Reliability of the Ely s Test for Assessing Rectus Femoris Muscle Flexibility and Joint Range of Motion

Transcription

1 Reliability of the Ely s Test for Assessing Rectus Femoris Muscle Flexibility and Joint Range of Motion J. Peeler, 1 J.E. Anderson 2 1 Department of Kinesiology and Applied Health, University of Winnipeg, 515 Portage Avenue, Winnipeg, Manitoba, Canada R3B 2E9, 2 Department of Human Anatomy and Cell Science Faculty of Medicine, University of Manitoba, Winnipeg, Manitoba, Canada Received 2 October 2006; accepted 20 September 2007 Published online 9 January 2008 in Wiley InterScience ( DOI /jor ABSTRACT: Rehabilitative protocols and orthopadic research are significantly influenced by the ability to perform reliable measures of specific physical attributes or functions. The hypothesis was that the Ely s test for evaluating rectus femoris flexibility and joint range of motion (ROM) is a reliable clinical tool. Participants (n ¼ 54) were between the ages of 18 and 45, and had no history of trauma. Three clinicians with orthopedic expertise assessed quadriceps flexibility and joint ROM using pass/fail and goniometer scoring systems. A retest session was completed 7 to 10 days later. Statistically, Kappa values for pass/fail scoring (intrarater X ¼ 0.52, interrater X ¼ 0.46) and ICC values (intrarater X ¼ 0.69, interrater X ¼ 0.66) for goniometer data both indicated that the Ely s test demonstrated only moderate levels of intra- and interrater reliability. Measurement error values (SEM ¼ 48,ME¼48, and CV ¼ 3%) and Bland and Altman plots (with 95% Limits of Agreement) further demonstrated the degree of intrarater variance for each examiner when executing the Ely s test in a clinical setting. Results call into question the statistical reliability of the Ely s test, and provide clinicians with important information regarding the reliability limits of the Ely s test when used to clinically evaluate flexibility and joint ROM in a physically active population. More research is required to determine the variables that may confound statistical reliability of this orthopedic technique that is commonly used in a clinical setting to assess function about the thigh region. ß 2008 Orthopaedic Research Society. Published by Wiley Periodicals, Inc. J Orthop Res 26: , 2008 Keywords: special tests; orthopedic evaluation; flexibility measurement In orthopedics, effective treatment and useful research are both dependent on the extent to which clinicians can perform reliable and accurate measures of a specific physical attribute or function. 1,2 Unreliable or inaccurate assessment confounds the use of a hypothesisdriven research model, compromises the clinician s ability to make informed decisions regarding treatment progression, and therefore complicates the effective prescription of treatment protocols. 1,3 Reliability, or consistency, refers to the extent that a measurement is reproducible and free of error, and assesses whether measurements are repeatable when all conditions are thought to be held constant. 1,4 A reliable examiner will be able to make repeated assessments as evidenced by consistent scoring. Examiner reliability can be conceptualized as either intrarater (or within examiner) or interrater (or between examiners) reliability. Intrarater reliability refers to the reproducibility of the measurements by the same examiner (i.e., the consistency with which one patient or subject is assessed by the same examiner over multiple examinations), and interrater reliability refers to the reproducibility of measurements taken by different examiners (i.e., the consistency with which one patient or subject is assessed by multiple examiners). 1,2,4,5 Goniometric assessment is a routine procedure used by clinicians to evaluate joint range of motion (ROM) in a real world clinical orthopedic setting. It allows the quantification of movement using linear (inches or centimeter) or angular units (degrees of an arc). Because of its widespread use in orthopedics, the reliability of goniometric assessment (continuous data) has been Correspondence to: J. Peeler (T: ; F: ; j.peeler@uwinnipeg.ca) ß 2008 Orthopaedic Research Society. Published by Wiley Periodicals, Inc. rigorously investigated. 6 9 Clinicians also utilize pass/ fail (or negative/positive) scoring systems (dichotomous data) to assess ROM about a particular joint. These orthopedic tests (sometimes referred to as special tests ) help to determine whether a particular type of dysfunction (i.e., impingement syndrome, agonist/antagonist imbalance in muscle strength or flexibility, etc.) or injury (i.e., muscle strain, ligament sprain, etc.) may be present. 10 Detailed procedures and established benchmarks are used to assess motion as either a passing score (when a ROM meets or exceeds a specified angle) or a failure (when a ROM fails to meet the specified angle). Previous research has shown that many special tests lack sensitivity, in that a failing score is suggestive of dysfunction, while a passing score does not necessarily rule out or exclude dysfunction. 11 In clinical orthopedics, the Ely s test is commonly used by practitioners to assess rectus femoris muscle flexibility and joint ROM. The face validity of this assessment technique is confirmed by its inclusion in several prominent textbooks on orthopedic physical assessment, 10,12,13 and its use as measurement tool in research examining function about the thigh region Unfortunately, few studies report rectus femoris muscle flexibility and joint ROM data that were collected using Ely s evaluation criteria with established reliability limits, and much of the research is population specific (i.e., in one type of athlete, sport, or disease). As a result, a reliability measurement for the Ely s test is not well established. HYPOTHESES AND SPECIFIC AIMS The purpose of this investigation was to test the hypothesis that the Ely s test provides reliable assessment of rectus femoris muscle flexibility and joint ROM. Specifically, the study had the following aims: (1) to investigate the intrarater reliability of the Ely s test; 793

2 794 PEELER AND ANDERSON (2) to investigate the interrater reliability of the Ely s test; and (3) to compare the reliability of goniometer (continuous data) and pass/fail (dichotomous data) scoring for the Ely s test. MATERIALS AND METHODS Data Collection Protocol Following approval by the Research Ethics Board at the University of Manitoba, healthy, physically active subjects between the ages of 18 and 45 years of age with no history of surgery or trauma to the hip, knee, or lower leg region were recruited for the study. All subjects took part in an initial intake session prior to beginning the study. During this session, subjects were assigned an identification (I.D.) number, and asked to complete informed consent and participant information forms. They also completed a physical activity questionnaire to provide baseline information regarding their habitual activity patterns at work, leisure and play 18 ; these data were used to confirm the sample was representative of a normal physically active population. Baseline anthropometric data such as height (m) and weight (kg) were also measured and recorded by a clinician with more than 13 years of clinical assessment experience in orthopedics. Subjects were instructed to refrain from starting new activities or exercise regimes during the course of the study. At the conclusion of the intake session, subjects were scheduled for two separate sessions (both occurring either over the lunch hour period or during the early evening) for assessment of rectus femoris muscle flexibility, which occurred 7 to 10 days apart. 19 Three experienced examiners with orthopedic assessment expertise were recruited from the community to participate in the study. All were Board-Certified Athletic Therapists, and possessed on average more than 12 years (experience ranged from 6 to 22 years) of clinical assessment and treatment experience in musculoskeletal disorders, and who routinely use the Ely s test to evaluate function about the thigh region. Prior to the start of the study, each examiner attended two 1-h instructional workshops in order to clarify details of the assessment protocol, and to reinforce the specific criteria used to define pass/fail scoring on the Ely s test, and the standardized procedures for collecting goniometric data. Assessments took place in the Rehabilitation Exercise Laboratory located in the School of Medical Rehabilitation at the University of Manitoba. Testing was conducted over the lunch hour, or during the early evening, in a standardized testing environment [i.e., consistent room temperature (208C), lighting, privacy, and plinth type]. Subjects were instructed to wear shorts and T-shirts for all assessments. Beyond this, participants were instructed to refrain from exercise a minimum of 4 h prior to each testing session, and to avoid starting/ initiating new sporting/training activities during their 7 to 10 days of participation in the study. Assessment by each examiner took approximately 5 min to complete; a maximum of four subjects were tested per half hour. Subjects underwent independent assessment by each of the three examiners in a random order, with the same order being used in the retest, as the initial test. Examiners assessed bilateral rectus femoris muscle flexibility and joint ROM of each participant using the Ely s Test (Fig. 1). Examiners determined pass/fail and goniometer scoring according to the protocol outlined in Magee s Orthopedic Physical Assessment textbook. 10 Pilot testing demonstrated that this goniometer measurement procedure required a subtle modification in order to be effectively executed by a single Figure 1. Ely s Test: visual representation of pass/fail grading. (A) Pass score: participant s hip remains stationary and on the plinth when performing active knee flexion of the test leg; (B) participant s hip spontaneously flexes off of the plinth when performing active knee flexion of the test leg; (C) visual representation of goniometer scoring. An adhesive marker was placed over the head of the fibula, and an 18-inch, semi-rigid plastic goniometer (centered over head of fibula) was used to measure the angle of active knee flexion of the test leg, while the participant maintained the anterior pelvis in contact with the plinth. examiner. The measurement procedure was modified to incorporate active knee flexion (rather than passive knee flexion) into the testing protocol in order to facilitate goniometer measurement by only one examiner. For pass/fail scoring, participants were instructed to actively flex one knee and bring their heel towards the buttocks, while maintaining a neutral pelvic posture and contact of the anterior aspect of the hip with the plinth. The test was scored as a pass if the hip was observed to remain in a stationary position against the examination table during active knee flexion. The test was scored as a fail, if during active flexion of the knee, the hip on the same side was observed to spontaneously flex, bringing the anterior hip off the table. 10 Knee joint ROM of the test leg was quantified using an 18- inch flexible and adjustable plastic goniometer (Baseline TM, Diagnostic and Measuring Instruments) that is commonly employed by healthcare practitioners working in a clinical setting. 8 All measurements were made according to visibly identifiable anatomical landmarks, and avoided procedures that would require examiners to estimate the exact center of rotation about which the knee joint moves. Pilot testing demonstrated that the greater trochanter of the femur (hip), head of the fibula (knee), and lateral malleolus of the fibula (ankle) were the most readily identifiable landmarks of the region. The easy availability of these points facilitated efficient,

3 ELY S TEST RELIABILITY 795 accurate, and reliable surface landmarking, and helped to minimize the confounding effect of inconsistent surface landmarking on the part of examiners. 20,21 Goniometer measurement procedures were standardized: each examiner placed an adhesive marker (1.5 cm in diameter) over the head of the fibula, and established the superior (greater trochanter of the femur) and inferior (lateral malleolus of the fibula) landmarks prior to each measurement. Subjects were instructed to maintain a neutral pelvic posture and contact of the anterior aspect of the hip with the plinth, while actively flexing one knee and bringing their heel towards the buttocks. The degree of knee flexion in the test leg was measured about an axis of rotation running through the head of the fibula, while the two arms of the goniometer were aligned with superior (greater trochanter of the femur) and inferior (lateral malleolus of the fibula) landmarks. Ely s test scores (goniometer measurement to the nearest degree and a pass/fail score) were recorded for each subject by each examiner on a standardized data collection sheet for each test session. Examiners were blinded as to their scoring from the first test session, and to the scoring by other examiners. At the end of each test session, data sheets were collected and collated according to subject I.D. numbers. Data Analysis Data were entered in a Microsoft 1 Excel spreadsheet. Descriptive statistics (mean SD) were generated for age, body weight, and height measurements, calculated body mass index (BMI), physical activity levels (scored out of a total of 15), and knee joint range of motion. Intraclass correlation coefficients (ICC) were calculated in order to evaluate the intra- and interrater reliability of goniometer scoring. An ICC (3,1) model was used to evaluate the intrarater reliability. ICC values were calculated using a two-way ANOVA and the equation: ICC ¼ (BMS EMS)/[BMS þ (K 1) EMS], where BMS is the between-subjects mean score, EMS is the error mean score, and K is the number of raters. 1,22 24 An ICC (2, 1) model was used to evaluate the interrater reliability. ICC values were again calculated using a two-way ANOVA and the equation: ICC ¼ (BMS EMS)/[BMS þ (K 1) EMS] þ [K (RMS EMS)/n], where BMS is the between-subjects mean score, EMS is the error mean score, RMS is the between-raters mean score, K is the number of raters, and n is the number of subjects tested. 1,22 24 Intra and interrater reliability of pass/fail scoring was measured using a Kappa statistic. This statistic uses a simple index of agreement, called percentage agreement, to measure how often raters agree on scoring for each individual subject. The advantage of the Kappa statistic is that it examines the proportion of observed agreement, and also considers the proportion of agreement that might be expected by chance. Therefore, the coefficient of agreement (proportion of observations on which there is agreement divided by the number of pairs of scores that were obtained) produced by the Kappa test is corrected for chance (number of expected agreements divided by number of possible agreements). This calculation provides a reasonable estimate of the reliability of dichotomous pass/fail data. 1,25 As cited by several clinical research publications, ICC and Kappa values above 0.75 should be considered representative of high levels of reliability, while values between 0.4 and 0.75 are indicative of a fair-to-moderate level of reliability. ICC values below 0.4 should be considered representative of a poor level of reliability. 1,3,22,23,26 Three forms of measurement error statistics [standard error of the measurement (SEM), method error (ME), and coefficient of variation (CV)] were used to examine the within-subject variation between testing sessions. The standard error of the measurement was defined by SEM ¼ SD 1 (1 ICC) 0.5,whereSD 1 is the standard deviation of all measurements, and the ICC value is derived from intrarater analysis. Method error was defined as ME ¼ SD 2 /H2, where SD 2 is the standard deviation of the differences between the two measurements. The coefficient of variation was defined as CV ¼ 100 ME/X 1,whereX 1 is the mean for all observations from test sessions 1 and Finally, Bland and Altman graphs (with accompanying 95% Limits of Agreement) provide a visual representation of the range of scoring by each of the examiners over the two testing sessions, and depict the difference score between test and retest plotted against the mean scores for each subject. The 95% upper and lower limits of agreement represent two standard deviations above and below the mean difference score. 27 RESULTS Descriptive statistics for study participants are presented in Table 1. Participants had a mean age of 29 years, and were representative of a population that is young, healthy, and physically active in a wide variety of leisure and sporting opportunities. Fifty-seven (57) subjects volunteered to participate in the study over a 6-month period. Fifty-four (54) subjects completed both testing sessions. For analysis, the flexibility measurements from 108 limbs were used to investigate intrarater reliability, while 222 flexibility measurements were available to examine interrater reliability. Descriptive statistics for knee joint range of motion are presented in Table 2. The mean knee joint range of motion for all participants was On average, there was little gender difference when comparing the ROM about the knee joint during Ely s testing. Intraclass correlation coefficients and a chance corrected Kappa statistic were used to evaluate the relative reliability of intra- and interrater scoring for the Ely s test. Intrarater results are presented in Table 3. Pass/fail corrected Kappa values ranged from a low of 0.46 to a high of 0.62 among the three examiners. Goniometer ICC values ranged from a low of 0.50 to a high of 0.83 among the three examiners. The intrarater results demonstrated that on average, the goniometer method of scoring was more consistent than the pass/fail method of scoring. As well, intrarater results revealed that examiner #3 was generally the most reliable in scoring knee joint ROM during the test retest protocol. Interrater results are presented in Table 4. The goniometer ICC values were, on average, higher than the pass/fail corrected Kappa values. However, analysis Table 1. Participant Anthropometric Data (Mean Standard Deviation) for the Present Study Parameter Total (n ¼ 57) Age (years) Weight (kg) Height (m) Body mass index Physical activity levels (/15)

4 796 PEELER AND ANDERSON Table 2. Goniometer Scoring of Knee Joint Range of Motion Male (34 Limbs) Female (74 Limbs) All (108 Limbs) Gender Test Retest Test Retest Test Retest Examiner Examiner Examiner Group Average Knee joint range of motion data (mean standard deviation) for Ely s Testing obtained through goniometer measurements (in degrees) for the present study. The group average is representative of the mean score of all examiners across both assessments (test and retest). of between-examiner scores using a two-way ANOVA revealed significant variation (p < 0.01) in goniometer scoring among the three examiners. Measurement error was analyzed using the goniometer data from the three examiners, and is presented in Table 5. The SEM can be used to determine the minimum difference (MD) that is required to assume that a real difference exists between test/retest scores. The MD can be calculated by the formula: SEM 1.96 H2, and allows any change in a subjects score to be considered real, if it is either above or below the previous score by greater than the MD value. The mean MD for the Ely s test among examiners was H2 ¼ 118. Bland and Altman plots with accompanying 95% Limits of Agreement (Fig. 2A C) illustrate the wide range in each examiner s scoring over the two test sessions, and indicate that the difference between test/ retest scoring can be expected to vary between (average of three examiners), or between 9.78 and Beyond this, Figure 2A C also depicts that examiners were unbiased in scoring over the two testing sessions (i.e., a higher or lower score was just as likely to occur in test session #1 as test session #2). Additionally, Figure 3 provides a visual representation of the large variation (p < 0.01) between-examiner s scoring for participants. It depicts a large number of outlying data points for each examiner, and is indicative of systematic examiner-dependent use of the Ely s test. DISCUSSION This study was conducted to examine the reliability of an orthopedic assessment technique that is commonly used in a clinical setting to assess rectus femoris muscle flexibility. The results show that the Ely s test demonstrated only moderate levels of statistical reliability for Table 3. Ely s Test Chance Corrected Kappa Statistics for Pass/Fail Scoring, and ICC (Model 3, 1) Values for Goniometer Scoring During the Present Study Intrarater Pass/Fail Goniometer Examiner 1 (n ¼ 108) Examiner 2 (n 108) Examiner 3 (n 108) Mean intra- and interrater comparisons among examiners during both pass/fail and goniometer scoring. It would appear that despite the use of well-defined methodology and examiner workshops that were designed to standardize the assessment protocol and clearly define pass/ fail criteria, each of the examiners used slightly different stringency (i.e., specified ROM) when grading rectus femoris muscle flexibility as either a pass or fail. Beyond this, despite the use of a readily identifiable anatomical landmark, it would appear that inaccurate or inconsistent surface landmarking during goniometer evaluation of joint ROM may have contributed to the large amount of variation between examiners scores for each participant. Because goniometric assessment evaluated joint ROM to within 18, small measurement differences may have also resulted in an overemphasis of the variation between examiners scores. Measurement error values for the goniometer data indicated that there was a large range (95% Limits of Agreement illustrated that scores could range by as much as 208) for each examiner s scoring over the two testing sessions. Beyond this, the SEM values also depict that the minimum difference (MD) required to assume that a real difference exists between an examiner s test/retest scores for the Ely s test was approximately 118. This information provides valuable insight into the clinical reliability limits of the Ely s test, and enables practitioners to make knowledgeable decisions regarding whether a real change has occurred between testing sessions, or whether the observed change is simply a product of measurement error. While this research project provided invaluable information on the reliability limits of the Ely s test when used in a clinical setting, it is important to acknowledge that participant variation both within and between assessment sessions (i.e., three consecutive Ely s test assessments conducted during each testing session; participant activities the day/week of assessment; time interval between testing sessions, etc.), as well as procedural modifications when executing the Ely s test, may have adversely affected the reliability scores for this technique. Specifically, it is important to acknowledge that a subject s active range of motion may not have remained the same upon repeated measurements in the same test session. Unfortunately,

5 ELY S TEST RELIABILITY 797 Table 4. Ely s Test Chance Corrected Kappa Statistics for Pass/Fail Scoring and ICC (Model 2, 1) Values for Goniometer Scoring During the Present Study Interrater Examiner 1 Examiner 2 Mean Pass/Fail Examiner 2 (n ¼ 222) Examiner 3 (n ¼ 222) Goniometer Examiner 2 (n ¼ 222) 0.54** 0.66 Examiner 3 (n ¼ 222) 0.77** 0.66** **p < this change in motion may or may not have been the same from examiner to examiner, and it is unclear what the degree of this variation was (particularly with regard to a known gold standard of bony orientation). It could be that this variation was different (e.g., less) when the limit of flexion was established by an examiner (e.g., a controlled flexion moment applied by an examiner may produce different findings in comparison to a flexion moment applied by a subject).therefore, the study s underlying assumption that the actual limit of knee flexion was exactly the same at each measurement time point may have resulted in an underestimation of between and within test variation. In order to limit or study the affect of these confounding variables, alternate methodological approaches should be investigated. The Ely s testing procedure could be done passively using two examiners, or could be further modified to incorporate a method for standardizing patient positioning during assessment (i.e., standardizing the position of the patients hips and pelvis by using a belt that straps the hips to the table and prevents horizontal and longitudinal movements of the pelvis), thereby limiting an examiner s ability to deviate from the established assessment protocol. As well, scoring of joint ROM and rectus femoris flexibility could be done from digital photos or film in order to minimize participant variation both within and between assessment sessions. These changes would serve to further standardize the data collection protocol, and thereby limit the number of variables that may confound rater reliability. The results of such a study would help to clarify the results of the present study and serve as a valuable comparison of the reliability differences between hands-on and secondary assessment of muscle flexibility and joint ROM. Having said this, the current findings do have important implications for the education, application, and evaluation of clinical skills within a real world orthopedic setting. The data indicate that even experienced examiners who possess advanced orthopedic assessment skills had difficulty attaining a high level of reliability when utilizing the Ely s test to assess rectus Table 5. Measurement Error Scores by Examiner for Goniometer Scoring During the Present Study Intrarater SEM8 ME8 CV% Examiner Examiner Examiner Mean Figure 2. Bland & Altman graphs (measured in degrees) depicting differences in goniometer scoring between test and retest, plotted against mean scores for each subject. Dashed line shows the mean difference score, solid line represents the 95% upper and lower limits (two standard deviations above and below mean difference score). (A) Examiner 1: mean SD ¼ , upper limit ¼ 10.9, lower limit ¼ 8.0; (B) Examiner 2: mean SD ¼ , upper limit ¼ 11.3, lower limit ¼ 12.2; and (C) Examiner 3: mean SD ¼ , upper limit ¼ 11.5, lower limit ¼ 9.0.

6 798 PEELER AND ANDERSON Figure 3. Bland & Altman graphs (measured in degrees) illustrate there was a large amount of interrater variation over the two test sessions. femoris muscle flexibility and joint ROM (for both pass/fail and goniometer scoring methods). Beyond this, the study sample was representative of a population that was young, healthy, and physically active. From a clinical standpoint, this type of patient would be hypothesized to provide the most accurate and consistent model for examining assessment techniques because it would limit confounding factors such as joint pathology, muscle contractures, and elevated BMI. If this notion is true, then one would predict that reliability values of the Ely s test would be even lower when examining a sedentary population or one demonstrating joint pathology. This point warrants consideration by clinicians evaluating rectus femoris muscle flexibility and joint ROM in relation to a specific pathology, and in recording day-to-day progress in rehabilitation programs designed to enhance flexibility and function about the thigh region. CONCLUSION The results of this study provide important information to practitioners regarding the limits of reliability for an orthopedic assessment technique (Ely s test) that is commonly used in a clinical setting. To our knowledge, the reliability of this special test has not been previously reported using a large normative population sample size within the scientific literature. Statistically, the data call into question the reliability of the technique when used to score rectus femoris muscle flexibility and joint ROM using both goniometer and pass/fail scoring methods. This means one measure may not be reproduced precisely on a second assessment, or during assessment by another clinician. Clinically, the results also serve as a guide for practitioners when evaluating and deciding whether a change observed between testing sessions is real, or simply a product of measurement error. Beyond this, the methodology employed for this study serves as a template to guide the evaluation or development of other clinically reliable musculoskeletal assessment techniques for the lower extremity. It should also assist in educating practitioners about evidence-based application and evaluation of clinical assessment skills used in the orthopedic and rehabilitative sciences. REFERENCES 1. Portney LG, Watkins MP Foundations of clinical research - applications to practice. Upper Saddle River, NJ: Prentice Hall Health. 2. Weir JP Quantifying test-retest reliability using the intraclass correlation coefficient and the SEM. J Strength Cond Res 19: Atkinson G, Nevill AM Statistical methods for assessing measurement error (reliability) in variables relevant to sports medicine. Sports Med 26: Bedard M, Martin NJ, Krueger P, et al Assessing reproducibility of data obtained with instruments based on continuous measurements. Exp Aging Res 26: Vela L, Tourville TW, Hertel J Physical examination of acutely injured ankles: an evidence based approach. Athletic Therapy Today 8: Boone DC, Azen SP, Chun-Mei L, et al Reliability of goniometric measurements. Phys Ther 58: Low JL The reliability of joint measurements. Physiotherapy 62: Rothstein JM, Miller PJ, Roettger RF Goniometric reliability in a clinical setting: elbow and knee measurements. Phys Ther 63: Somers DL, Hanson JA, Kedzierski CM, et al The influence of experience on the reliability of goniometric and visual measurement of the forefoot position. J Orthop Sports Phys Ther 25: Magee DJ Orthopedic physical assessment. Philadelphia: Saunders. 11. Ross MD, Nordeen MH, Barido M Test-retest reliability of Patrick s hip range of motion test in health college-aged men. J Strength Cond Res 17: Prentice WE The thigh, hip, groin, and pelvis. 12 th ed. In: Turenne M, editor. Arnheim s principles of athletic training: a competency-based approach. New York: McGraw- Hill; p Reid DC Problems of the hip, pelvis, and sacroiliac joint. In: Sports injury assessment and rehabilitation. 1st ed. Philadelphia: Churchill Livingstone; p Feldman DE, Shrier I, Rossignol M, et al Adolescent growth is not associated with changes in flexibility. Clin J Sport Med 9: Fredericson M, Yoon K Physical examination of patellofemoral pain syndrome. Arch Phys Med Rehabil 85: Marks MC, Alexander J, Sutherland DH, et al Clinical utility of the Duncan-Ely test for rectus femoris dysfunction during the swing phase of gait. Dev Med Child Neurol 45: Witvrouw E, Bellemans J, Lysens R, et al Intrinsic risk factors for the development of patellar tendonitis in an athletic population. Am J Sports Med 29: Baecke JAH, Burema J, Frijters JER A short questionnaire for the measurement of habitual physical activity in epidemiological studies. Am J Clin Nutri 36: Marx RG, Menezes A, Horovitz L, et al A comparison of two time intervals for test-retest reliability of health status instruments. J Clin Epidemiol 56: France L, Nester C Effects of errors in the identification of anatomical landmarks on the accuracy of Q angle values. Clin Biomech 16: Szulc P, Lewandowski J, Marecki B Verification of selected anatomic landmarks used as reference points for universal goniometer positioning during knee joint mobility range measurements. Med Sci Monit 7: Domholt E Physical therapy research - principles and applications. Philadelphia: Saunders.

7 ELY S TEST RELIABILITY Holmback AM, Porter MM, Downham D, et al Reliability of isokinetic ankle dorsiflexor strength measurements in healthy young men and women. Scand J Rehabil Med 31: Holmback AM, Porter MM, Downham D, et al Ankle dorsiflexor muscle performance in healthy young men and women: reliability of eccentric peak torque and work measurements. J Rehabil Med 33: Haley SM, Osberg JS Kappa coefficient calculation using multiple ratings per subject: a special communication. Phys Ther 69: Shrout PE, Fleiss JL Intraclass correlations: uses in assessing rater reliability. Psychol Bull 86: Bland JM, Altman DG Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1: