 |
 |
 |
|
| home | CME | help | feedback | subscriptions | archive |
|
© 1996 The Journal of Bone and Joint Surgery, Inc. Stress Radiographs of the Patellofemoral Joint* , WARREN , DETROIT, WARREN , DETROIT, MICHIGANInvestigation performed at the Division of Sports Medicine, Department of Orthopaedic Surgery, Hutzel Hospital, Wayne State University School of Medicine, Detroit, and Teitge Orthopaedic Associates, Warren
Patellar instability is usually diagnosed on the basis of the clinical presentation without radiographic confirmation. In the present report, we describe a new radiographic method to demonstrate patellar instability. Axial radiographs were made of the patellofemoral joint of ninety individuals (180 knees) and were then repeated while a medial or lateral force was applied to the patella. The applied force was kept constant with use of a specially designed instrument. The ninety individuals were divided into four groups on the basis of the clinical findings: normal, lateral instability, medial instability, and multidirectional instability. Stress radiographs differentiated the four groups and confirmed the clinical diagnosis in all patients who had unilateral symptoms. A four-millimeter increase in medial or lateral excursion of the patella of the symptomatic knee compared with the patellar excursion of the asymptomatic knee was significant (p < 0.0001). Stress radiographs offer a simple method for the measurement of force-displacement relationships in the patellofemoral joint and for the demonstration of patellofemoral instability.
Differentiation of the cause of pain in the anterior aspect of the knee continues to be an arduous challenge because of the multiple contributing factors, including malalignment of the axis of the limb, injury to the articular cartilage, weakness or contracture of the muscles, patellar or trochlear dysplasia, injury to the patellar ligaments, or overuse. Integration of these various factors into a comprehensive clinical picture is difficult because they frequently coexist, they are often interdependent, and methods to quantitate these abnormalities are not precise. Patellar subluxation is frequently associated with pain in the anterior aspect of the knee. A number of radiographic measurements have been shown to correlate with clinical patellofemoral instability (Fig. 1), and these are often used to verify the diagnosis23. Merchant et al.22 described the so-called congruence angle and found that asymptomatic knees had a congruence angle of -6 degrees and that any knee with an angle that was greater than 16 degrees had a 95 per cent chance of being abnormal. Brattström2 found that a sulcus angle of 142 degrees was normal and that larger angles were associated with patellar dislocation. The lateral patellofemoral angle of Laurin et al. opened laterally in ninety-seven of 100 normal patients (194 of 200 knees) but opened medially or was 0 degrees in all patients who had recurrent subluxation of the patella19. The lateral-to-medial condylar height ratio of Brattström2 averaged 1.65 in the normal knees, with a tendency to smaller ratios in the knees that had recurrent patellar dislocation. Finally, the patellar height index of Insall and Salvati9 normally is equal to 1.0, and an increase of 20 per cent is indicative of patella alta and a potential for instability.
Patellofemoral stability results from congruity of the articular surfaces and ligamentous restraint. A shallow femoral trochlea or a high-riding patella that does not seat in the femoral trochlea results in a less stable joint. However, a patellofemoral joint with a stable osseous environment may still subluxate because of weak ligaments or when subjected to excessive translational forces. Unstable joints are often congruous at rest, but stress may cause an abnormal displacement. The physical examination of a joint for instability should include demonstration of the displacement of the components of the joint when stress is applied. Radiographs made while stress is applied have been used to demonstrate instability of other joints1,3,5,10-16,18,20-29,31,35-37,39 but not of the patella. In the present report, we describe a technique for demonstrating patellar displacement on stress radiographs and compare this method of determining patellar instability with previously described radiographic methods.
A history was recorded, a physical examination was performed, and plain radiographs and stress radiographs were made for ninety individuals (thirty-five male and fifty-five female) (180 knees). The individuals ranged in age from thirteen to fifty-two years, with a mean age of twenty-six years. Twenty of the individuals were asymptomatic volunteers who had no signs or symptoms referable to the knees; they served as the normal control group. The remaining seventy individuals had been seen by the senior one of us (R. A. T.) for unilateral symptoms of patellofemoral instability. All patients who had a clear-cut clinical diagnosis of instability on the basis of commonly accepted signs and symptoms were included. The history included the onset and development of the symptoms, the mechanism of injury, and previous treatments. The patients were also asked if they had any history of limping, swelling, locking, giving-way, weakness, patellar sticking, patellar pain associated with sitting with the knee flexed or with ascending or descending stairs, abnormal patellar tracking, a feeling of insecurity related to the knee, or any sense of the patella coming out of place. A detailed physical examination was carried out with special attention to the physical signs associated with patellofemoral instability, including a clinical estimate of the alignment of the limb in all planes, inward pointing of the patella with the feet pointing forward, the alignment of the heel with the tibia, planovalgus feet, the range of motion of the knee, the ability to squat, the ability to extend the knee with the patient in the seated position, patellofemoral crepitation, a J sign (lateral deviation of the patella at terminal extension) or other abnormal tracking patterns, the sideways mobility of the patella both in extension and with the knee flexed at 20 degrees, the presence or absence of apprehension6, pain or crepitation when a stress was applied from the side of the patella, the comparative width of the two patellae, the Q angle, tenderness in the retinaculum or facets, the muscle tone of the quadriceps, the girth of the thigh, the stability of the knee, a tight iliotibial band or painful Ober test, the result of a prone rectus test for quadriceps contracture, internal and external rotation of the hip with the patient in the prone position with the hip extended, the foot-thigh axis, supination of the neutral forefoot on the hindfoot as an index of foot-pronation demand, the alignment of the tibia, and tightness of the Achilles tendon. Posteroanterior, lateral, and patellofemoral axial radiographs were made for all ninety subjects. The axial radiograph was made with the knee flexed 35 degrees. Stress radiographs of the patellofemoral joint were then made bilaterally for all subjects. The axial radiographs were first made with the knees flexed off the end of the x-ray table, with the legs resting on a platform that tilted down from the edge of the table at 35 degrees (Fig. 2). The actual degree of knee flexion was then measured with a goniometer; it ranged from 30 to 40 degrees. Since there is no substantial difference in the sulcus angle in a particular knee when flexion ranges between 30 and 40 degrees, this variation in actual flexion of the knee had no influence on the data reported here. The x-ray tube was located over the patient, and the cassette was placed anterior to the tibia and perpendicular to the x-ray beam. The angle of the beam was adjusted to obtain the widest image of the patellofemoral joint space at 35 degrees of knee flexion.
To make a radiograph with lateral stress applied, the examiner supports the lateral side of the knee with his or her hand to prevent rotation of the limb while pressure is applied to the medial side of the patella, displacing it laterally. The procedure is reversed for a radiograph made with medial stress applied (Fig. 2). To ensure that an equal pressure is applied to both of the knees, a spring-loaded scale with a curved rubber pad was developed by MedMetric (Patellar Pusher; MedMetric, San Diego, California). The radiograph is made while a constant pressure (usually sixteen pounds [7.3 kilograms]) is applied. Relaxation of the quadriceps must be maintained because any contraction forces the patella against the trochlea, reducing its freedom to displace. The displacement is measured on the radiograph with the technique described by Laurin et al.17. On the lateral stress radiograph, a line is drawn between the summits of the medial and lateral condyles and a perpendicular line is drawn anteriorly from the medial condyle tangent point (Fig. 1, F, G, and H). The distance between this line and the medial border of the patella is measured and compared with the measurement from the contralateral knee. Similar measurements are made for the medial stress radiograph but the perpendicular line is drawn anteriorly from the lateral femoral condyle. The comparison between the symptomatic and asymptomatic sides yields the difference in displacement and is a measure of the increased laxity or instability (Figs. 3-A, 3-B, and 3-C). A complete, unsuspected dislocation is sometimes seen and is obvious (Figs. 4-A, 4-B, 5-A, 5-B through 5-C).
A clinical diagnosis was established for each patient on the basis of the history and physical examination. A history of patellar dislocation or episodes of subluxation of the patella and patellar hypermobility associated with apprehension when the patella was manually displaced were considered the most important findings. Only patients who had unilateral symptoms in the knee were selected for this study. The subjects were separated into four groups. Group 1 (normal) consisted of the twenty volunteers who had no symptoms or signs related to the knee. Group 2 (lateral instability) consisted of twenty-seven patients who had a history and clinical findings consistent with a diagnosis of lateral patellar subluxation when they were first seen at the clinic. No patient in this group had had a previous operation, but all had had a twisting injury with subsequent persistent pain in the anterior aspect of the knee. On examination, all had increased lateral excursion of the patella with manual manipulation and all demonstrated apprehension when this maneuver was performed (a positive Fairbank sign6). Group 3 (medial instability) consisted of twenty-six patients who had persistent pain in the anterior aspect of the knee and instability, which had worsened after an isolated lateral retinacular release had been performed for the pain. No patient had a history that was consistent with preoperative dislocation or subluxation. Each patient had an increase in medial excursion of the patella with manual manipulation, and all patients had apprehension and pain with this maneuver, which reproduced the symptoms. None demonstrated increased lateral patellar excursion or a positive Fairbank sign when the patella was pushed laterally. Group 4 (multidirectional instability) consisted of seventeen patients who were seen for pain in the anterior aspect of the knee and instability, which had worsened following an isolated lateral retinacular release for previously diagnosed instability. On physical examination, all had increased excursion of the patella in both the medial and the lateral direction and all had apprehension and pain when the patella was displaced medially and laterally. Seven parameters were measured on the radiographs of each patient (Fig. 1). These seven parameters were the ratio of the length of the patellar ligament to the length of the patella, as measured on the lateral radiograph according to the method of Insall and Salvati9; the congruence angle, as defined by Merchant et al.22; the lateral patellofemoral angle, as described by Laurin et al.19; the sulcus angle, as defined by Brattström2; the lateral-to-medial condylar height ratio, as modified from the ratio described by Brattström; the lateral patellar displacement on the axial radiographs of both knees made with stress applied; and the medial patellar displacement on the axial radiographs of both knees made with stress applied. Statistical analysis was performed with use of the paired t test, sign test, Wilcoxon test, Pearson correlation coefficient, and Spearman rank correlation coefficient. The significance of the differences between the seven radiographic measurements from Group 1 and those from Groups 2, 3, and 4 was based on one-way analysis of variance followed by Tukey multiple comparison. By ranking the tests and analyzing differences in rank for non-parametric matched pairs of all groups, it was determined that a population of twenty normal individuals was of statistically sufficient size for a control group (p < 0.0005).
The mean Insall-Salvati ratio (and standard deviation) was 1.14 ± 0.13 for the normal individuals (Group 1), 1.26 ± 0.15 for the knees that had lateral instability (Group 2), 1.23 ± 0.15 for the knees that had medial instability (Group 3), and 1.30 ± 0.18 for the knees that had multidirectional instability (Group 4). Group 2 (p < 0.04) and Group 4 (p < 0.01) were significantly different from Group 1, indicating patella alta in those groups (Table I).
The mean congruence angle was -6.00 ± 11.56 degrees for Group 1, 6.41 ± 14.83 degrees for Group 2, -10.81 ± 10.80 degrees for Group 3, and -18.29 ± 15.10 degrees for Group 4. Groups 2 (p < 0.003) and 4 (p < 0.050) were significantly different from Group 1 but Group 3 was not (Table I). The negative congruence angle for Groups 3 and 4 suggests medial instability. The mean lateral patellofemoral angle was 9.15 ± 2.52 degrees for Group 1, 5.48 ± 5.96 degrees for Group 2, 9.92 ± 3.51 degrees for Group 3, and 8.24 ± 6.20 degrees for Group 4. The angle for Group 2 was significantly different from that for Group 1 (p < 0.043), but the angles for Groups 3 and 4 were not (Table I). All of the normal individuals and all of the patients who had medial instability had a normal patellofemoral angle. Despite the significant difference between Group 1 and Group 2, only six (22 per cent) of the twenty-seven patients who had lateral instability (Group 2) and three of the seventeen patients who had multidirectional instability (Group 4) had an abnormal patellofemoral angle, according to the definition of Laurin et al.19. The mean sulcus angle was 137.1 ± 4.38 degrees for Group 1, 143.2 ± 9.21 degrees for Group 2, 139.9 ± 5.24 degrees for Group 3, and 141.4 ± 5.55 degrees for Group 4 (Table I). The angle for Group 2 was significantly greater than that for Group 1 (p < 0.01). There was no significant difference between Group 1 and either Group 3 or Group 4. The mean lateral-to-medial condylar height ratio was 2.26 ± 0.53 for Group 1, 2.25 ± 1.07 for Group 2, 2.32 ± 0.67 for Group 3, and 2.22 ± 0.76 for Group 4. There was no significant difference among the four groups (Table I). Patellar displacement was measured on radiographs of both knees with stress applied in the medial and the lateral direction. Differences were then noted between the measurements for the symptomatic and the asymptomatic knee in both the lateral and the medial direction. The mean lateral displacement of the patella in Group 1 was 11.6 ± 8.18 millimeters on the side of greater displacement and 10.3 ± 8.18 millimeters on the side of less displacement. The mean lateral displacement on the symptomatic and asymptomatic sides, respectively, was 21.9 ± 6.20 and 14.4 ± 4.45 millimeters for Group 2, 13.2 ± 6.31 and 13.7 ± 6.44 millimeters for Group 3, and 20.9 ± 6.20 and 11.4 ± 6.06 millimeters for Group 4 (Table I and Fig. 6).
The mean difference in the lateral displacement between the two knees of each patient was 1.3 ± 1.10 millimeters for Group 1, 7.5 ± 3.43 millimeters for Group 2, -0.5 ± 1.56 millimeters for Group 3, and 9.5 ± 5.64 millimeters for Group 4. There was no significant difference (p = 0.18) in lateral patellar displacement between the two knees of the individuals in Groups 1 and 3. However, there was a significant increase in the lateral displacement of the patella in the symptomatic knee in Groups 2 and 4 (p < 0.0001 for both) (Table I and Fig. 7).
The mean medial displacement of the patella for Group 1 was 11.1 ± 4.21 millimeters on the side of greater displacement and 9.9 ± 4.25 millimeters on the side of less displacement. The mean medial displacement was 11.6 ± 3.61 millimeters on the symptomatic side compared with 11.8 ± 3.34 millimeters on the asymptomatic side for Group 2, 20.7 ± 6.07 millimeters compared with 10.4 ± 5.51 millimeters for Group 3, and 22.1 ± 7.73 millimeters compared with 11.5 ± 5.29 millimeters for Group 4 (Table I and Fig. 8).
The mean difference in the medial displacement between the two knees of each patient was 1.2 ± 1.08 millimeters for Group 1, -0.2 ± 1.43 millimeters for Group 2, 10.3 ± 5.12 millimeters for Group 3, and 10.6 ± 7.73 millimeters for Group 4. There were no significant differences in medial patellar displacement between the two knees of the individuals in Groups 1 and 2. There was a significant increase in the medial displacement of the patella in the symptomatic knee for Groups 3 and 4 (p < 0.0001 for both) (Table I and Fig. 9).
Patellofemoral instability may be caused by changes in the shape of the anterior aspect of the femur, the shape of the patella, the strength of the restraining ligaments, the tension in the quadriceps and patellar ligament, the direction at which this tension is applied, and the forces in the limb that would change the direction of pull. Because unstable joints are often congruous at rest, stress may be necessary for subluxation or dislocation to occur (Figs. 3-A, 3-B, 3-C, 4-A, through 4-B). The most reliable clinical diagnosis of patellofemoral instability is dependent on the patient's response to displacement with application of force (the Fairbank apprehension test6). A radiograph made with the application of stress demonstrates the degree of displacement, and a comparison with the asymptomatic knee reveals instability. We describe here a new technique to demonstrate lateral and medial instability, or both, of the patellofemoral joint when compared with the asymptomatic knee. Patients who had bilateral, symmetrical symptoms were not included in this study as it was noted that such patients often had bilateral instability evident on stress radiographs. One goal of the present study was to determine if stress radiographs made with the described technique could differentiate an asymptomatic knee from a knee with symptomatic instability. Because of the wide variation of soft-tissue elasticity, osseous geometry, and subsequent joint laxity in the general population, it is inappropriate or impossible to designate a normal absolute value of patellar displacement. Lateral displacement in the normal individuals (Group 1) in our study ranged from one to thirty-two millimeters, and medial displacement ranged from two to twenty-two millimeters. The absolute displacement on the symptomatic side was not in and of itself thought to be important. However, the difference in excursion between the right and left knees in Group 1 was found to be minimum, while differences in displacement between the symptomatic and asymptomatic knees were found to be significant (Figs. 7 and 9). Differences between symptomatic and asymptomatic knees have been noted to be significant in studies of other ligaments of the knee and, indeed, form the basis for standardized tests4. An attempt to develop a force-versus-displacement measurement technique for the patella that does not involve use of radiographic measurements has so far been unsuccessful34. The present study provides a comparison of various radiographic measurements routinely used in the diagnosis of patellar instability with a new radiographic technique. It was not meant to compare radiographic measurements with clinical measurements such as the Q angle. Patella alta has been shown2,23 to be a predisposing factor in recurrent dislocation and subluxation of the patella and, in the present study, a significantly increased Insall-Salvati ratio was noted in the knees with lateral instability (p < 0.04) and those with multidirectional instability (p < 0.01). The congruence angle has been used since the mid-1970's as a standard for the radiographic diagnosis of patellar subluxation. The average congruence angle in Group 1 in the present series (-6 degrees) was the same as that reported by Merchant et al.22. The congruence angle in Group 2 (+6.41 degrees) did differ significantly (p < 0.003) from that in the other three groups in that the median ridge of the patella was lateral to the bisector of the sulcus angle; however, this mean angle was still less than the 16-degree angle that Merchant et al. defined as the criterion for a diagnosis of abnormality and the 23-degree angle that they specified for a diagnosis of recurrent dislocation. Although the comparison of the congruence angles in Group 1 with those in Group 2 yielded the strongest association found for any static radiographic measurement (p < 0.003), only eight (30 per cent) of the twenty-seven patients in Group 2 had the diagnosis of subluxation confirmed by the criterion of Merchant et al. The small (negative) angles found in both Group 3 and Group 4 were perhaps due to the previous lateral release, which may have allowed the patella to move farther medially than is seen in normal subjects. Laurin et al. described the lateral patellofemoral angle as being 0 degrees, or reversed, in all of their thirty patients who had lateral subluxation19. Inoue et al.8 determined the normal lateral patellofemoral angle to be 11 degrees when measured at 30 degrees of knee flexion and 12 degrees when measured at 45 degrees of knee flexion. In knees with a subluxating patella, they noted the angles to be 0 and 4 degrees, respectively. The normal value of 9.15 degrees in our study is similar to these values. The lateral patellofemoral angle for Group 2 was slightly abnormal (5.48 degrees; p < 0.043); however, only six (22 per cent) of twenty-seven patients had the diagnosis of lateral subluxation confirmed with the use of the criteria of Laurin et al.19. It should be noted that our measurements were made at 35 degrees of knee flexion, not at 20 degrees as recommended by Laurin et al.19. The mean sulcus angle was significantly greater (p < 0.01) in Group 2 than in Group 1. This increase is consistent with Brattström's observation2 that distal femoral dysplasia (a greater-than-normal sulcus angle) was an important finding in recurrent dislocation of the patella. The mean difference in lateral displacement between the right and left knees in Group 1 was 1.3 ± 1.10 millimeters. We therefore determined that a difference in lateral displacement of 3.7 millimeters is abnormal (2 x standard deviation = 95 per cent confidence). In Group 4, the minimum difference was 4.0 millimeters, but two of the twenty-seven patients in Group 2 had a difference of only 3.0 millimeters. Perhaps this smaller-than-expected difference reflects the frequent prevalence of bilaterality. Brattström2 also noted a higher prevalence of femoral dysplasia in the asymptomatic knee of patients who had recurrent dislocation of the patella. The mean difference in medial displacement between the right and left knees in Group 1 was 1.2 ± 1.08 millimeters. We therefore determined that a difference in medial displacement of 3.46 millimeters is significant (2 x standard deviation = 95 per cent confidence). The minimum difference in patellar displacement was 4.0 millimeters in Group 4 and 5.0 millimeters in Group 3. Three different clinical presentations of patellar instability were defined in the present study, and it was found that the patterns of the three groups differed greatly on stress radiographs. While a great deal has been written regarding lateral instability and its predisposing causes, little has been reported regarding either medial instability (Group 3 in our study) or multidirectional instability (Group 4), and we know of no previous studies that have demonstrated radiographic confirmation. Patients with such instability were part of a group of seventy patients who were seen for iatrogenic medial instability38; however, they were not the focus of that paper. In the present study, the patients who had lateral instability (Group 2) had not had any operative intervention and, thus, represented a group with pure, untreated lateral instability. These knees, when compared with those in Group 1 and Group 3, were found to have a significantly increased Insall-Salvati ratio, mean congruence angle, mean sulcus angle, and mean difference in lateral displacement as well as a significantly decreased mean lateral patellofemoral angle, as determined from the radiographs. All of these measurements, except for the difference in lateral displacement, have been previously associated with lateral subluxation or dislocation2.23 but are not diagnostic. While the lateral stress radiographs for Group 2 showed substantial differences between the symptomatic and asymptomatic knees, the medial stress radiographs did not differ from those in Group 1. This finding would seem to indicate normal integrity of the lateral patellar retinaculum (the lateral patellofemoral ligament). As measured on the stress radiographs, the difference in the mean side-to-side difference in lateral displacement between Group 1 and Group 2 was a much more reliable predictor of lateral instability than any measurements made on static radiographs. The patients who had medial instability (Group 3) had all had an isolated lateral release for pain in the anterior aspect of the knee and for instability, but their histories were not consistent with lateral patellar instability. Medial subluxation of the patella as a complication of lateral retinacular release was described by Hughston and Deese7, but they had no objective way to measure this instability or to provide radiographic confirmation. The findings in the patients in Groups 3 and 4 fit their criteria for a clinical diagnosis of medial patellar subluxation. All had worsening of the symptoms following a lateral release, and the symptoms of severe apprehension and pain were duplicated when the patella was displaced medially. This instability was thought to be iatrogenic. The Insall-Salvati ratio, congruence angle, lateral patellofemoral angle, sulcus angle, and difference in lateral patellar displacement with stress were normal for Group 3. The lack of any increase in lateral patellar displacement probably indicates a normal medial retinaculum and normal lateral stability. The mean medial displacement measured on the stress radiograph was significantly greater (p < 0.0001) on the symptomatic side, most likely because of release of the normal tether provided by the lateral patellar retinaculum. Radiographic confirmation of the clinical diagnosis of medial dislocation was completely dependent on the stress radiographs. The patients who had multidirectional instability (Group 4) had had an isolated lateral retinacular release as treatment for lateral instability, and the symptoms worsened despite physical therapy. This group would also have been considered by Hughston and Deese7 as having iatrogenic medial patellar subluxation. Examination revealed hypermobility of one patella, both medially and laterally, and displacement of the patella in both directions reproduced the symptoms of pain and apprehension. The radiographs of these patients showed an increased sulcus angle and patella alta, which reflect a risk of instability, but the lateral patellofemoral angle was normal and the congruence angle failed to reveal the lateral instability in this group. The stress radiographs clearly showed these patients to have differences in the medial and lateral displacement of the two knees that were significantly greater (p < 0.0001) than normal. A lateral release in a patient who has a history of lateral instability can result in multidirectional instability. Radiographic confirmation of the clinical diagnosis of multidirectional instability was completely dependent on the stress radiographs. The treatment of these disorders depends on the anatomical factors that are contributing to the instability. However, it was not the purpose of this report to discuss treatment options. The present study indicates that conventional static radiographs may suggest a potential (propensity) for instability as indicated by anatomical factors, such as a shallow sulcus or a flat or high-riding patella, and may show some degree of displacement. However, static radiographs may not show the large amount of patellar displacement demonstrable on passive stress. Our technique allows measurement of the amount of displacement of the patella from the femoral trochlea that occurs when a measured force is applied, and it provides additional objective information regarding the degree of patellar instability. The technique also makes it possible to quantitate changes in stability after operative procedures. Techniques for the evaluation of instability of the patellofemoral joint are still being developed. Axial patellofemoral radiographs with lateral rotation of the leg21, computed tomography30, and dynamic magnetic resonance imaging32,33 have recently been suggested for the diagnosis of subluxation. Ultimately, a simple and reliable method is needed to quantitate force and the resultant displacement. In conclusion, patellofemoral stress radiographs may confirm a clinical diagnosis of patellofemoral instability and are especially useful when conventional static patellofemoral radiographs reveal normal findings. Measurements on patellofemoral stress radiographs are more sensitive for the assessment of unilateral instability of the patella than are other radiographic measurements, and they allow quantitation of the force-displacement relationship of the patella and thus measurement of the function of the patellofemoral ligaments. Finally, patellofemoral stress radiographs differentiate lateral, medial, and multidirectional instability and can be used to document any improvement after operative procedures performed to correct instability.
NOTE: The authors thank Grace B. Cannon, Ph.D., of Biomedical Analytics, Rockville, Maryland, for providing statistical analysis and consultation in this study.
*One or more of the authors has received or will receive benefits for personal or professional use from a commercial party related directly or indirectly to the subject of this article. No funds were received in support of this study.
  CiteULike  Connotea  Del.icio.us  Facebook  Technorati  Twitter What's this?
This article has been cited by other articles:
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Top
Abstract
Introduction
| home | CME | help | feedback | subscriptions | archive | site index |
|
Stanford University Libraries' HighWire Press (TM) assists in the publication of JBJS Online. |