The Correlation between Clinical and Arthroscopic Diagnosis of Meniscal Lesions
By Dr Jason L Crane
Department of Orthopaedic Surgery, Tygerberg Hospital, South Africa
Abstract
Methods: This is a retrospective study of 295 consecutive patients who underwent knee arthroscopy at 2nd Military Hospital Wynberg Cape Town from 26 February 2003 to 07 June 2007 for suspected knee pathology. Patient ages ranged from 14 to 83 years old with an average age of 41.5 years. All data was analyzed using the McNemar Chi-square test to extract the clinical significance.
Results: 209 meniscal lesions were diagnosed clinically, 143 medial and 66 lateral. Arthroscopically 185 meniscal lesions were confirmed, 107 medial and 78 lateral. The study found that clinical examination of the knee for suspected meniscal lesions resulted in an accuracy of 64.4%, a positive predictive value of 66.3%, negative predictive value of 60%, sensitivity of 79.1% and specificity of 56.1%.
Conclusion: The author is of the opinion that clinical evaluation alone is sufficient to diagnose meniscal pathology accurately and consistently.
Introduction
The aim of this thesis is to determine the predictive value of clinical examination in meniscal lesions. The outcome of the study may show that clinical examination alone is sufficient to diagnose meniscal lesions or it may find that clinical examination is unreliable and other diagnostic measures (e.g. MRI) may be indicated.
The thesis will discuss the anatomy, innervation and function of the meniscus as well as the pathogenesis of knee pain as a result of a meniscal injury. This thesis will include a literature review and analysis of previous international studies that have been done.
The thesis will conclude with an analysis of the statistical findings and use these findings to implement a protocol in the management of suspected meniscal lesions, taking into consideration the financial implications of these suggestions.
Anatomy of the Meniscus
Gross Anatomy1,2
The menisci are two wedge shaped semi lunar cartilages interposed between the convex femurs superiorly and the relatively flat tibial plateau inferiorly (fig 1,2). They measure approximately 35mm in diameter and (including their insertional ligaments) 110mm in length1. The menisci cover one half to two thirds of the articular surface of their corresponding tibial plateaus. The peripheral edges of the menisci are convex and attach to the capsule of the knee joint, except where the popliteus is interposed laterally. The anterior and posterior peripheral edges are attached loosely to the borders of the tibial plateaus by the coronary ligaments. The inner edges are concave, thin, and unattached. The inferior surface of each meniscus is flat, whereas the superior surface is concave in shape.
The medial meniscus is a c-shaped structure larger in radius than the lateral meniscus, with the posterior horn being wider than the anterior. The anterior insertional ligament of the medial meniscus is a fan shaped structure that inserts into the tibial plateau at the anterior intercondylar fossa, 6-7mm anterior to the enthesis of the anterior cruciate ligament. The posterior or upper fibres, of the anterior insertional ligament, are found to blend (64% of the cases) with the fibres of the transverse ligament, which connects the anterior horns of the medial and lateral menisci. The posterior insertional ligament of the medial meniscus attaches to the posterior intercondylar fossa of the tibia between the posterior enthesis of the lateral meniscus and the tibial enthesis of the posterior cruciate ligament. The peripheral border is firmly attached to the medial capsule and through the coronary ligament to the upper border of the tibia. The majority of the weight is transmitted through the posterior portion of the medial meniscus.
J: J Bone Joint Surg 46B:307, 1964.)
The lateral meniscus is more circular in form, covering up to two thirds of the articular surface of the underlying tibial plateau. The anterior insertional ligament of the lateral meniscus attaches to the anterior intercondylar fossa of the tibia, anterior to the intercondylar eminence just behind the tibial enthesis of the anterior cruciate ligament. Part of its fibres blend with those of the anterior cruciate ligament. The posterior insertional ligament of the lateral meniscus attaches to the tibia, posterior to the lateral intercondylar eminence and anterior to the posterior enthesis of the medial meniscus (fig 3). In 50% of the cases, the anterior fibres of the posterior insertional ligament of the lateral meniscus insert into the intercondylar fossa of the medial femoral condyle anterior to the origin of the posterior cruciate ligament, forming the anterior meniscofemoral ligament (Humphrey ligament). A posterior meniscofemoral ligament (Wrisberg ligament) is found in 76% of knees. It is formed by the posterior fibres of the posterior insertional ligament of the lateral meniscus that attach to the intercondylar fossa of the medial femoral condyle, posterior to the origin of the posterior cruciate ligament. The inner border, like that of the medial meniscus, is thin, concave, and free. The tendon of the popliteus muscle separates the posterolateral periphery of the lateral meniscus from the joint capsule and the fibular collateral ligament. The tendon of the popliteus is enveloped in a synovial membrane and forms an oblique groove on the lateral border of the meniscus.
Microscopic Anatomy1
Normal human meniscal tissue is composed of 72% water, 22% collagen, 0.8% glycosaminoglycans and 0.12% DNA. The posterior areas of meniscal tissue have higher water content. The meniscal body consists predominantly of a dense framework of circumferentially orientated coarse type I collagen fibres (fig 4). Radial fibres are found throughout but are less numerous. The radial fibres may act as a “tie” holding the circumferential fibres together, thereby resisting longitudinal splitting of the menisci and dissipating hoop. The collagens are extensively cross-linked by hydroxylpyridinium aldehydes. Type I collagen accounts for over 90% of meniscal tissue collagens while types II, III and V account for the remainder. This configuration provides the meniscus with greater elasticity and the ability to withstand compression forces.
Blood Supply1
Vascular supply to the menisci is provided by the lateral and medial geniculate arteries, which form from a perimeniscal capillary plexus (fig 5) with radial branches directed towards the centre of the joint. In the adult, the degree of vascular penetration from the periphery is 10-30% into the medial meniscus and 10-25% into the lateral meniscus. The anterior and posterior horns of the menisci are more vascularised than their bodies.
Innervation1,3,4
The nerve supply of the meniscus is, at present, debatable. Innervation arises mainly from the posterior articular nerve, but branches of the medial articular nerve provide part of the innervation of the medial meniscus. The menisci have a neural network of myelinated and unmyelinated nerves that extend from the perimeniscal connective tissue into the outer and middle thirds of the meniscus3. The perimeniscal connective tissue serves as a hilum for nerves entering the meniscus. The nerves enter along the outer edge of the meniscus and appear to funnel into the meniscus coming to a point at the boundary of the middle and inner thirds. Encapsulated end organs with a mechanoreceptor function predominate at the horns and attachment structures, while free nerve endings are found throughout, except for the inner third of the meniscal body. Three morphologically distinct mechanoreceptors namely, ruffini endings, golgi tendon organs and pacinian corpuscles have been identified.
No neural elements have been observed in the inner third of the meniscus. The horns and insertional ligaments have a sensory function that provides important proprioreceptive information related to the joint position. A greater concentration of nerves is found in the meniscal horns due to the need for afferent feedback at the extremes of flexion and extension. There is a greater concentration of nerves and receptors in the posterior horn of the meniscus, compared to the anterior horn; this may be related to the posterior location of collateral ligament attachment. During extension the collateral ligaments tighten because the greater curve of the anterior surface of the femoral condyles exerts leverage upon the ligament attachments. In turn, this may place the posterior horn of the meniscus under increasing pressure and tension, stimulating the mechanoreceptor system. The menisci could function as an “early warning” sensory device, coordinating the tension between the anterior meniscal horns, the posterior meniscal horns and/or posterior meniscofemoral ligament. It may be that the menisci aid in initiating protective muscular reflexes to compensate and adjust the tension in the different areas of the meniscus3.
In a study by Dye et al5 an attempted was made to correlate anatomical findings with actual physical findings. The researcher under went bilateral knee arthroscopy without anaesthesia and mapped out his neurosensory perceptions as each internal structure of the knee was stimulated. He experienced non- painful and poorly localized awareness at the inner rim, slight discomfort and poor localization at the capsular margins and moderate discomfort and poor localization at the anterior and posterior horns (fig 6). This observation may provide an explanation for the often poor localization that many patients experience with meniscal injury. The painful synovitis and capsular inflammation frequently associated with a meniscal injury may be a more important factor in the subjective localization of the site of possible cartilage damage rather than the sensation arising solely from the damaged meniscus.
Function of the meniscus
The menisci act as a joint filler, compensating for gross incongruity between femoral and tibial articulating surfaces. The menisci have an important role in load- bearing and shock absorption within the joint. They may also function as secondary stabilizers (particularly in the absence of a functioning anterior cruciate ligament), have a proprioceptive role and aid in the lubrication and nutrition of the articular cartilage.
Fairbank7 in 1948 was probably the first to suggest the weight-bearing role of the meniscus, resulting in articular surface protection. In 1974 Seedham7 confirmed that the menisci carry between 40% and 70% of the load across the knee joint. The intact menisci transmit load and absorb energy, thus protecting the articular cartilage, chondrocytes and extracellular matrix from compressive mechanical damage.
This is made possible by the strong anterior and posterior entheses to bone, which prevent the wedge shaped menisci from extruding from the joint during axial loading. Joint loading will tension the insertional ligaments and also the circumferential fibres of the meniscus. Thus part of the axial load will be transformed into hoop stresses at the meniscal periphery (fig 7A). Theoretically, a radial transection through the entire meniscal body or insertional ligaments will completely disable the load distribution function of the meniscus (fig 7B).1
Through their anatomical location the menisci prevent capsular and synovial impingement during flexion and extension of the knee2. They contribute to the stability of the knee in all planes but are especially important rotary stabilizers, and are probably essential for the smooth transmission from a pure hinge to a gliding or rotary motion as the knee moves from flexion to extension.
The menisci may attenuate the intermittent shock waves generated by impulse loading during normal gait. The shock absorbing capacity of normal knees is 20% higher than knees that have undergone a menisectomy.
As previously discussed, the horns and insertional ligaments have a sensory function that provides important proprioreceptive information related to the joint position.
Pathophysiology of Meniscal Tears
Meniscal tears occur in a surprisingly large portion of the population, with many of them being asymptomatic. Noble15 performed a series of 100 random necropsies (subjects had an average age of 65 years) where he found that of the 400 menisci studied, 29% contained a horizontal cleavage lesion and 60% of individuals had at least one other significant meniscal lesion. A further post mortem study of 70 subjects younger than 55 years of age revealed that 18.6% had at least one horizontal cleavage lesion of their menisci.
To understand the pathology of meniscal tears we have to understand the physiological movements of the meniscus. These were demonstrated by Vedi et al13 in an in-vivo study using dynamic MRI. They found that on weight bearing, the anterior horn of the medial meniscus moved through a mean of 7.1mm and the posterior horn through 3.9mm, with 3.6mm of mediolateral radial displacement. The height of the anterior horn increases by 2.6mm and that of the posterior horn by 2.0mm. The anterior horn of the lateral meniscus moves 9.5mm and the posterior horn 5.6mm, with 3.7mm of radial displacement. The height of the anterior horn increases by 4.0mm and that of the posterior horn by 2.4mm. (fig 8,9)
The lateral meniscus has greater movement than the medial meniscus and the anterior horns more than the posterior horns. These findings are in keeping with tibiofemoral kinematics and patterns of meniscal pathology. The relative immobility of the posterior part of the medial meniscus may account for the frequency with which tears occur in this area (the majority of all tears involve the medial meniscus).
Meniscal tears usually result from a single, acute rotational force applied to the weight-bearing knee, overloading the meniscus17 .
Degenerative tears occur from repetitive sub maximal forces applied to a meniscus having already undergone attritional wear from an irregular femoral articular surface. During axial rotation, the menisci follow exactly the displacements of the femoral condyles. Starting from the neutral position (fig 10b) they can be seen to move on the tibial condyles in the opposite direction19. During lateral rotation (fig 10a) the lateral meniscus is pulled towards the anterior part of the tibial condyle while the medial meniscus is drawn posterior. During medial rotation (fig 10c) the medial meniscus moves forward while the lateral meniscus recedes. During their movements the menisci distort around their fixed points, i.e. the attachments of their horns (lateral meniscus having a greater range of movement than the medial meniscus).
These displacements of the menisci during axial rotation are mostly passive, being drawn by the femoral condyles, but there is also an active mechanism involved. The meniscopatellar fibres become taught as a result of movement of the patella in relation to the tibia and the tension in these fibres draws the menisci anteriorly. During movements of the knee the menisci can be injured if they fail to follow the movements of the femoral condyles on the tibial condyles and as a result get caught between the two. This can occur during violent extension of the knee (e.g. kicking a ball), one of the menisci fails to move forwards (fig 11) and is caught between the femoral and tibial condyles as the tibia rotates anterior-superior to the femoral condyles. This mechanism leads to transverse tears (fig 12a) or to detachment of the anterior horn (fig 12b), which then folds on itself.
The other mechanism producing lesions of the menisci involves a twisting movement of the knee joint (fig 13), which combines lateral displacement (1) and lateral rotation (2). The medial meniscus is then pulled towards the centre of the joint under the convexity of the femoral condyle. When the joint is extended it is trapped between the two condyles with the following consequences (a) longitudinal splitting of the meniscus (fig 14) or (b) complete detachment of the meniscus from the capsule (fig15) or (c) complex tear of the meniscus (fig 16). In all these longitudinal lesions the central free part of the meniscus can rear itself up into the intercondylar notch so that the meniscus assumes the shape of a bucket handle. This type of lesion is common among footballers sustaining a fall onto a flexed knee.
Concomitant meniscal injury often accompany tears of the anterior cruciate ligament (ACL) as well as multiple knee ligament injuries. Lateral meniscal injuries are more frequent in acute ACL tears, while the medial meniscus is most often torn in chronic ACL insufficiency. In ACL injury (fig 17) The medial femoral condyle is no longer held back posteriorly and catches the posterior horn of the medial meniscus, which is pulled off its capsular attachment posteriorly or is torn horizontally.
Meniscal tears occur less frequently with complete medial collateral ligament (MCL) tears (grade III) than with partial tears (grade II) and in both instances, predominantly affect the lateral meniscus. As soon as a meniscus is torn, the injured part fails to follow its normal movements and becomes wedged between the femoral and tibial condyles. As a result the knee locks in a position of flexion thSat is more marked the more posterior the rupture occurs.
Classification of meniscal tears
Numerous classifications of meniscal tears exist14,18. Meniscal tears can be classified into the following categories (a) longitudinal or circumferential tears, (b) horizontal tears, (c) oblique tears, (d) radial tears, and (e) variations, which include flap tears, complex tears, and degenerative meniscal tears. Longitudinal tears most commonly occur as a result of trauma to a normal meniscus. The tear is usually vertically orientated and may extend completely or only partially through the meniscus. The tear is orientated parallel to the edge of the meniscus (fig 18). If the tear extends the full thickness of the meniscus, a mobile inner fragment is created. When the inner fragment displaces over into the intercondylar notch, it commonly is referred to as a bucket handle tear (fig 19).
Horizontal tears tend to be more common in older patients. The horizontal cleavage plane occurs from a shearing type injury, which divides the superior and inferior surfaces of the meniscus. These are more commonly seen in the posterior half of the medial meniscus or the midsegment of the lateral meniscus (fig 20).
Oblique tears are full-thickness tears running obliquely from the inner edge of the meniscus into the body of the meniscus. The tear can extend either posteriorly or anteriorly towards the anterior horn.
Radial tears, like oblique tears, are vertically oriented, extending from the inner edge of the meniscus toward its periphery. They can be complete or incomplete. Possible variations include flap tears, complex tears and degenerative meniscal tears. Flap tears are similar to oblique tears but usually have a horizontal cleavage element rather than being purely vertical in orientation (fig 21) .
Complex tears may contain elements of all of the above types of tears and are more common in chronic meniscal lesions or in older degenerative menisci. These are generally caused by chronic, long standing, altered mechanics of the abnormal meniscus.
Degenerative tears often refer to complex tears. These present with marked irregularity and complex tearing within the meniscus, seen mainly in elder patients (fig 22)
Clinical Examination16,20,21
In 1803 William Hey described a condition, which he termed “internal derangement of the knee”. He did not see it as a clear-cut clinical entity but his article did include the first, somewhat vague, description of locking of the knee. Hey suggested that this locking might be due to a meniscal injury21,30. Since then many famous orthopaedic surgeons have described clinical examinations for the torn meniscus, Sir Robert Jones, McMurray and Apley to name but a few.
Most of the tests are provocative in nature and rely on creating rotational and compressive forces to the knee, which result in impingement of the meniscus between the femoral and tibial condyles.
McMurray’s Test20
McMurray first described his test in 1942, and it has been used and modified ever since. From his paper entitled “The Semilunar Cartliages” Mcmurray describes his test as follows, “In carrying out the manipulation with the patient lying flat on their back, the knee is fully flexed until the heel approaches the buttock. The foot is then held by grasping the heel and using the forearm as a lever. The knee being now steadied by the surgeon’s other hand, the leg is rotated on the thigh with the knee still in full flexion (fig 23). During this movement the posterior section of the cartilage is rotated with the head of the tibia, and if the whole cartilage, or any fragment of the posterior section is loose, this movement produces an appreciable snap in the joint. By external rotation of the leg the internal cartilage is tested, and by internal rotation any abnormality of the posterior part of the external cartilage can be appreciated. By altering the position of flexion of the joint the whole of the posterior segment of the cartilages can be examined, from the middle to the posterior attachments. Thus if the leg is rotated with the knee at right angles the cartilages in their midsection come under pressure, but anterior to this point, the pressure exerted on the cartilage is so diminished that accurate examination is impossible. When a loose segment of the cartilage is caught between the bones during the rotation, the sliding of the femur over the loose fragment is accompanied by a thud or click, which can sometimes be heard but can always be felt, and the size of the detached portion can be judged by the rocking of the tibia, and usually also by the severity of the sound produced.” He goes on further to say, “This method of examination is not easy to master, the rotation requires a considerable amount of practice, and the whole procedure must be carried out systematically if success is to be attained.”
The Apley Test16,21
In 1947 A. Graham Apley published an article proposing new clinical methods. He criticised McMurrays test stating, “A click is not a truly reliable sign. Its cause may lie in a different knee injury or in a different joint entirely. Many surgeons fail to find it at all, although McMurray declares that the diagnostic click is reliably constant in his personal cases.” His argument was that, using McMurray’s test, the examiner is unable to distinguish between ligamentus injury and meniscal injury. He went about developing a test that is capable of differentiating between the two pathologies. In his original article he describes his examination as, “For this examination the patient lies on his face. He should be on a couch not more than two feet high, or the tests become difficult, and he must be well over to the edge of the couch nearest the surgeon. To start the examination, the surgeon grasps one foot in each hand, externally rotates as far as possible, and then flexes both knees together to their limit (fig 24). When the limit has been reached, he changes his grasp, rotates the feet inward, and extends the knees together again. This preliminary manoeuvre demonstrates limited rotation, painful rotation, and the exact angles of flexion at which these occur; the estimation of these angles proves useful later in the examination. The surgeon then applies his left knee to the back of the patient’s thigh (fig 25). It is important to observe that in this position his weight fixes one of the levers absolutely. The foot is grasped in both hands, the knee is bent to a right angle, and powerful external rotation is applied. This test determines whether simple rotation produces pain. Next, without changing the position of the hands, the patient’s leg is strongly pulled upward while the surgeons weight prevents the femur from rising off the couch (fig 26). In this position of distraction, the powerful external rotation is repeated. Two things can be determined: (1) whether or not the manoeuvre produces pain and (2), still more important, whether the pain is greater than in rotation alone without the distraction. If the pain is greater, the distraction test is positive, and a rotation sprain may be diagnosed. Then the surgeon leans well over the patient and, with his whole body weight, compresses the tibia downward into the couch (fig 27) Again he rotates powerfully, and asks the two questions: (1) does it hurt? (2) How much does it hurt? If the addition of compression has produced an increase of pain, this grinding test is positive, and meniscal damage is diagnosed. Incidentally, this question of the amount of pain is not a matter of fine hairline distinction; the patient must be sure of a considerable difference, and indeed he usually is.”
Joint Line Tenderness
Probably the most important finding in patients with a meniscal tear is localized tenderness along the joint line.
Steinmann Test
First Steinmann test. With the patient supine and the knee and hip flexed at 90°, the examiner forcefully and quickly rotates the tibia internally and externally (fig 28). Pain in the lateral compartment with forced internal rotation indicates a lateral meniscus lesion. Medial compartment pain during forced external rotation indicates a lesion of the medial meniscus.
Second Steinmann test. This test is indicated when point tenderness is located along the anterior joint line. When the examiner moves the knee from extension into flexion, the meniscus is displaced posteriorly, along with its lesion resulting in posterior point tenderness.
Bounce Home Test
The patient is supine with their heel cupped in the examiner's hand. The examiner fully flexes and then passively extends the knee. If the knee does not reach complete extension or has a rubbery or springy end feel, a torn meniscus may be blocking the knee movement.
Bohler Test
This test is performed in the same way as one tests the stability of the collateral ligaments. A valgus stress will result in pain in the presence of a lateral meniscus lesion, and a varus stress should result in pain in the presence of a medial meniscus lesion.
Payr Test
With the patient sitting cross- legged, the examiner exerts downward pressure along the medial aspect of the knee. Medial knee pain indicates a posterior horn lesion of the medial meniscus.
Bragard Test
This test may be used if anterior joint-line point tenderness is present. To test for a medial lesion, the examiner extends and externally rotates the tibia, which displaces a meniscal lesion forward, if one exists. Palpable tenderness along the anterior medial joint line is reduced with flexion and internal rotation.
Childress Test
The patient squats with the knee fully flexed and attempts to "duck walk." If the motion is blocked, a posterior horn meniscal lesion is indicated; however, pain in this position may also indicate patellofemoral joint pathology.
Merkel Test
The patient stands with their knees extended and rotates their upper body. This movement causes compression of the menisci. Medial compartment pain during internal rotation of the femur indicates a medial meniscal lesion. Lateral compartment pain occurring during external rotation of the femur indicates a lateral meniscal lesion.
Modified Helfet Test
With the patient sitting on the edge of a table with their knee flexed to 90°, the patient extends their knee. If knee mechanics are within normal limits, the tibial tuberosity can be seen in line with the midline of the patella in full flexion. During extension, the tibia rotates and the tibial tubercle moves into line with the lateral border of the patella (Qangle). Failure of the tibia to rotate during extension indicates a meniscal lesion or cruciate ligament pathology.
O'Donoghue Test
With the patient prone, the examiner flexes the knee to 90°. The examiner rotates the tibia internally and externally twice, then fully extends the knee and repeats the rotations. Increased pain during rotation in either or both knee positions indicates a meniscal tear or joint capsule irritation. If the patient experiences pain and instability to a valgus force on a flexed and laterally rotated knee, the medial meniscus, MCL and the ACL may all be injured, representing the O`Donoghue triad.
Literature Review
Since 1975, many studies have been undertaken to ascertain the reliability of clinical examination in meniscal lesions. There is little consensus between studies. Some find clinical examination to be reliable, some unreliable and others have found equality in both clinical and specialised diagnostic investigations (mainly MRI scans). With this vast variation in the outcomes of various studies, I was interested in determining the reliability of clinical examination, in meniscal lesions, as experienced by the 2nd Military Hospital Orthopaedic Unit.
Pro-Clinical Examination
A 1975 prospective study by Kenneth et al12 of 100 patients suffering “internal derangement of the knee” found clinical diagnosis to be correct in 72% of cases, correct but incomplete in 10% and incorrect in 18% of cases. When existing disease was accurately predicted but additional lesions were encountered that also required surgical treatment, the clinical diagnosis was considered correct but incomplete. They therefore reported an 82% positive predictive value (PPV).
Of further interest, the article stated, “Recently arthroscopy has been added in the diagnostic techniques for the evaluation of the knee, although interest in arthroscopy is increasing. There is little published information to indicate its practical value and the purpose of this report is to assess its role.” They did conclude that arthroscopy proved to be the most accurate diagnostic tool with a 94% accuracy rate.
Stanitski11 correlated clinical examination, MRI (magnetic resonance imaging) and arthroscopic findings of injured knees in children and adolescents. He found a highly positive correlation between clinical and arthroscopic findings (total agreement of 78.5%), a highly negative correlation between arthroscopic and MRI findings (total disagreement of 78.5%) and a negative correlation between clinical and MRI findings (total disagreement of 75%) (Table 1).
In this study, accuracy, positive predictive value, negative predictive value, sensitivity and specificity were more favourable in clinical examination than from MRI (Table 2,3). Overall, MRI diagnosis added little guidance to the patient management and at times even provided spurious information.
Clinical assessment of meniscal injuries resulted in 14 true positive, 12 true negative, 1 false positive and 1 false negative result, resulting in a clinical accuracy of 92.8%, positive predictive value of 93.3%, negative predictive value of 92.3%, sensitivity of 93.3% and specificity of 92.3% (Table 2).
In the MRI versus arthroscopy group, 2 of 28 patients (7.1%) had complete agreement of diagnosis between artroscopic examination and MRI findings, 4 (14.3%) had partial agreement and 22 (78.5%) had total disagreement. Meniscal injuries scanned by MRI had an accuracy of 37.5%, positive predictive value of 50%, negative predictive value of 50% and specificity of 45% (Table 3).
They found that an experienced clinician provided greater sensitivity and specificity for injuries to the ACL, menisci and articular cartilage than MRI. They concluded that MRI added little to the treatment plan and outcome of their treatment group.
Pro-Magnetic Resonance Imaging
In a prospective study of 58 patients with suspected internal derangement of the knee, Spiers et al9 found preoperative clinical assessment to have a diagnostic sensitivity of 77%, a specificity of 43%, positive predictive value of 59% and negative predictive value of 63% (Table 4), compared with 100% sensitivity, 63% specificity, 74% positive predictive value and 100% negative predictive value for MRI (Table 5). Comparison of MRI and arthroscopy confirmed the accuracy of MRI in the diagnosis of internal derangement of the knee (Figure 30).
They concluded that for the detection of meniscal tears, MRI was highly sensitive (100%) but less specific (lateral meniscus, 92% and medial meniscus, 71%) (Table 5).
A possible reason for MRI being so sensitive but less specific lies in the fact that both tears and degenerative changes produce areas of high signal within the meniscus, and the diagnosis of a tear depends on seeing that the high signal extends to the articular surface of the meniscus. The most contentious source of error is the over diagnosis by MRI of tears of the posterior horn of the meniscus.
Clinical and MRI equality
Mininder et al8 investigated the diagnostic performance of clinical examination and selective MRI in the evaluation of intraarticular knee disorders. 139 varied internal knee lesions were diagnosed clinically, 128 by MRI and 135 arthroscopically.
There was no significant difference between clinical examination and MRI with respect to arthroscopic findings of all internal lesions (Table 6). For isolated meniscus lesions they found the sensitivity for clinical examination to be 56.3% and for MRI to be 73%. The specificity was 85% for clinical examination and 87.4% for MRI. The positive predictive value for meniscal lesions was low (24.3%) for clinical examination and 32.2% for MRI. The negative predictive value for meniscal injuries was 95.9% for clinical examination and 97.2% for MRI. They concluded that selective MRI does not provide enhanced diagnostic utility over clinical examination, particularly in children, and should be used judiciously in cases where the diagnosis is uncertain and MRI input will alter the treatment plan.
The study showed a very low positive predictive value for diagnosing meniscal injuries in both clinical and MRI examination. Unfortunately the study makes no attempt to explain this low predictive value.
The Correlation between Clinical and Arthroscopic Diagnosis of Meniscal Lesions
The aim of this thesis is to determine the positive predictive value of clinical examination in meniscal lesions. The outcome of the study may show that clinical examination alone is sufficient to diagnose meniscal lesions or it may find that clinical examination is unreliable and other diagnostic measures (e.g. MRI) may be indicated.
Materials and Methods
This is a retrospective study of 295 consecutive patients who have undergone arthroscopy for suspected knee pathology identified through clinical examination at 2nd Military Hospital Wynberg Cape Town, from 26 February 2003 to 07 June 2007. Ethical approval for this study has been obtained through the South African National Defence Force.
Patients were all seen in an orthopaedic outpatients department under the supervision of two full time consultants. Patient ages ranged from 14 to 83 years old with an average age of 41.5 years. Patients were examined either by a consultant, registrar, medical officer or intern. Clinical diagnosis was made on the basis of history, a minimum of three recognised clinical examination techniques (predominantly joint line tenderness, McMurray’s test, Steinmann’s test and bounce home test) and standard knee radiographs. All patients were booked for arthroscopy within 6 weeks of clinical diagnosis.
Arthroscopy was carried out by either a consultant or registrar (who was under consultant supervision), using a standard Smith and Nephew arthroscope and two inferior portal technique. Patient’s clinical and arthroscopic data was captured and stored on the military database. Location and type of meniscal tear was recorded. If the tear extended into the posterior or anterior horn it was recorded as a posterior or anterior horn tear respectively. Arthroscopy was considered the standard against which the clinical findings were compared.
All data was analyzed by Dr M Kidd of the Centre for Statistical Consultations, University of Stellenbosch, using the McNemar Chi-square test to extract the clinical significance. P value of <0.05 was considered significant.
Results
Meniscal lesions
209 meniscal lesions were diagnosed clinically (143 medial and 66 lateral). Arthroscopically, 185 meniscal lesions were confirmed (107 medial and 78 lateral) (Graph 1). Of the meniscal injuries confirmed arthroscopically, 58% were medial and 42% lateral. Of these 21% (37) were radial tears, 24% (43) posterior horn tears, 18% (32) anterior horn tears, 11% (19) bucket handle tears, 5% (9) horizontal type tears and 21% (36) were not documented in the operative notes (Graph 2).
The most common type and location of tears were, the posterior horn of the medial meniscus (29 tears), anterior horn of the medial meniscus (28 tears), radial tear of the lateral meniscus (25 tears) and radial tear of the medial meniscus (16 tears).
Clinical findings
Suspected medial meniscus lesion on clinical examination resulted in an accuracy of 65%, positive predictive value of 51.8%, negative predictive value of 78.3%, sensitivity of 69.2% and specificity of 63.3%. While suspected lateral meniscus lesions resulted in an accuracy of 73%, positive predictive value of 48.5%, negative predictive value of 79.9%, sensitivity of 41% and specificity of 84.3% (Table 7). The clinical accuracy of diagnosing lateral meniscus pathology is statistically more predictable than medial meniscus pathology, p<0.05 (Graph 3).
The study showed that overall, for clinical examination of meniscal lesions the accuracy was 64%, positive predictive value was 66.3%, negative predictive value 60%, sensitivity 79.1% and specificity 56.1% (Table 7).
Individual meniscal lesions
24% (19 lesions) of lateral meniscal lesions were misdiagnosed, while 19% (20 lesions) of medial lesions were misdiagnosed. 22% (8 tears) of radial tears were missed, 19% (8 tears) of posterior horn tears, 16% (3 tears) of bucket handle tears, 13% (4 tears) of anterior horn tears and 11% (1 tear) of horizontal type tears were missed (Table 8).
The radial tear of the lateral meniscus was statistically the most misdiagnosed lesion (15 out of 25 lesions misdiagnosed), p<0.05 (Graph 4). While the posterior horn of the medial meniscus was statistically the most accurately diagnosed lesion (25 out of 29 lesions correctly diagnosed), p<0.05. (Graph 5)
Health care provider
The study also divided the examiners into two specific groups, consultant orthopaedic surgeons and junior doctors. Consultant orthopaedic surgeon’s clinical findings for suspected meniscal lesions resulted in an accuracy of 67.1%, positive predictive value of 71.4%, negative predictive value of 59.3%, sensitivity of 76.1% and specificity of 46.7%. While junior doctors clinical examinations resulted in an accuracy of 63.4%, positive predictive value of 64.7%, a negative predictive value of 60.3%, sensitivity of 80.2% and specificity of 59.1% (Table 9). These differences were not found to be statistically significant (p=0.57) (Graph 6).
Discussion
This study incorporates a large and diverse population group who have undergone arthroscopy at a single orthopaedic department over a four and a half year period. The study is retrospective, utilizing a well- maintained database. Changes in protocol, data capturing and information loss are still factors that could have compromised obtaining accurate patient clinical and operative data.
There were multiple clinical examiners, with vastly different levels of orthopaedic experience and training. This is reflective of a normal government orthopaedic outpatient clinic. The study does, however, make a distinction between consultant examination (76 patients examined) and junior doctor examinations (219 patients examined). Multiple examination techniques were implemented and the finding of each individual recognised test was not documented, but an overall clinical diagnosis was recorded in the patient’s notes. This study looks at the clinical diagnostic skills of any examining doctor in an orthopaedic outpatients department using any examination technique that they are confident in.
A shortcoming in this study, which seems to be present in most studies of this nature, is the confirmation of an accurate negative clinical finding. If no obvious clinical meniscal injury is present (diagnosed as either a suspected ligament sprain or soft tissue injury), the patient is treated conservatively and not booked for an arthroscopy to confirm the clinical diagnosis. Therefore a large number of negative clinical findings will not be included in this study, resulting in a misrepresented low accuracy and negative predictive value. The author is of the impression that most of the negative clinical diagnosis were accurate and that the negative predictive values should be higher than they are represented in this study. If the clinical diagnosis was falsely negative then it is likely that the patient would return with ongoing pain and an arthroscopy would have been performed. Sensitivity and specificity are calculated using true negative and false negative values and would therefore also be affected by the misrepresented low values. Therefore the positive predictive values obtained are used to determine the clinical outcome of this study. This study achieved a positive predictive value within the average range of the previously reviewed studies (Table 10).
Meniscal lesions
There was a greater amount of medial meniscal (107 tears) compared with lateral meniscal (78 tears) pathology confirmed by arthroscopy. Clinically, medial pathology (143 tears suspected) was overdiagnosed while lateral pathology (66 tears suspected) tendered to be under diagnosed (Graph 1).
Individual meniscal lesions
The most common meniscal tear was the posterior horn of the medial meniscus (29 tears), which corresponds with the literature13, this was also the most accurately diagnosed lesion. The most commonly misdiagnosed lesion was a radial tear of the lateral meniscus (15 out of 22 tears misdiagnosed), despite the fact that lateral meniscal pathology is more accurately predictable with clinical examination. The most likely reason for this is that lateral meniscal pathology, is only 41% clinically sensitive and therefore many lesions are missed.
Health care provider
Although the Orthopaedic consultants obtained a higher positive predictive value (71.4%) as opposed to the junior doctor (64.7%) there is no statistical difference in the clinical accuracy of diagnosing meniscal pathology between the two health care providers (p>0.05). All levels of health care providers, using simple, recognized clinical examination techniques, are capable of making a reliable and accurate clinical diagnosis of meniscal injuries.
Financial implications
Another goal of this study was to compare efficacy and cost between clinical examination and Magnetic Resonance Imaging (MRI). Although MRI is a non-invasive investigation it is purely diagnostic and any diagnosed meniscal pathology would necessitate further surgical intervention. In South African Government Hospitals the cost of a knee MRI scan (code P6235), is R3824, and the total price for a knee arthroscopy and menisectomy (including surgery, anaesthetic and day theatre, code P0667 and P0673) is R4860 (any surgical complications and their subsequent costs have not been taken into consideration). In this study 295 patients underwent arthroscopy, at a total cost of R1433700. If all patients underwent MRI initially, the cost would have been R1128080, but all positive findings would then require surgical management. In this study there were 185 confirmed meniscal tears (Even if it is assumed that MRI has a 100% accuracy, which in the best results of previous studies9 is know only to have a positive predictive value of 74% and negative predictive value of 100%). This would result in a further cost of R899100 for curative arthroscopic treatment of the meniscal pathology, giving a total cost of R2027180. If all suspected meniscal injuries in this study underwent a MRI scan, it would have cost an additional R593480 (30% cost increase).
Conclusion
Through the outcome of this study at 2nd Military Hospital, the author is of the opinion that clinical evaluation alone is sufficient to diagnose meniscal pathology accurately and consistently, regardless the experience of the examiner. Specialized tertiary investigations (e.g. MRI) provide no diagnostic or financial benefit for the patient.
References
- Messner A. The menisci of the knee joint. Anatomical and functional characteristics, and a rationale for clinical treatment. J anat 1998.193;161-178
- Canale S.T. Cambell’s Operative Orthopaedics tenth edition, Volume three: 21822191
- Marilyn L. Mechanoreceptors in the human medial meniscus. Acta anat. 1998;133:35-40
- Antonios P. The innervation of the human meniscus. CORR 1992; number 275:232-235
- Dye SF. Conscious neurosensory mapping of the internal structures of the human knee without intraarticular anesthesia. Am J Sports Med. 1998 Nov-Dec;26(6):773-7
- Mcdermott I.D. The consequences of menisectomy. J Bone Joint Surgery (Br)
- Noble J. The function, pathology and surgery of the meniscus. CORR 1985;number 210 september:62-67
- Minder S. Diagnostic performance of clinical examination and selective magnetic resonance imaging in the evaluation of intraarticular knee disorders in children and adolescents. American journal of sports medicine 2001;vol 29,3:292-295
- Spiers A.S.D. Can MRI of the knee affect arthroscopic practice. J Bone Joint Surg (Br) 1993;75-B:49-52
- Karachalios T. Diagnostic accuracy of a new clinical test (the Thessaly test) for early detection of meniscal tears. J Bone and Joint Surg May 2005;87-A(5):955-961
- Stanitski. C. Correlation of arthroscopic and clinical examinations with magnetic resonance imaging findings of injured knees in children and adolescents. American journal of sports medicine 1998;vol26, no 1:2-6
- DeHaven K.E. Diagnosis of internal derangements of the knee. The role of arthroscopy. J. Bone Joint Surg (Am).1975;57:802-810
- Verdi V. Meniscal movement. J. Bone Joint Surg (Br) 1999;81B: 37-41
- http://www.kneeguru.co.uk/html/ steps/step_04_meniscus/tear_t ypes.html
- Noble J. Function, Pathology, and Surgery of the Meniscus. Clinical orthopaedics and related research. 1986;number 210:63 16.
- Magee. D.J. Orthopaedic Physical Assessment. Third Edition. Chapter 12: The knee:560-563
- Chapmans. M.W. Chapmans Orthopaedic Surgery. Third Edition. Section IV: 2300-2301
- Canale S.T. Cambell’s Operative Orthopaedics tenth edition, Part XIV: 2528-2529
- Kapandji. The Physiology of the Joints. Volume Two,1987:94-97
- McMurray T.P. The semilunar cartilages. Br J Surg. 1942;29:407-414
- Apley A.G. The diagnosis of meniscus injuries: Some new clinical methods. J Bone Joint Surg Am. 1947;29:78-84