Reliability and Validity of Goniometric Measurements at the Knee

PREM P. GOGIA, JAMES H. BRAATZ, STEVEN J. ROSE, and BARBARA J. NORTON

The purposes of this study were to assess 1) the intertester reliability of gonio­metric measurements at the knee and 2) the validity of the clinical measurements by comparing them to measurements taken from roentgenograms. Thirty healthy subjects between the ages of 20 and 60 years were studied. The subjects were positioned on their right side on a roentgenographic table with their left lower extremity on a stabilizing board that was elevated 15 cm above the table's surface. For standardization of the position, an assistant placed the posterior aspect of the subject's left thigh in contact with two 15-cm pegs, which had been inserted perpendicularly into the stabilizing board. The assistant then moved the left leg to achieve an arbitrary angle of the knee joint and held the limb in that position. Two physical therapists then independently used a standard plastic goniometer to measure the knee joint angle in the sagittal plane using the greater trochanter, the lateral condyle of the femur, the head of the fibula, and the lateral malleolus as bony landmarks. A roentgenogram was taken of the extremity before the subject was moved. Pearson product-moment correlation coefficients (r's) and intraclass correlation coefficients (ICCs) were used to analyze the data. The data analysis revealed that the intertester reliability (r = .98; ICC = .99) and validity (r = .97-.98; ICC = .98-.99) were high. The results of this study indicate that goniometric measurements of the knee joint are both reliable and valid.

Key Words: Knee joint, Physical therapy.

Goniometry is a technique commonly used in physical therapy for assessing the limitation of a patient's joint motion. The usefulness of goniometric measurements for providing objective assessments of a patient's initial status and progress depends on the reliability and validity of the measurements.

Reliability is the degree to which a measurement yields the same results when taken on at least two different occasions or by a minimum of two different examiners.1 Because pa­tients often are treated and reevaluated by different physical therapists, the intertester reliability of goniometric measure­ment techniques is essential.

Several studies about assessing the reliability of the gonio­metric measurements have been published.2-9 One study em­phasizes interdevice reliability,2 one is about the assessment

methods used,3 and others focus on intertester and intratester reliability.4"9

Hellebrandt et al, in 1949, were the first authors to report on the reliability of goniometric measurements.4 They tested patients with a variety of disabilities and investigated inter­tester reliability of shoulder, elbow, radioulnar, and wrist joints. They found that well-trained physical therapists could measure the range of motion of the specified joints with a high degree of reliability and concluded that "measurements made by different physical therapists may be used inter­changeably without destroying the reliability of clinical rec-ords"4(pp306-307) when intraindividual variability is known. They recommended that when interrater reliability is not known, different observers should not be used interchangea­bly to obtain measurements on the same patient.

Hamilton and Lachenbruch found significant variance among the seven therapists who measured the ROM of three joints of the index and middle fingers.2 They concluded that an individual therapist is capable of making accurate repeated observations to provide reliable information of joint function in the hand. Low compared estimates and measurements of elbow flexion and extension.5 In his first of two experiments, 50 observers estimated and then measured both joint motions. The mean error of the measurements was less than that of the estimates. In the second experiment, five testers measured elbow flexion on 10 separate occasions. Both intertester and intratester variations were smaller in the second experiment than in the first experiment. Low concluded that measuring joint motion with a goniometer is more reliable than estimat-

Mr. Gogia is Orthopaedic Physical Therapist, The Methodist Hospital, 6565 Fannin, Houston, TX 77030 (USA). At the time of this study, he was Associate Director of Physical Therapy, St. Mary's Hospital, 129 N Eighth St., East St. Louis, IL 62201.

Mr. Braatz is Director of Physical Therapy, St. Mary's Hospital. Dr. Rose is Director and Associate Professor, Program in Physical Therapy,

Washington University School of Medicine, and Co-Director, Department of Physical Therapy, Irene Walter Johnson Rehabilitation Institute, St. Louis, MO 63110.

Ms. Norton is Instructor and Coordinator, Applied Kinesiology Laboratory, Program in Physical Therapy, Washington University.

This study was completed in partial fulfillment of Mr. Gogia's Master of Health Sciences Degree in Physical Therapy, Washington University.

The results of this study were presented at the Sixty-First Annual Conference of the American Physical Therapy Association, New Orleans, LA, June 16-20, 1985.

This article was submitted July 10, 1985; was with the authors for revision 22 weeks; and was accepted May 1, 1986. Potential Conflict of Interest: 3.

192 PHYSICAL THERAPY

RESEARCH

ing the degree of joint angle. He also concluded that using the average of several goniometric measurements increases relia­bility. He recommended that one person should make all the measurements in a series because the intratester error was notably less than the intertester error.

Boone et al also found intratester error to be less than intertester error.6 In their study, four testers measured the movements of six different joints in both the upper and lower extremities of six healthy subjects. In each of the four meas­urement sessions, the movements of each joint were measured three times for each subject. They found greater reliability for three upper extremity motions than for the lower extremity motions.

In their study of 22 healthy subjects, Ekstrand et al found intratester error to be less than intertester error when meas­uring six motions of the lower extremity joints.7 Mayerson and Milano found high intertester reliability of goniometric measurements of the major joints of the upper and lower extremities.8 Rothstein et al found high intertester reliability in the clinical setting for flexion and extension of the elbow and for knee flexion, but poor intertester reliability for knee extension.9

Validity is the degree to which an instrument measures what it is purported to measure and the extent to which it fulfills its purpose.10 Studies of the validity of goniometric measurements are important because of their extensive use in physical therapy and because of the conclusions based on their use.

The validity of goniometric measurements has not been studied as extensively as reliability. To date, no studies have been reported that assess directly the validity of goniometric measurements that have been obtained using a "gold stand­ard" and that have been analyzed statistically. A few articles, however, have provided some information about the validity of goniometry.

Hellebrandt et al, in Part 3 of their series on the reliability of goniometry, discuss their approach to assessing validity.4

Although they were assisted by an experienced roentgenolo­gist, which is an acceptable approach for assessing validity, they did not actually compare the measurements obtained by the roentgenologist with those obtained by the other exam­iners. They, instead, compared the differences in the meas­urements recorded in the two trials by the other examiners and discussed the relative reliability of their data, but not the validity.

In their article on the reliability of goniometers, Hamilton and Lachenbruch reported using roentgenograms to assess the ability of their specially designed positioning device to place the hand in a "reproducibly static position."2 Although they used roentgenograms, a potential gold standard, they did not report using statistical analysis to assess the validity of their measurements. They also did not compare the measure­ments taken from the roentgenograms to those obtained by the examiners.

Finally, in an article comparing several methods for meas­uring hip flexion, Ahlback and Lindahl reported on measure­ments taken both clinically and from roentgenograms.11 Al­though they did not report any statistical analysis of the data, they concluded on the basis of differences between pairs of measurements that close agreement existed between the meas­urements taken with their clinical technique and those ob­tained from the roentgenograms.

In summary, the consensus of these investigators is that, although exceptions exist, the level of reliability in goniomet­

ric measurement generally is acceptable. Their reports, how­ever, contain minimal information about the validity of the goniometric measurements.

The purposes of this study were to assess 1) the intertester reliability and 2) the validity of goniometric measurements at the knee joint. We chose to study the knee joint because it could be stabilized easily on a roentgenographic table.

METHOD

Subjects Thirty volunteers (13 women and 17 men) between the

ages of 20 and 60 years ( = 3 5 years) participated in the study. None of the subjects had a history of knee injuries. The purposes of the study and the procedure to be used were explained, and each subject signed an informed consent form.

Materials

A 60- × 90-cm piece of plywood, similar to a powder board, was used to support the left lower extremity. The board was supported by two 15-cm spacers. Two 15-cm pegs were in­serted perpendicularly into the board to stabilize the thigh posteriorly (Fig. 1). One large standard plastic goniometer with the scale marked in 1-degree increments and with 30-cm arms was used to measure the angular position of the knee joint, both on the subjects and on the roentgenograms.

Procedure

The subjects were positioned on their right side on a roent­genographic table with their right lower extremity resting on the table and their left lower extremity resting on the stabiliz­ing board. For standardization of the position, an assistant placed the posterior aspect of the subjects' left thigh in contact with the pegs projecting perpendicularly from the stabilizing board (Fig. 2). A physical therapist assistant then moved the left leg to achieve an arbitrary angle of the knee joint and held the limb in that position. The knee joint angle for all subjects was in the range of 0 to 120 degrees.

For each subject, two physical therapists (P.P.G., J.H.B.) independently and privately used a standard plastic goniom­eter to measure the knee joint angle in the sagittal plane and recorded the measurement on a piece of paper containing the subject's number. The knee joint measurements were ob­tained according to the procedure cited in the Manual of Joint Motion Measurements.12 The stationary arm of the goniom­eter was placed parallel to the long axis of the femur along a line extending from the greater trochanter to the lateral con­dyle, and the moveable arm was placed parallel to the long axis of the fibula in line with the head of the fibula and the lateral malleolus (Fig. 3).

Before the subject was moved, a roentgenogram of the knee joint was made by a radiology technician. The knee joint was held in the same fixed position for both measurements and for the roentgenogram. A roentgenologist drew lines on the roentgenogram parallel to the long axes of the femur and the fibula, then measured and recorded the angle formed by the intersection of the two lines (Fig. 4).

Data Analysis

Two statistical methods were used to calculate the reliability and validity of our goniometric measurements. The first method used was the Pearson product-moment correlation

Volume 67 / Number 2, February 1987 193

Fig. 1. Plywood stabilizing board for goniometric measurements.

Fig. 2. Left lower extremity positioned on the stabilizing board.

Fig. 3. Stabilized position of the left lower extremity during knee measurement.

Fig. 4. Measurement of joint angle using the roentgenogram.

coefficient (r), an assessment of covariance. The second method used was the intraclass correlation coefficient (ICC: 2,1) as identified by Bartko and Carpenter.13 The correlation coefficients were used to compare the measurements recorded by two therapists and to compare the measurements obtained by each therapist to those derived from the roentgenograms.

TABLE 1 Means and Standard Deviations of Measurements Obtained by Therapists and Roentgenologist (N = 30)

Source Therapist 1 Therapist 2 Roentgenologist

60.37 58.73 57.30

s 14.94 15.02 14.93

TABLE 2 Intertester Reliability and Validity Correlation Coefficients of Measurements Obtained at the Knee Joint (N = 30)

Comparison Therapist 1 vs therapist 2 Therapist 1 vs roentgenograms Therapist 2 vs roentgenograms

ra

.98

.97

.98

ICCb

.99

.98

.99

In addition, the means and standard deviations for all of the measurements were computed.

RESULTS

The joint angles used ranged from 0 to 120 degrees; their average mean and standard deviation were about 59 and 15 degrees, respectively (Tab. 1). The average correlation coeffi­cients for intertester reliability were .98 (r) and .99 (ICC). The correlation coefficients for validity ranged from .97 to .98 (r) and from .98 to .99 (ICC) (Tab. 2).

DISCUSSION

The findings of our study agree with those of previous investigators who also demonstrated good intertester reliabil­ity.4,8,9 Hellebrandt et al found that well-trained physical therapists could measure joint motion of the upper extremity with a high degree of reliability.4 Boone et al also found high intertester reliability for goniometric measurements of upper extremity joint motion; however, they reported a relatively poor intertester reliability for goniometric measurements of lower extremity joint motion.6 Rothstein et al found poor intertester reliability for goniometric measurements of knee extension, although they found high intertester reliability for goniometric measurements of elbow flexion and extension and knee flexion.9 Ahlback and Lindahl11 reported the raw data from their study for which we calculated a Pearson product-moment correlation coefficient of .82 and a Bartko and Carpenter13 ICC of .91.

Several potential sources of difficulty exist, however, in obtaining reliable measurements. Some investigators have suggested that certain joint motions might be more difficult to measure than others even though both the instrument and the examiner have been shown to be reliable.4,6 A particular patient may be difficult to assess for specific technical reasons14 or therapists may use different patient positions for measuring the same joints.9 The subjects' measurements also may fluctuate regardless of either the patient's level of coop­eration or the skills of the examiner.15 Both intratester and intertester reliability, therefore, should be assessed for all joints of interest. Future studies to assess the validity of goniometric measurements might include all of the joints, patients with

a r = Pearson product-moment correlation coefficient. b ICC = intraclass correlation coefficient.

194 PHYSICAL THERAPY

RESEARCH

various dysfunctions, a variety of clinical settings, and thera­pists with different levels of experience.

CONCLUSIONS

Both the reliability and the validity of our clinical measure­ment techniques are essential. In this study, we demonstrated

high intertester reliability and validity for goniometric meas­urements at the knee joint. Based on this study, we concluded that 1) goniometric measurements obtained by different ther­apists to assess the position of a patient's knee can be used interchangeably and 2) measurements recorded using the specified technique reflect the actual position of the knee joint.

REFERENCES

1. Design of Research and Analysis of Data in the Clinic: An Introductory Manual for Clinical Research. Washington, DC, American Physical Therapy Association, 1985

2. Hamilton GF, Lachenbruch PA: Reliability of goniometers in assessing finger joint angle. Phys Ther 49:465-469, 1969

3. Loessin Grohmann JE: Comparison of two methods of goniometry. Phys Ther 63:922-925, 1983

4. Hellebrandt FA, Duvall EN, Moore ML: The measurement of joint motion: Part 3. Reliability of goniometry. Phys Ther Rev 29:302-307, 1949

5. Low JL: The reliability of joint measurement. Physiotherapy 62:227-229, 1976.

6. Boone DC, Azen SP, Lin C-M, et al: Reliability of goniometric measure­ments. Phys Ther 58:1355-1360, 1978

7. Ekstrand J, Wiktorsson M, Oberg B, et al: Lower extremity goniometric measurements: A study to determine their reliability. Arch Phys Med Rehabil 63:171-175, 1982

8. Mayerson NH, Milano RA: Goniometric measurement reliability in physical medicine. Arch Phys Med Rehabil 65:92-94, 1984

9. Rothstein JM, Miller PJ, Roettger RF: Goniometric reliability in a clinical setting: Elbow and knee measurements. Phys Ther 63:1611-1615, 1983

10. Currier DP: Elements of Research in Physical Therapy, ed 2. Baltimore, MD, Williams & Wilkins, 1984, p 166

11. Ahlback SO, Lindahl O: Sagittal mobility of the hip joint. Acta Orthop Scand 34:310-322, 1964

12. Manual of Joint Motion Measurements. Washington, DC, US Government Printing Office, US Army Technical Manual 8-640/US Air Force Publication 160-14,1968

13. Bartko JJ, Carpenter WT: On the methods and theory of reliability. J Nerv Ment Dis 163:307-317, 1976

14. Iddings DM, Smith LK, Spencer WA: Muscle testing: Part 2. Reliability in clinical use. Phys Ther Rev 41:249-256, 1961

15. Cobe HM: The range of active motion at the wrist of white adults. J Bone Joint Surg 26:763-774, 1928

Volume 67 / Number 2, February 1987 195