.91. Booher et al (3) also examined the test-retest reliability of functional performance tests using a single-leg hop for distance, a single-leg 6m hop for time, and a 30-m single-leg agility hop with normal subjects (N = 18). They reported ICCs ranging from.77 to.97 when analyzing mean scores between test sessions; however, an analysis of variance (ANOVA) with repeated measures revealed significant differences between individual scores both within trials and between sessions. These researchers attributed the consistently improving scores to the practice effect and suggested that additional trials may stabilize this effect. Most clinicians have assumed that scores on functional performance tests provide a reliable measure of lower extremity performance; limited research exist. concerning the reli- JOSFT Volume 26 Number 3 September 1997 ability of lower extremity performance tests. The primary purpose of our study was to further investigate the test-retest reliability of lower extremity functional performance tests to confirm assumptions made by clinicians with respect to these tests. A secondary purpose of our study was to examine the effect that three practice trials would have on the stability of measures taken for actual test trials. METHODS Subjects Five males and 15 females (age = years, height = m, weight = kg) with no history of lower extremity dysfunction participated in this study. All subjects were informed of the procedures and inherent risks associated with the investigation and signed an informed consent approved by the institutional review board. The principal investigator interviewed each subject to ensure that no individual had a previous lower extremity injury or any other medical problem that would disqualify the person from safely participating in the investigation. Subjects performed all functional performance tests using their dominant limb. Limb dominance was determined by asking subjects which leg they would use to naturally kick a ball (3). Measuring Instrument Subjects performed four functional performance tests as described by Noyes et a1 (1 1): a one-legged single hop for distance (single hop), a one-legged 6-m timed hop (timed hop), a one-legged triple hop for distance (triple hop), and a one-legged cross-over hop for distance (cross over hop). The Figure shows an illustration of these tests. The single hop test was measured using a standard tape measure se- Snde Hap la DMurr Clos-orrr Hap la Dbbrrc mp* Hop la Dblmce 6 Meter Tlmd Hap FIGURE. The four functional performance tests modified from Noyes et a1 (101. cured to the floor. Each subject began the test by standing on the dominant limb with the toes lined at the tape measure's zero mark. The recorded measure was the distance from the zero mark to the place where the back of the subject's heel hit the ground upon completing the single hop on the dominant limb. The timed hop test had a 6m distance marked off. Subjects stood on the dominant limb with their toes lined at the starting point, began the test after the principal investigator said "1, 2, 3, Go," and ended the test when the back of their heel crossed the finish line. The test score was the time it took for the subject to hop 6 m on the dominant limb as measured to the nearest one-tenth of a second using a standard stop watch. The triple hop test was measured using a standard tape measure secured to the floor. Each subject began the test by standing on the dominant limb with the toes lined at the tape measure's zero mark. The recorded measure was the distance from the zero mark to the place where the back of the subject's heel hit the ground upon completing three consecutive hops on the dominant limb.

3 TABLE 1. Means and standard deviations for average day 1 and average day 2 scores on the iunctional pehrmance tests. "- Single hop (cm) Triple hop (cm) Timed hop (seconds) Crossimr hop (cm) SEM = Standard error oi measure. TABLE 2. Reliability estimates for the functional performance tests. The cross-over hop test had a line 6 m long and 15 cm wide secured to the floor. A tape measure was secured on top of the line. Sub jects began the test by standing on the right side of the line on the dominant limb. They hopped over to the left side, back over to the right side, and then back over to the left side using only the dominant limb. As done in the single and triple hop tests, each subject began the test with the toes lined at the tape measure's zero mark. The recorded measure was the distance from the zero mark to the place where the back of the subject's heel hit the ground upon completing the third hop. The test score was the total distance hopped. Procedure The test procedures consisted of a general warm-up, a task-specific warm-up, actual testing, and a cool down. The general warm-up required participants to ride a stationary bike at a steady, comfortable speed for 3 minutes followed by gentle quadriceps, hamstring, and calf stretches. Stretching involved three repetitions of each stretch using a 10-second hold. The principal investigator demonstrated each stretch and supervised all subjects during the warm-up period. The task-specific warm-up allowed subjects the opportunity to practice each functional performance test. First, the test administrator demonstrated each functional performance test. Participants then practiced each test three times in the following order: single hop, triple hop, timed hop, and cross-over hop. Participants rested approximately 30 seconds between each practice trial and rested 1 minute prior to actual testing. Actual testing had participants perform each functional performance test in a randomly determined order. Testing consisted of three consecutive trials for each functional performance test. Subjects were allowed a 30-second rest between each trial and a 45second rest between functional performance tests. Participants received no verbal encouragement during actual testing. Actual testing lasted approximately 15 minutes. A cooldown period followed actual testing. Subjects were verbally instructed to perform the gentle stretching as previously done during the general warm-up period. Participants were also told to return to the testing area approximately 48 hours later for retesting. Retesting was done following the same procedures used on the first day of testing. After the cooldown period, the principal investigator asked participants not to engage in any new physical activities during the 48hour period between test sessions; however, the investigators made no attempt to monitor sub jects' activities during this time. Analysis We compared average day 1 and day 2 scores for each functional performance test to determine the testretest reliability of the functional performance tests using intraclass correlation coefficients (ICCs) according to the Shrout and Fleiss (14) equation (2,k). The equation (2,k) was chosen because the trial scores were considered a random sample from a larger population and scores represented a composite value (the mean of three trials) (4). We also used the average day 1 and average day 2 scores for a repeated measures ANOVA in order to determine differences between test sessions for each functional performance test. Standard error of measure (SEM) was used to describe the precision of the measurement reported by the rater. The SEM possesses the unit of measure and calculates a range where the subject's true score is located. The SEM is equal to the standard deviation of the measurements multiplied by the square root of one minus the reliability coefficient (R) (4) - We also tested for significant differences of average scores between the six individual trials using a repeated measures ANOVA. The purpose of this analysis was to determine the stability of all measures used in this investigation. Level of significance was set at the.05 level for all statistical tests. RESULTS The means and standard deviations for average day 1 and average day 2 scores for each functional performance test are presented in Table 1. A repeated measures ANOVA of mean scores between days 1 and 2 for each functional performance test revealed no significant differences (p >.05). Intraclass correlation coefficients and SEMs for these mean scores are summarized in Table 2. Intraclass correlation coefficients Volume 26 Number 3 September 1997 JOSFT

4 TABLE 3. Means and standard deviations for the functional performance tests for trials 1-3 on day I. TABLE 4. Means and standard deviations for the functional performance tests for trials 4-6 on day 2. ranged from.66 to.96. Standard error of measure for the distance hop tests ranged from 4.56 cm to cm; SEM for the timed hop test was.13 seconds. The means and standard deviations for the repeated measures ANOVA for the average scores within the individual trials are shown in Tables 3 and 4. The repeated measures ANOVA showed no significant differences between individual trials for all functional performance tests except for the single hop test. The F values for each functional performance test are summarized in Table 5. Tukey post hoc testing indicated significant (p <.05) differences between the first and last trial of the single hop test. DISCUSSION The purpose of this investigation was to further explore the reliability S = S~jinificant. NS = Nonsignificant. TABLE 5. F values for the functional performance tests for repeated measures analysis of variance for average scores within six trials. of the functional performance tests described by Noyes et al (1 1). Our findings are consistent with estimates of reliability reported by other researchers. Booher et al (3) tested 18 normal subjects using a single-leg hop for distance, a single-leg 6-m hop for time, and a 30-m single-leg agility hop and reported ICCs ranging from.77 to.99. Greenberger et a1 (6) tested 33 normal subjects using a hip adduction excursion test, anterior lunge test, and balance leg reach test to assess functional performance and reported ICCs ranging from.73 to.91. Our study, reporting ICCs ranging from.95 to.96 for the onelegged hop for distance tests, suggested a high level of reliability for these tests. The 6-m timed hop had an ICC value of.66 and a SEM value of.13 seconds; its lower ICC value resulted from limited variability between measurements. A limitation of this study was that we used a manual stop watch and were unable to record changes not detectable by the human eye. Since small errors in manual timing could potentially result in large differences, future studies should use an electrical timing device to see if such differences exist. Assuming that such minute differences did not exist, the timed hop would still be a reliable measurement because a small SEM value inferred that the inconsistency of measurement would occur in an acceptably small range (4). Irrespective of the limitation concerning the timed hop test, we believe that the ICCs and SEMs reported in this study indicate that functional performance tests are a reliable measure of lower extremity performance when following a standardized protocol. Sports medicine clinicians routinely use functional performance tests in deciding when an individual can safely resume unrestricted sporting activities, and we believe that reliable measures of lower extremity performance depend on the use of a standardized protocol. The standardized protocol will ensure that clinicians set up the testing area properly and administer the functional performance tests in a consistent manner on any given test day. Other components of the standardized protocol include warm-up procedures, measuring techniques, the number of practice and test trials, the rest period within trials and between functional performance tests, and cooldown activities. Together, these factors help ensure that the scores on the functional performance tests reflect changes in lower extremity performance. Booher et al (3) reported significant differences of scores for trials within each test day and between each test day and suggested that future investigators perform additional trials to stabilize the individual scores. These researchers had subjects practice each test one time on each lower extremity, and testing consisted of one trial on each limb. The current study differed from the Booher et al study because our subjects practiced each test three times on the dominant limb and performed three trials of each functional performance test on the dominant limb for actual testing. We used the dominant limb only because previous JOSPT Volume 26 Number 3 September 1997

5 researchers reported no differences between dominant and nondominant limbs (3,6,11). The additional practice and test trials resulted in more stable measures for the triple hop, cross-over hop, and timed hop tests, as a repeated measures ANOVA revealed no statistical differences (P >.05) between average scores within the six trials. A repeated measures ANOVA indicated a statistical difference between average scores within trials for the single hop test, with the variance being attributable to a difference between the first and last trials. Comparison of scores from the second to the sixth trial revealed more stable These findings imply that sports medicine clinicians allow patients adequate practice trials before actual testing in order to account for the motor learning effect. scores, implying that the practice effect was not eliminated for the single hop test when using three practice trials. A possible reason for the significant difference for the single hop test was that subjects subjectively a p peared more confident performing this test than the other tests and seemed to hop more aggressively. These findings imply that sports medicine clinicians allow patients adequate practice trials before actual testing in order to account for the motor learning effect. We suggest that rehabilitation protocols include components of the functional perfor- mance tests to control for the learning effects. We also believe that patients should practice the tests on a day prior to actual testing. Finally, clinicians might consider more practice time for the single hop test because our study suggested that a greater learning effect occurred with this test. Although our investigation did not analyze the optimal number of practice trials needed to stabilize the learning effect, it inferred that three practice and three test trials resulted in scores deemed statistically the same within six trials for all functional performance tests except the single hop for distance. Future studies should be designed to investigate the optimal number of practice and test trials needed to account for the learning effect through the use of both a control and test group. CONCLUSION The purpose of this study was to investigate the reliability of lower extremity functional performance tests. We believe that these tests provide a reliable measure of lower extremity performance when following a standardized protocol and accounting for the learning effect through sufficient practice and test trials. By establishing the reliability of lower extremity functional performance tests, sports medicine clinicians can use them in better determining when an individual can safely return to unrestricted sporting activities. JOSPT REFERENCES Barber SD, Noyes FR, Mangine RE, De- Maio M: Rehabilitation after ACL reconstruction: Functional testing. Sports Med Rehabil Series 15(8): , 1992 Barber SD, Noyes FR, Mangine RE, Mc- Closkey JW, Hartman W: Quantitative assessment of functional limitations in normal and anterior cruciate limment- deficient knees. Clin Orthop 255:2O4-214, 1990 Booher LD, Hench KM, Worrell TW, Stikeleather J: Reliability of three singleleg hop tests. J Sports Rehabil 2: , 1993 Denegar C, Ball D: Assessing reliability and precision of measurement: An introduction to intraclass correlation and standard error of measurement. J Sports Rehabil2:35-42, 1993 Domholdt E: Physical Therapy Research, Philadelphia: W. B. Saunders Company, 1993 Greenberger HB, Sperling L, Frankford K: The rationale and reliability of three functional performance tests in the assessment of lower extremity function. Presented at the Combined Sections Meeting of the American Physical Therapy Association, Atlanta, GA, June, 1996 Keskula OR, Dowling IS, Davis VL, Finley PW, Dell'Omo DL: Interrater reliability of isokinetic measures of knee flexion and extension. J Athl Train 30(2): , 1995 Lephart SM, Perrin DH, Fu FH, Gieck jh, McCue FC, Irrgang JJ: Relationship between selected physical characteristics and functional capacity in the anterior cruciate ligament-insufficient athlete. J Orthop Sports Phys Ther 16(4): , 1992 Lephart SM, Perrin OH, Fu FH, Minger K: Functional performance tests for the anterior cruciate ligament insufficient athlete. Athl Train 26:44-50, Noyes FR, Barber SD, Mangine RE: Abnormal lower limb symmetry determined by functional hop tests after anterior cruciate ligament rupture. Am J Sports Med 1 9(5): , Noyes FR, Barber SD, Mooar LA: A rationale for assessing sports activity levels and limitations in knee disorders. Clin Orthop 246: , 1989 Portney LG, Watkins MP: Reliability. In: Foundations of Clinical Research. Applications to Practice, pp East Nonvalk, CT: Appleton & Lange, 1993 Risberg MA, Ekeland A: Assessment of functional tests after anterior cruciate ligament surgery. J Orthop Sports Phys Ther 1 9(4): , Shrout PE, Fleiss JL: lntraclass correlations: Uses in assessing rater reliability. Psych01 Bull 86: , 1979 Tegner Y, L ysholm J: Rating systems in the evaluation of knee ligament injuries. Clin Orthop 198:43-49, 1985 Volume 26 Number 3 September 1997 JOSPT