In U.S. high-school-aged athletes, knee injuries account for approximately 29% of all severe sports injuries (>21 d missed sport participation), 54% of severe sports injuries requiring surgery (10), and 10% to 25% of all adolescent injuries (18). The rate of meniscal tears is estimated to be 61 per 100,000 in the general population, approximately one-third are sports-related, and they occur in almost all sports (5). Traumatic tears occurring in the menisci periphery are seen in patients younger than 30 yr, while more complex tears and degenerative patterns tend to occur in patients older than 30 (25).
Given the magnitude of the problem, it is important for clinicians to diagnose meniscal tears accurately. Although imaging techniques have a role to play in confirming the diagnosis, it is inappropriate for clinicians to order imaging tests for every patient with a knee injury. This is because every test has false positive and false negative results, and if an imaging test shows a meniscal tear in a patient with no signs of a meniscal tear, it likely represents a false positive test and should not be operated on. Therefore, clinicians must be able to obtain a proper history and physical examination in order to appropriately select the patients where imaging will be helpful. There have been several recent meta-analyses on the strengths of the physical examination in this context (13,22,27). Unfortunately, the conclusion of each article is that there is no method to diagnose a meniscal tear that has an "acceptable" sensitivity and specificity (Table 1). Although this appears discouraging at first, upon further reflection, it actually is an expected finding. In this article, we argue that it is not appropriate to use the meta-analytic approach towards diagnostic tests to evaluate the physical examination for meniscal tears.
First, we briefly review the context in which a meta-analytical diagnostic test approach is expected to provide valuable information. Second, we show that the context of a physical examination test for meniscal tears (as for most conditions) is different than diagnostic tests, and therefore the meta-analysis almost always will yield unfavorable results. Because of these limitations, we have chosen a different approach for this review and will not attempt to pronounce one test as superior to another. Rather, we provide the clinician with a list of each of the different published physical signs, and we categorize them according to exactly what they are testing from a mechanical point of view. By clearly documenting each test along with its strengths and weaknesses, we hope to provide clinicians with a broader scope of tools to assess patients with potential meniscal tears and provide a foundation from which future research should proceed.
THE DIAGNOSTIC TEST APPROACH
Meta-analyses of diagnostic studies are different from meta-analyses of interventions because they are prone to several unique types of biases (28), and more recent methods (26) are not yet being used routinely. In this section, we focus on three underlying concepts of diagnostic test evaluation that are of particular relevance to the physical examination: reporting bias, selection bias, and heterogeneous pathology. We assume that the reader already understands sensitivity (ability to identify true positives), specificity (ability to identify true negatives), positive predictive value (probability of having the disease if a test is positive), negative predictive value (probability of not having the disease if a test is negative), positive likelihood ratio (increase in odds of disease when the test is positive), and negative likelihood ratio (decrease in odds of disease when the test is negative). In addition, readers should consult other references concerning general issues such as the need for a gold standard, blinding, availability of other information, and loss to follow-up (7,8).
When a test is first described, it usually has very good sensitivity and specificity. There are several reasons for this and clinicians need to understand why they need to carefully consider the study before adopting any test.
- Reporting Bias. As with studies on treatments, studies are more likely to be published if the findings are positive. In the case of physical examinations, it is fairly easy to conduct studies if one uses standardized sheets for the physical examination. With the large amount of academic institutions, there probably are many studies showing a test is not good, and these would not be submitted for publication. However, even if a particular test is not very good, a few studies would have promising results by chance alone, and these would be published. This is why confirmatory evidence from independent sources is so important. As an example, Table 1 shows the specificity and sensitivity of four studies (in order of publication time) investigating a relatively new standing test with rotation to detect a torn meniscus (Thessaly test). Note that the later studies report less promising results; "negative" studies become much more important (and publishable) if they suggest current practice is flawed or results from previous studies were exaggerated or biased.
- Selection Bias. This is more specific to the context of physical examinations. Although sensitivity and specificity are not affected by the prevalence of disease, they are affected by the choice of patients in the study (8). This is a subtle but important difference. The logical first step in developing a test is to compare the results in those who clearly have or do not have the disease; if the test does not work in this context, then it is not worth pursuing. If a test is promising, the next step is to evaluate it among the population where it would normally be used, i.e., where it is unclear if disease is present or absent. Unfortunately, the sensitivity and specificity may be very different in this context. For example, an effusion will show very high sensitivity and specificity if the patient population is a mix of 1) healthy young adults with normal knees and 2) young adult patients who experienced knee swelling 24 h after a twisting mechanism while fully weight-bearing. However, if the patient population is over 65 yr with no trauma, effusions will show low sensitivity and specificity because most tears are degenerative and there is a higher prevalence of osteoarthritis (OA). Because the history is essential to understanding the context of the population, and history is a complex construct involving many components, the standard diagnostic test approach towards evaluating the physical examination is problematic.
- Heterogeneity of Pathology. Diagnostic tests are designed to evaluate signs that are specific characteristics of the disease. These characteristics are expected to be present in every patient (or almost every patient) and absent in subjects without disease. For example, a tuberculin skin test is expected to be positive in patients with tuberculosis because every patient has been infected with the bacterium. Some false negative skin tests may occur because the patient does not mount a sufficient immune response to the challenge, and some false positive skin tests may occur because of cross-reactivity. Regardless, the test still assesses a pathology that exists in every patient with disease and absent in patients without disease. This fundamental principle must be present to have excellent diagnostic test properties. In the physical examinations of musculoskeletal conditions (and many other conditions), this rarely applies. Further, it particularly is absent for meniscal pathology where the pathology is quite heterogeneous. For example, a test designed to stress the posterior horn of the meniscus is expected to be negative for anterior horn tears. This leads to a poor sensitivity if all meniscal tears are combined. Similarly, the pain caused by meniscal tests is thought to occur because of increased stress on an inflamed synovium; specificity will be low because there are many different causes of synovial inflammation. Therefore, making comparisons of different tests designed to test different pathologies is not logical.
In summary, only four of the physical examination signs for meniscal tears actually have been investigated using meta-analytical techniques (Table 1). In addition, the results of individual studies for the Thessaly test (Table 1) are indicative of the problems associated with using meta-analyses of diagnostic test studies to evaluate the physical examination. Aside from the issue of reporting bias already mentioned, the Thessaly test appears superior to other tests in some studies but not others, suggesting that the populations or methods or heterogeneity of pathology of the different studies affect the sensitivity and specificity of the test. Further, the true value of any diagnostic test is the added information it provides in the context of what already is known about the patient. For the evaluation of meniscal signs, this includes other aspects of the physical examination in addition to history. For example, lateral joint-line pain of the knee in a 25-yr-old patient is expected if the history or physical strongly suggests an acutely torn anterior cruciate ligament and provides little additional information; the same sign provides important information in a 25-yr-old patient with an acutely injured knee and no ligament damage. Because of all these limitations, a more promising approach may be predictive modeling where one calculates probabilities under a variety of scenarios (9,19,30), but much more work needs to be done in this area. Until new approaches are found, clinicians should not rely on reported sensitivity, specificity, positive and negative predictive values, and positive and negative likelihood ratios to decide which test to use. Rather, we believe their choice of physical examination signs should be based on the logic underlying how different tests stress different parts of the meniscus. In addition, to avoid unnecessary pain, clinicians should first use tests that produce lesser amounts of stress (and hence less pain) on the meniscus (e.g., palpation, rotation without compression), and proceed to tests that produce greater amounts of stress (and hence more pain) on the meniscus (e.g., rotation of the knee with meniscus under compression using full body weight) only if the lesser stress tests do not provide enough information to confirm the presence or absence of a meniscal tear.
PHYSICAL EXAMINATION FOR MENISCAL TEARS
This section summarizes the specific components of the physical examination related to the meniscus. Table 2 groups meniscal signs according to the mechanism by which they distinguish knees with meniscal tears from those without meniscal tears and provides additional details about the differential diagnosis for a positive test. The following text gives a more detailed description for each specific physical examination sign and the acronym that is associated with it.
- Effusion. This will be a useful sign in the context of a 20-yr-old person with trauma, but of minimal help in a 60-yr-old person with an atraumatic injury. Atraumatic injuries usually represent degenerative tears of the meniscus and usually occur in older populations with coexisting OA. Because an effusion is common in patients with OA, its presence is expected even in the absence of a meniscal tear.
- Joint-Line Tenderness. The meniscus itself does not have pain fibers, and the mechanism underlying the tenderness is believed to come from an inflammation of the surrounding capsule. Therefore, any condition that causes inflammation in this area also will cause joint-line tenderness. Furthermore, in the physical examination of the joint line, we are palpating all the structures from the skin to the meniscus. Therefore, damage to any structure that lies between the skin and meniscus (including any joint irritation) also would cause joint-line tenderness.
- Range of Motion (ROM). The lack of terminal extension is considered a sign for an anterior horn tear or bucket-handle tear of the meniscus. A lack of full flexion is considered a sign for a posterior horn tear. That said, there are many reasons for losing ROM. For example, the lack of terminal extension with an anterior horn meniscal tear is caused by the meniscus being pinched anteriorly, which stresses the anterior synovium and results in anterior knee pain. Similarly, lack of flexion due to a posterior meniscal tear should be accompanied by posterior knee pain; the presence of anterior knee pain with flexion is common with large effusions of any cause and is not a specific sign of a meniscal tear.
Special Tests for Meniscal Tears
There are many acronyms for tests developed to diagnose meniscal tears. In general, all of these tests try to compress the meniscus in some way, and some add a rotary component. When inflammation is diffuse, all tests would be expected to be positive. However, if the inflammation is localized, one theoretically would expect tests to identify only those types of meniscal tears where the torn meniscus actually is being compressed or stressed by that particular maneuver. Some authors consider a positive test when there is pain, and some require that a "click" also occur. For brevity, we have not distinguished which authors require a click to be present in order for the test to be considered positive.
- Isolated Anterior Displacement: Finochietto test or Jump Sign (11). This simply is the anterior drawer test of the knee for a tear of the anterior cruciate ligament but the knee is held in 130°-140° of flexion and more force is applied. The examiner feels a "jump" as the torn posterior horn of the meniscus is displaced anterior to the tibio-femoral point of contact: the jump rarely is audible or visible.
- Isolated Valgus or Varus Stress. The pain must be on the side of compression, which distinguishes it from a ligament or capsular sprain that causes pain on the side of distraction.
- Boehler Test (24). The examiner applies a varus force to compress the medial meniscus (or a valgus stress to compress the lateral meniscus) while the knee is almost fully extended. With the knee extended, most of the compressive force will occur in the mid- to anterior section of the meniscus, and therefore, the test theoretically would be most sensitive for tears in these regions.
- Payr Test (24). Similar to the Boehler test, the examiner applies a varus force to compress the medial meniscus. However in this test, the knee is examined at 90° flexion.
- Isolated Rotation: The First Steinmann Sign (6). The examiner holds the knee at 90° (stresses the midsection of the meniscus) and rotates the leg internally and externally to try and "catch" the torn piece of the meniscus.
- Rotation with Flexion or Extension.
- The McMurray's Test (21). This is perhaps the best-known test. As originally described, the knee first is flexed fully. The examiner then externally rotates the leg (and later internally rotates) as it is extended to 90°. Because only the posterior horn to the midsection of the meniscus is under compression within this range, McMurray suggested this test is only to diagnose tears in these regions.
- Bragard Sign (6) or Steinman Sign (also known as Second Steinmann Sign) (3). This test is for anterior meniscal tears and is based on the principle that extending the knee while in rotation will displace the meniscus forward and increase anterior joint line tenderness. Therefore, a positive test is when the joint line tenderness decreases (Bragard) or moves posteriorly (Second Steinman Sign) as the knee is flexed. However, this theory contradicts the knowledge that the meniscus does not have pain fibers, so tenderness is not due to the location of the meniscus. Rather, we believe it is more likely that flexion would reduce tenderness because it tightens the anterior structures between the examiner's hand and the synovium, thereby reducing pressure on the inflamed synovium (similar to how abdominal muscle contraction reduces tenderness in suspected peritonitis from any cause).
- Flexion or Extension with Rotation, and Valgus or Varus Stress (#3 and #4 combined).
- Medial-Lateral Grind Test (2). This test is a combination of the McMurray and Boehler tests (some authors mistakenly attribute the description of this test to only McMurray 14). The examiner applies a valgus force to the knee while internally rotating the leg to further compress the lateral meniscus (or a varus force while externally rotating the leg to further compress the medial meniscus). In some descriptions, the pressure is applied as the knee is extended beyond the 90° of flexion that McMurray originally described. If the knee is extended beyond 90°, the test also should be positive with midsection and anterior horn tears.
- Flexion or Extension with Rotation, and Valgus or Varus Stress, and Axial Loading.
- Pivot-Shift Test (17). This test is only for lateral meniscal tears. It is similar to the medial-lateral grind test in that there is valgus stress and internal rotation of the leg. However, the lateral meniscus is further stressed by applying an axial force to the heel, thereby compressing the meniscus in a cranial-caudal direction. The knee is moved only from terminal extension to 90° flexion (one cannot apply an axial load to the flexed knee) so the compression on the posterior horn of the meniscus is less than in the McMurray test or its variations with valgus or varus stress. Although it never has been described, a logical addition would be to apply an axial load during varus stress of the knee with external rotation of the leg.
- Standing Tests: Thessaly (15), Merkel (6), Weight-Bearing McMurray Test (Ege's Test) (3).
- Thessaly and Merkel. This test is similar in principle to the Pivot Shift test in that it combines an axial load to the valgus or varus stress and rotation of the leg. In this test, the axial load occurs because the patient is standing, and he or she then flexes the knee (20° has been reported to be better than 5° 15) and rotates it him or herself.
- Weight-Bearing McMurray Test. This test is similar to the others in the category, but the feet are kept internally rotated to examine the lateral meniscus and externally rotated to examine the medial meniscus. Anterior horn tears create pain in early knee flexion and posterior horn tears create pain in late knee flexion.
- Childress Test (Duck-Walk) (6). This basically is similar to the standing tests described in 6b, in that it is an active test with compression applied by the patient's weight and movement. The patient attempts a full squat, which applies both a flexion load and an axial load (like the weight-bearing McMurray test). The patient then attempts to "duck walk," which means he or she moves forward without extending the knee very much. In doing so, the patient will also externally and internally rotate the leg. This test applies a considerable compressive force and would not be appropriate in a patient for whom the McMurray test already was positive or where the patient was lacking full flexion.
- Apley Test (4). The Apley test is similar in principle to the other tests in that there is a compressive and rotary force applied to the meniscus. However, it is distinct from the other tests because there are three phases in the original description (all with the patient prone) used to increase the specificity of the test. A full description is provided here because the test has been described as having poor sensitivity and specificity (Table 1), but this may be because the test is not being conducted appropriately; many studies omit important aspects of the test when they describe it in the methods section. In phase 1, the legs are externally rotated and then fully flexed, followed by internal rotation and complete extension; the examiner notes the position in which pain occurred. In phase 2, the examiner places the knee in 90° flexion and applies axial distraction of the leg (thereby removing any compression of the meniscus) and repeats the internal and external rotation. If there is a notable increase in pain, then the synovium or capsule are inflamed and the test cannot be interpreted as specific for a meniscal tear. This important part of the Apley is sometimes not described in articles reviewing examination for meniscal tears of the knee (e.g., 24,29). In phase 3, the examiner applies axial compression to the heel while the knee is held in the amount of flexion that caused pain in phase 1 (this often is not described in articles), and then rotates the leg internally (for lateral meniscus) and externally (for medial meniscus). Again, there should be a noticeable increase in pain for the test to be considered positive and suggestive of a meniscal tear. Readers should note that the test specifically was designed to reduce the number of false positive tests due to capsular sprains and is not designed to have high sensitivity if one interprets a pain on distraction as a "negative" test instead of the more appropriate "at least the presence of capsular or ligament sprain."
There are many different clinical examination tests that stress the meniscus in different areas and to varying degrees. It is unrealistic to expect that one test will be superior to all others for all types of pathologies. Further, stressing the torn meniscus causes pain. A logical approach is to use the less stressful (less painful) tests first, and proceed to more stressful (more painful) tests if the former are negative. For example, if the examiner finds that the supine patient has pain with passive full flexion of the knee, it is not necessary to ask the patient to duck walk (full flexion while weight-bearing). An awareness of all the different methods that stress the meniscus provides the clinician with additional tools to diagnose tears, and future research should explore predictive models based on the different components accounting for the different contexts of patients.
1. Akseki D, Ozcan O, Boya H, Pinar H. A new weight-bearing meniscal test and a comparison with McMurray's test and joint line tenderness. Arthroscopy
. 2004; 20:951-8.
2. Anderson AF, Lipscomb AB. Clinical diagnosis of meniscal tears. Description of a new manipulative test. Am. J. Sports Med.
3. Anonymous. Injuries of the semilunar cartilage (Article 45) [abstract of H. Prinz. Beitr. Klin. Chir
. 28 Apr 1937: 337]. Epitome of the current medical literature. Br. Med. J.
4. Apley AG. The diagnosis of meniscus injuries. J. Bone Joint Surg.
5. Baker BE, Peckham AC, Pupparo R, Sanborn JC. Review of meniscal injury and associated sports. Am. J. Sports Med
. 1985; 13:1-4.
7. Bossuyt PM, Reitsma JB, Bruns DE, et al
. Towards complete and accurate reporting of studies of diagnostic accuracy: the STARD initiative. Br. Med. J.
8. Bossuyt PM, Reitsma JB, Bruns DE, et al
. The STARD statement for reporting studies of diagnostic accuracy: explanation and elaboration. Ann. Intern. Med.
9. Breiman L, Friedman JH, Olshen RA, Stone CJ. Classification and Regression Trees
. Wadsworth, 1984.
10. Darrow CJ, Collins MA, Yard EE, Comstock RD. Epidemiology of severe injuries among United States high school athletes 2005-2007. Am. J. Sports Med.
11. Finochietto R. Semilunar cartilages of the knee: the "jump sign". J. Bone Joint Surg. Am.
12. Harrison BK, Abell BE, Gibson TW. The Thessaly test for detection of meniscal tears: validation of a new physical examination technique for primary care medicine. Clin. J. Sports Med.
13. Hegedus EJ, Cook C, Hasselblad V, et al
. Physical examination tests for assessing a torn meniscus in the knee: a systematic review with meta-analysis. J. Orthop. Sports Phys. Ther.
14. Jackson JL, O'Malley PG, Kroenke K. Evaluation of acute knee pain in primary care. Ann. Intern. Med.
15. Karachalios T, Hantes M, Zibis AH, et al
. Diagnostic accuracy of a new clinical test (the Thessaly test) for early detection of meniscal tears. J. Bone Joint Surg. Am.
16. Konan S, Rayan F, Haddad FS. Do physical diagnostic tests accurately detect meniscal tears? Knee Surg. Sports Traum. Arth.
17. Kurosaka M, Yagi M, Yoshiya S, et al
. Efficacy of the axially loaded pivot shift test for the diagnosis of a meniscal tear. Int. Orthop.
18. Louw QA, Manilall J, Grimmer KA. Epidemiology of knee injuries among adolescents: a systematic review. Br. J. Sports Med.
19. Lowery DJ, Farley TD, Wing DW, et al
. A clinical composite score accurately detects meniscal pathology. Arthroscopy
. 2006; 22:1174-9.
20. Malanga GA, Andrus S, Nadler SF, McLean J. Physical examination of the knee: a review of the original test description and scientific validity of common orthopedic tests. Arch. Phys. Med. Rehabil.
21. McMurray TP. The semilunar cartilages. Br. J. Surg.
22. Meserve BB, Cleland JA, Boucher TR. A meta-analysis examining clinical test utilities for assessing meniscal injury. Clin. Rehabil.
23. Mirzatolooei F, Yekta Z, Bayazidchi M, et al
. Validation of the Thessaly test for detecting meniscal tears in anterior cruciate deficient knees. Knee
. 2010; 17(3):221-3.
24. Muellner T, Nikolic A, Vecsei V. Recommendations for the diagnosis of traumatic meniscal injuries in athletes. Sports Med.
25. Poehling GG, Ruch DS, Chabon SJ. The landscape of meniscal injuries. Clin. Sports Med.
26. Reitsma JB, Glas AS, Rutjes AW, et al. Bivariate analysis of sensitivity and specificity produces informative summary measures in diagnostic reviews. J. Clin. Epidemiol.
27. Scholten RJ, Deville WL, Opstelten W, et al
. The accuracy of physical diagnostic tests for assessing meniscal lesions of the knee: a meta-analysis. J. Fam. Pract.
28. Smidt N, Deeks JJ, Moore T. Cochrane Handbook for Diagnostic Test Accuracy Reviews
. The Cochrane Collaboration, 2009.
29. Solomon DH, Simel DL, Bates DW, et al
. The rational clinical examination. Does this patient have a torn meniscus or ligament of the knee? Value of the physical examination. JAMA
. 2001; 286:1610-20.
30. Wagemakers HP, Heintjes EM, Boks SS, et al
. Diagnostic value of history-taking and physical examination for assessing meniscal tears of the knee in general practice. Clin. J. Sport Med.