- Email: [email protected]
Ultrasound evaluation of sacroiliac motion in normal volunteers
Ultrasound Evaluation of Sacroiliac Motion in Normal Volunteers Pamela J. Lund, MD 1, Elizabeth A. Krupinski, PhD 1, William J. Brooks, DO 2
Rationale and Objectives. We demonstrated quantitatively, using ultrasound imaging, the passive range of motion of the normal sacroiliac (SI) joint. Methods. Ultrasound images of the SI joints of 22 adults at rest and during a manual medicine m a n e u v e r designed to induce a passive range of motion in the SI joint were obtained. Differences b e t w e e n the baseline alignment of the SI joint and alignment during induced passive motion were observed and measured by six radiologists. Results. Significant m o v e m e n t (> 2 m m ) of at least one SI joint was demonstrated in 82% of the subjects using ultrasound recordings. Interobserver ( r = .49-.81) and intraobserver (r = .87) correlations were high.
Conclusion. The results suggest that the range of passive SI joint motion is more than 2 mm, and m a y be up to 10 m m in some normal subjects, and that ultrasound imaging could be a useful m e t h o d for assessing passive SI movement. K e y W o r d s . Sacroiliac joint; motion; ultrasound; manual medicine. A
n u m b e r of studies [1-4] have b e e n conducted to describe and quantify sacroiliac (SI) joint motion. Many of those studies seemed to have had flawed methods that prevented the researchers from maximizing and properly recording $I motion. The first problem is that although some researchers [1-3] found no more than 0.5-2.0 m m of active range of motion (i.e., motion with activities of daily living such as lying, sitting, and standing), passive or induced motion, as evaluated routinely in the manual medicine musculoskeletal examination, was not adequately tested. That passive range of motion of the SI joint m a y well exceed 2.0 m m is supported b y Weisl [4], w h o found up to 6 m m of St motion using methods of motion generation similar to manual medicine spring testing (i.e., a technique that passively moves the SI joint). Toward this goal, a variety of imaging techniques [5, 6] have been used to quantify St movement. For example, Sturesson et al. [3] used a stereophotogrammetric imaging technique to image SI motion. Tantalum balls were
192
From the Departments of 1Radiologyand 2Surgery, University of Arizona, Tucson, AZ. Address reprint requests to P. J. Lund, MD, Department of Radiology, University of Arizona, P.O. Box 245067, Tucson, AZ 85724-5067. Received June 12, 1995, and accepted for publication November 2, 1995. Acad Radio11996;3:192-196 © 1996, Association of University Radiologists
Vol. 3, NO. 3, March 1996
inserted in the sacrum and pelvis and participants were asked to carry out various motions and adopt various postures (i.e., active motions). Sturesson et al. [3] concluded that active m o v e m e n t of the SI is small ( M = 0.5 mm), never exceeding 1.6 mm. We argue that their technique may be inadequate for detecting SI motion for two reasons. First, this and m a n y other studies have b e e n invasive (e.g., inserting Kirschner wires into the pelvis [1] or inserting tantalum balls into the sacrum and two iliac bones [2]). Introducing intraosseous devices might limit potential motion because of pain or biomechanical alteration. Second, in this and other studies, imaging techniques were used that are incapable of recording the extremely rapid passive range of motion of the SI produced during manual medicine spring testing. In an effort to overcome these problems, we attempted to use a manual medicine technique (spring testing) to induce a passive range of motion of the SI joint and to use a noninvasive imaging technique to record the movement. In addition to being noninvasive, an imaging technique was needed that could capture the rapid movement (i.e., about 1 sec total) of the anatomically complex SI joint during spring testing. Therefore, we chose to image the SI joint using real-time ultrasound, a technique that, to our knowledge, has never been used for this purpose. Unlike computed tomography scanning, fluoroscopy, and magnetic resonance imaging, ultrasound also provides a method in which simultaneous spring testing and imaging can be carried out without equipment or physical examination constraints. The goals of the current study were to demonstrate that the SI joint moves and that ultrasound can be used to reliably record and quantify this motion. MATERIALS AND METHODS
Twenty-two volunteers (9 m e n and 13 women, aged 19-51 years) served as subjects in our study. Informed consent was obtained from all subjects after the nature of the experimental procedures were fully explained both orally and in writing. To ensure that all subjects had adequate SI m o v e m e n t without pain, initial ratings of perceived (palpable) SI mobility at the posterosuperior iliac spine (PSIS) level during manual medicine spring testing were m a d e using the following scale: 0 = immobile, 1 = mild mobility, 2 = moderate mobility, 3 = substantial mobility, and 4 = very mobile. Each side was given a separate mobility rating by the manual medicine practitioner, w h o conducted all the manual medicine procedures. A separate determination per-
ULTRASOUND EVALUATION OF SACROILIAC MOTION
formed by the manual medicine practitioner using a pressure meter (Empi, St. Paul, MN) showed that forces of approximately 40 lb (18 kg) were generated for a typical spring test maneuver. Manual medicine spring testing of each SI joint was then performed while simultaneous ultrasound recordings (transferred to videotape) were made. Two to four spi'ing testing maneuvers were recorded at each position. A total of four overall ultrasound records per subject were obtained using a Toshiba 270 system (Toshiba America, Tustin, CA) as follows: at the PSIS and approximately 2 cm caudal to tile PSIS (inferior [INF]) levels for the right and left SI joints. A 5.0-MHz curvilinear transducer showed optimal penetration and resolution compared with 3.5MHz and 7.5-MHz transducers and was used for all recordings. The image plane was axial or axial-coronal from the posterior approach. For consistency in recording, only one sonographer conducted all the recordings. During the procedure, the patient was prone and the manual medicine physician straddled him or her (at thigh level). The physician then located the sacral margin of the SI joint at the PSIS level and the opposite inferior sacral margin using palpation. Downward (posterior to anterior) pressure was then applied rapidly by the examiner on the opposite inferior sacral margin in order to induce sacral displacement in an oblique axis. This motion (nodding or nutation) was seen in the axial scan plane as movement of the sacrum toward the transducer. A less prominent coronal or axial plane motion could sometimes be seen as m o v e m e n t of the sacrum toward the ilium or "closing" of the SI joint (Fig. 1). Ultrasound videorecordings for each subject were then s h o w n to six radiologists w h o typically read ultrasound images. Because this technique has not been used previously, the radiologists were s h o w n a videotape of another subject not included in the original group of subjects. This tape was used to acquaint the radiologists with the SI ultrasound anatomy and demonstrate the type of m o v e m e n t that resulted from the spring testing. During this training session, the radiologists also were s h o w n types of m o v e m e n t that should be distinguished from SI motion, such as m o v e m e n t of the transducer and overall m o v e m e n t of the pelvis. The discrimination of potential types of m o v e m e n t was not identified as a problem for the readers, and they were confident that they were successfully isolating SI movement w h e n it occurred. Radiologists viewed each subject's record in a splitscreen videotape format. On the left half of the screen
193
LUND
ET AL.
vol. 3, No. 3, March 1996
FIGURE 1. Potential planes of sacroiliac motion. A, Anteroposterior drawing. Arrows indicate motion planes including rotation, translation, and vertical displacement. B, Lateral diagram shows direction of sagittal plane motion (nutation) reproduced passively with the spring test maneuver.
A
B
Ps,,
--y--
I
L/-
INF
Static
A
B
AP
Lat
Real-time
C
FIGURE 2. Ultrasound evaluation of motion at PSIS and INF levels. Posterior transverse static ultrasound at the PSIS (A) and INF (B) levels. C, Static and dynamic representations of position of sacroiliac joint from left to right. PSIS = posterosuperior lilac spine, INF = inferior, I = ilium, AP = anteroposterior, ML = mediolateral, S = sacrum, lat = lateral.
was a static image of the SI level at rest and on the right was a real-time recording of the SI during application of the two to four spring test maneuvers at each location (right and left joints, PSIS and INF levels). Readers could repeatedly view each record to assess the SI movement. They were asked to judge the maximum displacement (in millimeters) observed in the anteroposterior and mediolateral directions on the basis of the maneuver that demonstrated the best and clearest anatomic detail (Fig. 2). The millimeter scale displayed on the ultrasound videotape was used to quantitatively estimate motion. In addition to recording movement in millimeters for each joint and level, the radiologists also were asked to grade the degree of overall perceived movement using the following scale: 0 = no motion, 1 = mild motion, 2 = moderate motion, and 3 = marked
194
motion. One of the radiologists read each record twice (on two separate occasions about 3 weeks apart) to obtain an estimate of reader reliability. RESULTS
Correlation analyses were done initially to determine the degree to which the six radiologists agreed with each other on the degree of movement in millimeters and the subjective ratings of overall joint movement, as well as to ensure that all readers were observing and judging the same phenomena. Overall, the radiologists were highly consistent in judging whether a joint moved significantly (2 mm or more) or not (0-2 mm). tnterobserver correlation coefficients comparing the millimeter judgments of the radiologists ranged from
Vol. 3, No. 3, March 1996
ULTRASOUND
.49 to .75. Interobserver correlation coefficients c o m paring the subjective grading of m o v e m e n t b y the radiologists r a n g e d from .57 to .81. The intraobserver correlation coefficient for the radiologist w h o read all cases twice was .87. The initial ratings of perceived SI mobility for eac h joint w e r e as follows: no joints w e r e given a rating o f 0, three w e r e rated 1, 20 w e r e rated 2, 19 w e r e rated 3, and two w e r e rated 4 (22 subjects = 44 joints total). Results of readers' ultrasound evaluations s h o w e d maxi m u m m o t i o n at the PSIS level bilaterally, and m o v e m e n t w a s judged to occur m o r e in the anteroposterior than in the mediolateral direction. A cutoff point of 2 m m or m o r e w a s used to indicate significant SI m o v e ment. Overall, 82% (18 of 22) of the subjects w e r e judged, o n average, to have significant m o v e m e n t (>2.0 ram) in at least o n e joint. The percentage of cases for each joint (at a given level and given direction of m o v e m e n t ) that w e r e j u d g e d by all six radiologists as m o v i n g 2 m m or m o r e are s h o w n in Table 1. A r e p e a t e d measures analysis of variance indicated that there w e r e significant differences b e t w e e n the eight SI motions studied (Table 1), F(7, 91.7) = 27.44, p < .0001. Post h o c Scheff6 tests indicated that there w e r e significant differences (p < .05) b e t w e e n each c o r r e s p o n d i n g right a n d !eft joint pairs (e.g., left PSIS anteroposterior versus right PSIS anteroposterior). There also w e r e significant differences (p < .05) for each anteroposterior versus mediolateral direction c o m p a r i s o n for e a c h
EVALUATION
OF SACROILIAC
MOTION
TABLE 1 : Percentage of Cases Judged to Move 2 mm or More on Spring Thrusting Joint
Cases (%)
Left PSIS AP PSIS ML INF AP INF ML Right PSIS AP PSIS ML INF AP INF ML
70.45 59.85 47.73 51.52 68.18 49.24 35.61 36.36
PSIS = poster0superior iliac spine, AP = anteropostedor motion, ML = mediolateral motion•
given jo{nt at a given level (e.g., left PSIS anteroposterior versus left PSIS mediolateral), except for the right INF comparison. There w e r e no significant differences in the degree of m o v e m e n t b e t w e e n m e n a n d w o m e n or b e t w e e n w o m e n w h o had b e e n p r e g n a n t ( n = 6) an d tl~ose w h o had never b e e n p r e g n a n t (n = 7). The distributions of m o v e m e n t judgments (in millimeters) for the eight SI motions studied w e r e analyzed to demonstrate graphically w h a t percentage o f cases w e r e judged to m o v e a given distance (data w e r e b i n n e d in l-ram increments). As can be seen in Figure 3, a significant a m o u n t o f data fell b e y o n d the 2-mm displacement point, and s o m e joints (primarily left PSIS anteroposterior) w e r e judged to m o v e up to 10 mm.
50,
R I G•H T
[]
PSIS-AP
[]
PSIS-ML
[]
!NF~ A P
LEFT
I I PSIS-AP []
40
(.~
PSIS-ML
[]
INF-AP
[]
INF-ML
30"
20
,,=, i'~
10
0 1
2
AVERAGE
3
4
5
6
7
8
DISPLACEMENT
g
10
(mm)
1
2
AVERAGE
3
4
~;
6
7
8
DISPLACEMENT
9
10
(ram)
FIGURE 3. Averaged displacements for six readers at the right (left panel) and left (right panel) sacroiliac joint of cases moving a given distance. PSIS = posterosuperior iliac spine, INF = inferior, AP = anteroposterior, ML = mediolateral.
195
LUND
ET AL.
Vol. 3, No. 3, March 1996
DISCUSSION
than it being a true difference in right versus left mobility of the SI joint. To k e e p the subjects comfortable and to conduct the sessions quickly, we asked the subjects to lay prone on the examination table. The ultrasound equipment and the sonographer were located to the patient's left. The manual medicine physician was up on the table straddling the patient (at thigh level) while he performed the spring testing on the SI joint. Because he w a s right-handed and his right hand was on the patient's right inferior lateral angle during spring testing of the left SI joint, this action might have produced the greatest and most consistent application of pressure to the joint. Using. his left hand to apply the spring testing pressure to the patient's left side duringthe right SI maneuver might have resulted in less consistent force on the right SI joint. Other potential error sources included limitations in the quantitative evaluation of the rapid spring test maneuver, which occurs within approximately t sec. Although the spring test is designed to isolate SI joint motion as m u c h as possible, invariably accessory pelvic motion occurs at the pubis, hips, and lumbosacral junction, which could interfere with accurate assessment of SI joint motion. In addition, a small amount of transducer m o v e m e n t necessarily occurred during spring testing, which could cause a false or inaccurate impression of SI motion. I n conclusion, we have demonstrated passive $I motion sonographically. This technique m a y have future applications for evaluating SI function in normal and diseased states.
The data support the hypothesis that ultrasound techniques are valid for evaluating passive SI movement. Note, however, that because of the rapid m o v e m e n t of the SI during spring testing, direct measurements from the analog images were difficult, although the qualitative motion assessments were judged to be fairly easy b y all six radiologists. All radiologists had little trouble identifying the relevant anatomic features necessary to make the judgments; as the correlation results indicate, they were consistent in the measures of h o w much a given joint area moved. Digitizing the images and using sophisticated image-processing techniques to isolate and evaluate the movements might obtain more precise measures of SI displacement. As noted, however , the radiologists were confident that they had accurately judged any movements that occurred, and the correlation data confirmed this.Our data indicate that manual medicine evaluation of the SI joint using spring testing maneuvers to maximize passive joint movement d o e s produce significant SI joint movement (i.e., >2 ram), as recorded by ultrasound and judged by radiologists. As indicated in Table 1, 36-71% (right INF anteroposterior and left PSIS anteroposterior, respectively) of all joint motions were judged to be more than 2 mm. That more movement was judged to occur in the anteroposterior than in the mediolateral direction in the majority of cases might have been due to an easier perception of or lower threshold for m o v e m e n t in the anteroposterior direction because of larger and more distinct reference surfaces for the anteroposterior motion than for the mediolateral motion (i.e., PSIS and sacrum versus St joint .margins, respectively). We think that this p h e n o m e n o n is more likely an accurate reflection of the functional mechanics of the SI joint, which require that the joint rotate in the sagittal plane (nutate) during daily activities. Other investigators have documented and quantified active nutation (nodding) movements in recumbent-to-sitting and sitting-to-standing positions [13]. During spring testing, which is designed to elicit maxim u m unilateral sacral nutation, the resulting motion should occur predominately in the anteroposterior or sagittal plane and the maximum displacement should occur at the more cranial level of the PSIS using our combined spring test and ultrasound scanning technique. The differences in right versus left SI joint movement are most likely caused by a methodologic variable rather
196
ACKNOWLEDGMENTS
We thank Hazel Dickey for secretarial assistance and Barbara Brooks for subject recruitment assistance. REFERENCES 1. Colachis SC, Worden RE, Bechtol CO, Strohm BR. Movement of the sacroiliac joint in the adult male: a preliminary report. Arch Physic Med Rehab 1963;44:490-498. 2. Egund N, Olsson TH, Schmid H, Selvik G. Movements in the sacroiliac joints demonstrated with roentgen stereophotogrammetry. Acta Radiol Diag 1978;19:833'-846. 3. Sturesson B, Selvik G, Uden A. Movements of the sacroiliac joints: a roentgen stereophotogrammetric analysis. Spine 1989;14:162-165. 4. Weisl H. The movements of the sacro-iliac joint. Acta Anatomica 1955; 23:80-91. 5. Brunner C, Kissling R, Jacob HAC. The effects of morphology and histopathologic findings on the mobility of the sacroiliac joint. Spine 1991 ;16:1111-1117. 6. Kissling RO, Brunner C, Jacob HAC. Zur beweglichkeit der iliosacralgelenke in vitro. Z Orthop 1990;128:282-288.
Rationale and Objectives. We demonstrated quantitatively, using ultrasound imaging, the passive range of motion of the normal sacroiliac (SI) joint. Methods. Ultrasound images of the SI joints of 22 adults at rest and during a manual medicine m a n e u v e r designed to induce a passive range of motion in the SI joint were obtained. Differences b e t w e e n the baseline alignment of the SI joint and alignment during induced passive motion were observed and measured by six radiologists. Results. Significant m o v e m e n t (> 2 m m ) of at least one SI joint was demonstrated in 82% of the subjects using ultrasound recordings. Interobserver ( r = .49-.81) and intraobserver (r = .87) correlations were high.
Conclusion. The results suggest that the range of passive SI joint motion is more than 2 mm, and m a y be up to 10 m m in some normal subjects, and that ultrasound imaging could be a useful m e t h o d for assessing passive SI movement. K e y W o r d s . Sacroiliac joint; motion; ultrasound; manual medicine. A
n u m b e r of studies [1-4] have b e e n conducted to describe and quantify sacroiliac (SI) joint motion. Many of those studies seemed to have had flawed methods that prevented the researchers from maximizing and properly recording $I motion. The first problem is that although some researchers [1-3] found no more than 0.5-2.0 m m of active range of motion (i.e., motion with activities of daily living such as lying, sitting, and standing), passive or induced motion, as evaluated routinely in the manual medicine musculoskeletal examination, was not adequately tested. That passive range of motion of the SI joint m a y well exceed 2.0 m m is supported b y Weisl [4], w h o found up to 6 m m of St motion using methods of motion generation similar to manual medicine spring testing (i.e., a technique that passively moves the SI joint). Toward this goal, a variety of imaging techniques [5, 6] have been used to quantify St movement. For example, Sturesson et al. [3] used a stereophotogrammetric imaging technique to image SI motion. Tantalum balls were
192
From the Departments of 1Radiologyand 2Surgery, University of Arizona, Tucson, AZ. Address reprint requests to P. J. Lund, MD, Department of Radiology, University of Arizona, P.O. Box 245067, Tucson, AZ 85724-5067. Received June 12, 1995, and accepted for publication November 2, 1995. Acad Radio11996;3:192-196 © 1996, Association of University Radiologists
Vol. 3, NO. 3, March 1996
inserted in the sacrum and pelvis and participants were asked to carry out various motions and adopt various postures (i.e., active motions). Sturesson et al. [3] concluded that active m o v e m e n t of the SI is small ( M = 0.5 mm), never exceeding 1.6 mm. We argue that their technique may be inadequate for detecting SI motion for two reasons. First, this and m a n y other studies have b e e n invasive (e.g., inserting Kirschner wires into the pelvis [1] or inserting tantalum balls into the sacrum and two iliac bones [2]). Introducing intraosseous devices might limit potential motion because of pain or biomechanical alteration. Second, in this and other studies, imaging techniques were used that are incapable of recording the extremely rapid passive range of motion of the SI produced during manual medicine spring testing. In an effort to overcome these problems, we attempted to use a manual medicine technique (spring testing) to induce a passive range of motion of the SI joint and to use a noninvasive imaging technique to record the movement. In addition to being noninvasive, an imaging technique was needed that could capture the rapid movement (i.e., about 1 sec total) of the anatomically complex SI joint during spring testing. Therefore, we chose to image the SI joint using real-time ultrasound, a technique that, to our knowledge, has never been used for this purpose. Unlike computed tomography scanning, fluoroscopy, and magnetic resonance imaging, ultrasound also provides a method in which simultaneous spring testing and imaging can be carried out without equipment or physical examination constraints. The goals of the current study were to demonstrate that the SI joint moves and that ultrasound can be used to reliably record and quantify this motion. MATERIALS AND METHODS
Twenty-two volunteers (9 m e n and 13 women, aged 19-51 years) served as subjects in our study. Informed consent was obtained from all subjects after the nature of the experimental procedures were fully explained both orally and in writing. To ensure that all subjects had adequate SI m o v e m e n t without pain, initial ratings of perceived (palpable) SI mobility at the posterosuperior iliac spine (PSIS) level during manual medicine spring testing were m a d e using the following scale: 0 = immobile, 1 = mild mobility, 2 = moderate mobility, 3 = substantial mobility, and 4 = very mobile. Each side was given a separate mobility rating by the manual medicine practitioner, w h o conducted all the manual medicine procedures. A separate determination per-
ULTRASOUND EVALUATION OF SACROILIAC MOTION
formed by the manual medicine practitioner using a pressure meter (Empi, St. Paul, MN) showed that forces of approximately 40 lb (18 kg) were generated for a typical spring test maneuver. Manual medicine spring testing of each SI joint was then performed while simultaneous ultrasound recordings (transferred to videotape) were made. Two to four spi'ing testing maneuvers were recorded at each position. A total of four overall ultrasound records per subject were obtained using a Toshiba 270 system (Toshiba America, Tustin, CA) as follows: at the PSIS and approximately 2 cm caudal to tile PSIS (inferior [INF]) levels for the right and left SI joints. A 5.0-MHz curvilinear transducer showed optimal penetration and resolution compared with 3.5MHz and 7.5-MHz transducers and was used for all recordings. The image plane was axial or axial-coronal from the posterior approach. For consistency in recording, only one sonographer conducted all the recordings. During the procedure, the patient was prone and the manual medicine physician straddled him or her (at thigh level). The physician then located the sacral margin of the SI joint at the PSIS level and the opposite inferior sacral margin using palpation. Downward (posterior to anterior) pressure was then applied rapidly by the examiner on the opposite inferior sacral margin in order to induce sacral displacement in an oblique axis. This motion (nodding or nutation) was seen in the axial scan plane as movement of the sacrum toward the transducer. A less prominent coronal or axial plane motion could sometimes be seen as m o v e m e n t of the sacrum toward the ilium or "closing" of the SI joint (Fig. 1). Ultrasound videorecordings for each subject were then s h o w n to six radiologists w h o typically read ultrasound images. Because this technique has not been used previously, the radiologists were s h o w n a videotape of another subject not included in the original group of subjects. This tape was used to acquaint the radiologists with the SI ultrasound anatomy and demonstrate the type of m o v e m e n t that resulted from the spring testing. During this training session, the radiologists also were s h o w n types of m o v e m e n t that should be distinguished from SI motion, such as m o v e m e n t of the transducer and overall m o v e m e n t of the pelvis. The discrimination of potential types of m o v e m e n t was not identified as a problem for the readers, and they were confident that they were successfully isolating SI movement w h e n it occurred. Radiologists viewed each subject's record in a splitscreen videotape format. On the left half of the screen
193
LUND
ET AL.
vol. 3, No. 3, March 1996
FIGURE 1. Potential planes of sacroiliac motion. A, Anteroposterior drawing. Arrows indicate motion planes including rotation, translation, and vertical displacement. B, Lateral diagram shows direction of sagittal plane motion (nutation) reproduced passively with the spring test maneuver.
A
B
Ps,,
--y--
I
L/-
INF
Static
A
B
AP
Lat
Real-time
C
FIGURE 2. Ultrasound evaluation of motion at PSIS and INF levels. Posterior transverse static ultrasound at the PSIS (A) and INF (B) levels. C, Static and dynamic representations of position of sacroiliac joint from left to right. PSIS = posterosuperior lilac spine, INF = inferior, I = ilium, AP = anteroposterior, ML = mediolateral, S = sacrum, lat = lateral.
was a static image of the SI level at rest and on the right was a real-time recording of the SI during application of the two to four spring test maneuvers at each location (right and left joints, PSIS and INF levels). Readers could repeatedly view each record to assess the SI movement. They were asked to judge the maximum displacement (in millimeters) observed in the anteroposterior and mediolateral directions on the basis of the maneuver that demonstrated the best and clearest anatomic detail (Fig. 2). The millimeter scale displayed on the ultrasound videotape was used to quantitatively estimate motion. In addition to recording movement in millimeters for each joint and level, the radiologists also were asked to grade the degree of overall perceived movement using the following scale: 0 = no motion, 1 = mild motion, 2 = moderate motion, and 3 = marked
194
motion. One of the radiologists read each record twice (on two separate occasions about 3 weeks apart) to obtain an estimate of reader reliability. RESULTS
Correlation analyses were done initially to determine the degree to which the six radiologists agreed with each other on the degree of movement in millimeters and the subjective ratings of overall joint movement, as well as to ensure that all readers were observing and judging the same phenomena. Overall, the radiologists were highly consistent in judging whether a joint moved significantly (2 mm or more) or not (0-2 mm). tnterobserver correlation coefficients comparing the millimeter judgments of the radiologists ranged from
Vol. 3, No. 3, March 1996
ULTRASOUND
.49 to .75. Interobserver correlation coefficients c o m paring the subjective grading of m o v e m e n t b y the radiologists r a n g e d from .57 to .81. The intraobserver correlation coefficient for the radiologist w h o read all cases twice was .87. The initial ratings of perceived SI mobility for eac h joint w e r e as follows: no joints w e r e given a rating o f 0, three w e r e rated 1, 20 w e r e rated 2, 19 w e r e rated 3, and two w e r e rated 4 (22 subjects = 44 joints total). Results of readers' ultrasound evaluations s h o w e d maxi m u m m o t i o n at the PSIS level bilaterally, and m o v e m e n t w a s judged to occur m o r e in the anteroposterior than in the mediolateral direction. A cutoff point of 2 m m or m o r e w a s used to indicate significant SI m o v e ment. Overall, 82% (18 of 22) of the subjects w e r e judged, o n average, to have significant m o v e m e n t (>2.0 ram) in at least o n e joint. The percentage of cases for each joint (at a given level and given direction of m o v e m e n t ) that w e r e j u d g e d by all six radiologists as m o v i n g 2 m m or m o r e are s h o w n in Table 1. A r e p e a t e d measures analysis of variance indicated that there w e r e significant differences b e t w e e n the eight SI motions studied (Table 1), F(7, 91.7) = 27.44, p < .0001. Post h o c Scheff6 tests indicated that there w e r e significant differences (p < .05) b e t w e e n each c o r r e s p o n d i n g right a n d !eft joint pairs (e.g., left PSIS anteroposterior versus right PSIS anteroposterior). There also w e r e significant differences (p < .05) for each anteroposterior versus mediolateral direction c o m p a r i s o n for e a c h
EVALUATION
OF SACROILIAC
MOTION
TABLE 1 : Percentage of Cases Judged to Move 2 mm or More on Spring Thrusting Joint
Cases (%)
Left PSIS AP PSIS ML INF AP INF ML Right PSIS AP PSIS ML INF AP INF ML
70.45 59.85 47.73 51.52 68.18 49.24 35.61 36.36
PSIS = poster0superior iliac spine, AP = anteropostedor motion, ML = mediolateral motion•
given jo{nt at a given level (e.g., left PSIS anteroposterior versus left PSIS mediolateral), except for the right INF comparison. There w e r e no significant differences in the degree of m o v e m e n t b e t w e e n m e n a n d w o m e n or b e t w e e n w o m e n w h o had b e e n p r e g n a n t ( n = 6) an d tl~ose w h o had never b e e n p r e g n a n t (n = 7). The distributions of m o v e m e n t judgments (in millimeters) for the eight SI motions studied w e r e analyzed to demonstrate graphically w h a t percentage o f cases w e r e judged to m o v e a given distance (data w e r e b i n n e d in l-ram increments). As can be seen in Figure 3, a significant a m o u n t o f data fell b e y o n d the 2-mm displacement point, and s o m e joints (primarily left PSIS anteroposterior) w e r e judged to m o v e up to 10 mm.
50,
R I G•H T
[]
PSIS-AP
[]
PSIS-ML
[]
!NF~ A P
LEFT
I I PSIS-AP []
40
(.~
PSIS-ML
[]
INF-AP
[]
INF-ML
30"
20
,,=, i'~
10
0 1
2
AVERAGE
3
4
5
6
7
8
DISPLACEMENT
g
10
(mm)
1
2
AVERAGE
3
4
~;
6
7
8
DISPLACEMENT
9
10
(ram)
FIGURE 3. Averaged displacements for six readers at the right (left panel) and left (right panel) sacroiliac joint of cases moving a given distance. PSIS = posterosuperior iliac spine, INF = inferior, AP = anteroposterior, ML = mediolateral.
195
LUND
ET AL.
Vol. 3, No. 3, March 1996
DISCUSSION
than it being a true difference in right versus left mobility of the SI joint. To k e e p the subjects comfortable and to conduct the sessions quickly, we asked the subjects to lay prone on the examination table. The ultrasound equipment and the sonographer were located to the patient's left. The manual medicine physician was up on the table straddling the patient (at thigh level) while he performed the spring testing on the SI joint. Because he w a s right-handed and his right hand was on the patient's right inferior lateral angle during spring testing of the left SI joint, this action might have produced the greatest and most consistent application of pressure to the joint. Using. his left hand to apply the spring testing pressure to the patient's left side duringthe right SI maneuver might have resulted in less consistent force on the right SI joint. Other potential error sources included limitations in the quantitative evaluation of the rapid spring test maneuver, which occurs within approximately t sec. Although the spring test is designed to isolate SI joint motion as m u c h as possible, invariably accessory pelvic motion occurs at the pubis, hips, and lumbosacral junction, which could interfere with accurate assessment of SI joint motion. In addition, a small amount of transducer m o v e m e n t necessarily occurred during spring testing, which could cause a false or inaccurate impression of SI motion. I n conclusion, we have demonstrated passive $I motion sonographically. This technique m a y have future applications for evaluating SI function in normal and diseased states.
The data support the hypothesis that ultrasound techniques are valid for evaluating passive SI movement. Note, however, that because of the rapid m o v e m e n t of the SI during spring testing, direct measurements from the analog images were difficult, although the qualitative motion assessments were judged to be fairly easy b y all six radiologists. All radiologists had little trouble identifying the relevant anatomic features necessary to make the judgments; as the correlation results indicate, they were consistent in the measures of h o w much a given joint area moved. Digitizing the images and using sophisticated image-processing techniques to isolate and evaluate the movements might obtain more precise measures of SI displacement. As noted, however , the radiologists were confident that they had accurately judged any movements that occurred, and the correlation data confirmed this.Our data indicate that manual medicine evaluation of the SI joint using spring testing maneuvers to maximize passive joint movement d o e s produce significant SI joint movement (i.e., >2 ram), as recorded by ultrasound and judged by radiologists. As indicated in Table 1, 36-71% (right INF anteroposterior and left PSIS anteroposterior, respectively) of all joint motions were judged to be more than 2 mm. That more movement was judged to occur in the anteroposterior than in the mediolateral direction in the majority of cases might have been due to an easier perception of or lower threshold for m o v e m e n t in the anteroposterior direction because of larger and more distinct reference surfaces for the anteroposterior motion than for the mediolateral motion (i.e., PSIS and sacrum versus St joint .margins, respectively). We think that this p h e n o m e n o n is more likely an accurate reflection of the functional mechanics of the SI joint, which require that the joint rotate in the sagittal plane (nutate) during daily activities. Other investigators have documented and quantified active nutation (nodding) movements in recumbent-to-sitting and sitting-to-standing positions [13]. During spring testing, which is designed to elicit maxim u m unilateral sacral nutation, the resulting motion should occur predominately in the anteroposterior or sagittal plane and the maximum displacement should occur at the more cranial level of the PSIS using our combined spring test and ultrasound scanning technique. The differences in right versus left SI joint movement are most likely caused by a methodologic variable rather
196
ACKNOWLEDGMENTS
We thank Hazel Dickey for secretarial assistance and Barbara Brooks for subject recruitment assistance. REFERENCES 1. Colachis SC, Worden RE, Bechtol CO, Strohm BR. Movement of the sacroiliac joint in the adult male: a preliminary report. Arch Physic Med Rehab 1963;44:490-498. 2. Egund N, Olsson TH, Schmid H, Selvik G. Movements in the sacroiliac joints demonstrated with roentgen stereophotogrammetry. Acta Radiol Diag 1978;19:833'-846. 3. Sturesson B, Selvik G, Uden A. Movements of the sacroiliac joints: a roentgen stereophotogrammetric analysis. Spine 1989;14:162-165. 4. Weisl H. The movements of the sacro-iliac joint. Acta Anatomica 1955; 23:80-91. 5. Brunner C, Kissling R, Jacob HAC. The effects of morphology and histopathologic findings on the mobility of the sacroiliac joint. Spine 1991 ;16:1111-1117. 6. Kissling RO, Brunner C, Jacob HAC. Zur beweglichkeit der iliosacralgelenke in vitro. Z Orthop 1990;128:282-288.