[

RESEARCH REPORT

MICHAEL P. REIMAN, PT, DPT, OCS, SCS, ATC, FAAOMPT, CSCS1 JANICE K. LOUDON, PT, PhD, SCS, ATC, CSCS1 ADAM P. GOODE, PT, DPT, PhD1

Diagnostic Accuracy of Clinical Tests for Assessment of Hamstring Injury: A Systematic Review

amstring injuries are one of the most common soft tissue injuries in athletes.4,8-10,31,35,44 Treatment and management of hamstring injuries,20,24,25 as well as injury prevention and return to sport,1,4,6,16,30,32 have received signicant research effort in the past 10 years. Examination of athletes with
T STUDY DESIGN: Systematic literature review.
injury has traditionally relied on various clinical measures (eg, palpation, swelling, manual resistance), as well as the use of diagnostic imaging. But a few studies have suggested the use of specic clinical tests that may be helpful for the diagnostic process. potential articles, of which only 3 articles met the inclusion criteria, with only 1 of these 3 articles being of high quality. Two of the studies investigated a single special test, whereas the third article examined a composite clinical assessment employing various special tests. The SN values ranged from 0.55 (95% condence interval [CI]: 0.46, 0.69) for the active range-of-motion test to 1.00 (95% CI: 0.97, 1.00) for the taking-off-the-shoe test. The SP values ranged from 0.03 (95% CI: 0.00, 0.22) for the composite clinical assessment to 1.00 (95% CI: 0.97, 1.00) for the taking-off-the-shoe test, active range-of-motion test, passive range-of-motion test, and resisted range-of-motion test. The use of a single special test demonstrated stronger SP than SN properties, whereas the composite clinical assessment demonstrated stronger SN than SP properties.

T BACKGROUND: The diagnosis of a hamstring

T OBJECTIVE: To summarize the current

literature on the diagnostic accuracy of orthopaedic special tests for hamstring injuries and to determine their clinical utility.

T METHODS: A computer-assisted literature

search of the MEDLINE, CINAHL, and Embase databases (along with a manual search of grey literature) was conducted using key words related to diagnostic accuracy of hamstring injuries. To be considered for inclusion in the review, the study required (1) patients with hamstring or posterior thigh pain; (2) a cohort, case-control, or cross-sectional design; (3) inclusion of at least 1 clinical examination test used to evaluate hamstring pathology; (4) comparison against an acceptable reference standard; (5) reporting of diagnostic accuracy of the measures (sensitivity [SN], specicity [SP], or likelihood ratios); and (6) publication in English. SN, SP, and positive and negative likelihood ratios were calculated for each diagnostic test.

T CONCLUSION: Very few studies have investi-

gated the utilization of clinical special tests for the diagnosis of hamstring injuries. Further studies of higher quality design are suggested prior to advocating independent clinical utilization of these special tests.

T LEVEL OF EVIDENCE: Diagnosis, level 3b. J

Orthop Sports Phys Ther 2013;43(4):222-231. Epub 14 January 2013. doi:10.2519/jospt.2013.4343

T RESULTS: The search strategy identied 602

1

T KEY WORDS: diagnosis, sensitivity, specicity,

strain

hamstring injury has traditionally relied on combinations of pain with palpation of the injured area, traditional manual muscle testing, passive straight leg raise testing,3-6,12,24,41 magnetic resonance imaging (MRI),4,5,34,38,41 and isokinetic testing.34 However, the studies on diagnostic accuracy of palpation, traditional manual muscle testing, and leg raise testing have not provided sufficient information to quantify the clinical ability of these tests to differentiate between those with and without conrmed hamstring injury. In addition, some of these tests have been described only for assessment of readiness for return to sport.2-5 MRI4,5,34,38,41 and ultrasonography (US) are considered the criterion reference standards for diagnosis of hamstring injuries.7,13,38 However, both MRI and US are not practical alternatives for diagnosis of hamstring injury due to the high incidence of this injury and the costs associated with these diagnostic tests. Therefore, clinical tests with strong psychometric properties for use in diagnosing this condition are needed. The purpose of this study was to conduct a systematic review of the literature reporting on the diagnostic accuracy of clinical tests that have been proposed to be helpful in the diagnosis of hamstring injury. The studies included had cohort, casecontrol, and/or cross-sectional designs

Doctor of Physical Therapy Division, Department of Community and Family Medicine, Duke University School of Medicine, Durham, NC. The authors certify that they have no affiliations with or nancial involvement in any organization or entity with a direct nancial interest in the subject matter or materials discussed in the manuscript. Address correspondence to Dr Michael P. Reiman, Duke University School of Medicine, Duke University Medical Center, Doctor of Physical Therapy Division, DUMC 104002, 2200 West Main Street, Suite B 230, Durham, NC 27705. E-mail: reiman.michael@gmail.com t Copyright 2013 Journal of Orthopaedic & Sports Physical Therapy

222 | april 2013 | volume 43 | number 4 | journal of orthopaedic & sports physical therapy

TABLE 1

Quality Assessment of the Studies Included in the Review

Item*

Article Cacchio et al11 Schneider-Kolsky et al37 Zeren and Oztekin45

1 N/U Y N/U

2 Y Y N/U

3 Y Y Y

4 N/U Y N/U

5 Y Y Y

6 Y Y Y

7 Y Y Y

8 Y Y Y

9 Y Y N/U

10 N/U Y N/U

11 N/U N/U N/U

12 Y Y Y

13 N/U N/U N/U

14 N/U Y N/U

Total 8 12 6

Abbreviations: N, no; U, unclear; Y, yes. *Item 1: was the spectrum of patients representative of those in clinical practice? Item 2: were selection criteria clearly described? Item 3: is the reference standard likely to classify the target condition correctly? Item 4: is the period of time between the reference standard and index test acceptable? Item 5: did the whole sample of patients receive verication using the reference standard? Item 6: did patients receive the same reference standard regardless of the index test result? Item 7: was the reference standard independent of the index test? Item 8: was the execution of the index test described in sufficient detail for replication? Item 9: was the execution of the reference standard described in sufficient detail for replication? Item 10: were the index test results interpreted without knowledge of the reference standard? Item 11: was the reference standard interpreted without knowledge of the results of the index test? Item 12: were the same clinical criteria available when test results were interpreted as would be in clinical practice? Item 13: were uninterpretable/intermediate test results reported? Item 14: were withdrawals from the study explained?

that enabled comparison of the diagnostic accuracy of clinical tests to their appropriate criterion reference standards.

METHODS

he PRISMA guidelines were utilized during the search-and-reporting phase of this review. The PRISMA statement includes a 27-item checklist designed to be used as a basis for reporting systematic reviews of randomized trials, 33 but can also be applied to multiple forms of research methodologies.40

Search Strategy
A systematic, computerized search of the literature in the MEDLINE, CINAHL, and Embase databases was conducted in February 2012. The MeSH search string in MEDLINE was (((hamstring[ti] OR semitendinosus[ti] OR semimembranosus[ti] OR posterior thigh[ti] OR biceps femoris[ti])) AND ((strain) OR strained OR (tear) OR tears OR (injury) OR injuries AND evaluation OR physical examination OR orthopedic clinical examination OR diagnosis OR diagnose)) NOT (cruciate ligament[ti] OR ACL[ti] OR PCL[ti]), with limits for English language and humans. Two authors (M.P.R. and J.K.L.) independently performed

the search. Because computerized search results for diagnostic accuracy data frequently omit relevant articles,19 the reference lists of all selected publications were checked to retrieve relevant publications that were not identied in the computerized search. The grey literature, which included publications, posters, abstracts, or conference proceedings, was hand searched. The reference lists and grey literature were searched by 1 author (M.P.R.). To identify relevant articles, titles and abstracts of all identied citations were independently screened by both authors. Full-text articles were retrieved if the abstract provided insufcient information to establish eligibility or if the article passed the rst eligibility screening.

Selection Criteria
Articles examining clinical tests for hamstring injuries were eligible if they met all of the following criteria: (1) patients presented with hamstring or posterior thigh pain; (2) a cohort, case-control, or cross-sectional design was used; (3) the study included at least 1 clinical examination test to evaluate hamstring pathology; (4) the results of the clinical test were compared against an acceptable reference standard (MRI or US)7,13,38; (5) the study reported diagnostic accuracy of the measures (sensitivity [SN], specicity

[SP], positive likelihood ratio [+LR], and negative likelihood ratio [LR]); and (6) the study was published in English. An article was excluded if (1) the reported pathology was associated with a condition located elsewhere (eg, lumbar spine) that referred pain to the hamstring/posterior thigh, (2) the study did not provide either SN or SP data, (3) the clinical examination test was performed under any form of anesthesia or on cadavers, (4) the study used specialized instrumentation not readily available to all clinicians, and (5) the study was performed on infants/toddlers. All criteria were independently applied by both reviewers to the full text of the articles that passed the rst eligibility screening. Disagreements among the reviewers were discussed and resolved during a consensus meeting.

Quality Assessment
The Quality Assessment of Diagnostic Accuracy Studies (QUADAS) tool43 was used to determine the quality of the studies. The QUADAS consists of 14 items (TABLE 1), each with response categories of yes, no, or unclear. A yes score indicates sufficient information, with bias considered unlikely; a no score indicates sufficient information, but with potential bias from inadequate design or conduct; and an unclear score indicates that the article

journal of orthopaedic & sports physical therapy | volume 43 | number 4 | april 2013 | 223

[
or methodology provided insufficient information or the methodology was unclear. The total score was a count of all of the criteria that scored yes (valued as 1, whereas no and unclear scores were valued as zero), with a maximum attainable score of 14. The methodological quality of each of the studies was independently assessed by both reviewers. Disagreements were discussed and resolved during a consensus meeting. Qualitatively, studies that exhibit higher QUADAS values are associated with less risk of design bias than those with lower values. Similar to previously published reviews, the studies were stratied as high quality/low risk of bias if their QUADAS score was 10 or greater or as low quality/high risk of bias if their QUADAS score was less than 10.23

RESEARCH REPORT
Identication 915 abstracts identied through MEDLINE (n = 596), CINAHL (n = 143), and Embase (n = 176)

]
8 abstracts identied through hand search

602 titles included after duplicates removed Screening 576 abstracts rejected because each did not reect diagnosis 18 articles rejected for providing quantitative numbers that did not allow measurement of sensitivity or specicity 5 articles rejected for failing to calculate diagnostic accuracy or failing to report both sensitivity and specicity

602 titles screened

26 abstracts screened

Data Extraction
One author (M.P.R.) independently gathered data regarding study population, setting, special test performance, pathology, diagnostic reference standard, and number of true positives, false positives, false negatives, and true negatives for calculation of SN, SP, +LR, and LR when these were not provided. The other authors (J.K.L. and A.P.G.) veried data extraction accuracy once completed. Cell counts of zero are common in diagnostic accuracy studies, and in such instances 0.5 was added to all cells, as suggested by Cox.15 SN is dened as the percentage of people who test positive for a specic disease among a group of people who have the disease. SP is the percentage of people who test negative for a specic disease among a group of people who do not have the diagnosis/disorder. A +LR is the ratio of a positive test result in people with the pathology to a positive test result in people without the pathology. A +LR identies the strength of a test in determining the presence of a nding, and is calculated by the formula SN/(1 SP). A LR is the ratio of a negative test result in people with the pathology to a negative test result in people without the pathology, and is calculated by the for8 full text articles screened Inclusion 3 studies included in the qualitative analysis
FIGURE. Flow diagram for study inclusion.

mula (1 SN)/SP. The higher the +LR and the lower the LR, the greater the posttest probability is altered. Posttest probability can be altered to a minimal degree with +LRs of 1.0 to 2.0 or LRs of 0.5 to 1.0, to a small degree with +LRs of 2.0 to 5.0 or LRs of 0.2 to 0.5, to a moderate degree with +LRs of 5.0 to 10.0 and LRs of 0.1 to 0.2, and to a large and almost conclusive degree with +LRs greater than 10.0 and LRs less than 0.1. Pretest probability is dened as the probability of the target disorder before a diagnostic test result is known. It represents the probability that a specic patient with a specic past history, presenting to a specic clinical setting with a specic symptom complex, has a specic diagnosis.26

RESULTS

Selection of Studies
he systematic search through MEDLINE, CINAHL, and Embase netted 915 abstracts, and 8 additional papers were identified through an extensive hand search. In total, 602 titles were initially retained after duplicates were removed. Abstract and full-text review reduced the acceptable papers to 3 ( FIGURE, TABLE 2). The sample sizes of the 3 studies were 46, 11 140,45 and 58 37 athletes, respectively. Cacchio et al11 and Zeren and Oztekin45 investigated individual special tests, whereas Schneider-Kolsky et al 37 employed a composite clinical assessment.

224 | april 2013 | volume 43 | number 4 | journal of orthopaedic & sports physical therapy

TABLE 2
Author/Test

Summary of Studies Included in the Review

Subjects Sport Sprinters, distance runners, long jumpers, hurdlers, soccer and rugby players Symptom Duration 15.0 7.2 mo

Cacchio et al11 46 symptomatic athletes (mean  Puranen-Orava test SD age, 22.8 2.3 y; height,  Bent-knee stretch test 1.79 m; weight, 76.8 11.3 kg;  Modied bent-knee 12 females, 34 males) stretch test 46 asymptomatic athletes (mean SD age, 23.2 1.8 y; height, 1.78 m; weight, 75.9 12.6 kg; 11 females, 35 males) Zeren and Oztekin45  Taking-off-the-shoe test  Active rangeof-motion test  Passive rangeof-motion test  Resisted rangeof-motion test Schneider-Kolsky et al37  Composite clinical assessment 140 professional male soccer players (age range, 17-33 y) with a history and clinical ndings of a proximal hamstring muscle strain injury

Soccer

Examined between 0 and 42 d from injury

58 professional footballers who received a diagnosis from the trainer or physician of an acute hamstring injury (mean SD age, 24 3.8 y; height, 186.3 6.0 cm; weight, 87.9 8.7 kg)

Professional footballers playing in the Australian Football League

Examined within 3 d of injury

Quality Scores
The kappa value between testers for the overall score using the QUADAS was 0.68 (95% condence interval [CI]: 0.44, 0.91), with this point estimate reecting substantial agreement.27 Of the individual items of the QUADAS, items 1, 3, 5, 6, 7, 8, 12, and 14 had 100% agreement; and items 2, 4, 9, 10, 11, and 13 had 83% agreement between raters. Quality scores for each of the studies are shown in TABLE 3. Using our previously established stratication of the QUADAS, the Schneider-Kolsky et al37 article was considered of high quality/low risk of bias, whereas the Cacchio et al11 and the Zeren and Oztekin45 articles had a QUADAS score of less than 10 points, suggesting low quality/high risk of bias (TABLE 3). The most poorly scored items of the QUADAS were items 1 (spectrum representative of those in clinical practice), 4 (time period between reference standard and index test), 10 (index test results interpretation without knowledge of reference standard), 11 (reference standard interpre-

tation without knowledge of index test results), 13 (uninterpretable test results), and 14 (explanation of withdrawals).

Diagnostic Clinical Tests

Schneider-Kolsky et al37 and Cacchio et al11 used MRI as the reference standard, whereas Zeren and Oztekin,45 the study with the lowest score on the QUADAS, used diagnostic US. Based on recent advances in technology, diagnostic US is now considered comparable to MRI for the diagnosis of muscle injury.7,13,38 Seven individual special tests were investigated in the Cacchio et al11 and Zeren and Oztekin45 studies. Schneider-Kolsky et al37 used a composite method based on the interpretation of 3 special tests (TABLE 3, APPENDIX). Cacchio et al11 examined the PuranenOrava, bent-knee stretch, and modied bent-knee stretch tests in 46 symptomatic and 46 asymptomatic athletes. The interrater reliability of the tests, based on intraclass correlation coefficients, was 0.84 (95% CI: 0.72, 0.87) for the

Puranen-Orava test, 0.86 (95% CI: 0.78, 0.93) for the bent-knee stretch test, and 0.88 (95% CI: 0.82, 0.94) for the modied bent-knee stretch test. The PuranenOrava test was determined to have an SP of 0.82 (95% CI: 0.68, 0.92) and an SN of 0.76 (95% CI: 0.61, 0.87). The bentknee stretch test had SN and SP values of 0.84 (95% CI: 0.71, 0.93) and 0.87 (95% CI: 0.73, 0.95), respectively. The modied bent-knee stretch test had SN and SP values of 0.89 (95% CI: 0.76, 0.96) and 0.91 (95% CI: 0.79, 0.97), respectively. Zeren and Oztekin45 examined 4 individual special tests for proximal hamstring muscle strain injury in 140 male professional soccer players: the takingoff-the-shoe test, active range-of-motion test, passive range-of-motion test, and resisted range-of-motion test. The criterion standard utilized by the authors was US. Reliability of the tests was not determined in this study. SP for all of these tests was 1.00 (95% CI: 0.97, 1.00). All 140 noninvolved legs (control side) tested negative, resulting in an SP of 1.00 (95% CI: 0.97, 1.00) for all of these tests.45 The SN for these tests ranged from 0.55 (95% CI: 0.46, 0.63) for the active range-ofmotion test to 1.00 (95% CI: 0.97, 1.00) for the taking-off-the-shoe test. Schneider-Kolsky et al37 investigated the diagnostic accuracy of a composite clinical assessment in 58 professional footballers (rugby) with hamstring injuries, with a positive test result being the reproduction of the patients concordant pain/stiffness during any of the 3 individual tests. Reliability of the testing was not investigated. The composite clinical assessment had an SN of 0.95 (95% CI: 0.83, 0.99) and an SP of 0.03 (95% CI: 0.00, 0.22).

DISCUSSION

ur study investigated the diagnostic accuracy of selected orthopaedic special tests for hamstring injury. There were only 3 studies illustrating tests that included both SN and SP values. Our review also found limited

journal of orthopaedic & sports physical therapy | volume 43 | number 4 | april 2013 | 225

[
TABLE 3
Author/Test Cacchio et al11 Puranen-Orava test Bent-knee stretch test Modied bent-knee stretch test Zeren and Oztekin45 Taking-off-the-shoe test Active range-of-motion test Passive range-of-motion test Resisted range-of-motion test Schneider-Kolsky et al37 Composite clinical assessment

RESEARCH REPORT

Summary of Studies Reporting on the Diagnostic Accuracy of Orthopaedic Special Tests for Hamstring Pathologies*
SN 0.76 (0.61, 0.87) 0.84 (0.71, 0.93) 0.89 (0.76, 0.96) 1.00 (0.97, 1.00) 0.55 (0.46, 0.63) 0.57 (0.49, 0.66) 0.61 (0.52, 0.69) 0.95 (0.83, 0.99) SP 0.82 (0.68, 0.92) 0.87 (0.73, 0.95) 0.91 (0.79, 0.97) 1.00 (0.97, 1.00) 1.00 (0.97, 1.00) 1.00 (0.97, 1.00) 1.00 (0.97, 1.00) 0.03 (0.00, 0.22) +LR 4.2 (NR) 6.5 (NR) 9.9 (NR) 280.0 (17.6, 4454.6) 154.6 (9.7, 2468.7) 160.6 (10.1, 2564.0) 170.6 (10.7, 2722.9) 0.97 (0.88, 1.08) LR 0.29 (NR) 0.18 (NR) 0.12 (NR) 0.00 (0.00, 0.06) 0.50 (0.38, 0.54) 0.43 (0.36, 0.52) 0.40 (0.32, 0.49) 1.9 (0.2, 16.0) QUADAS 8 6 12 ICC 0.84 (0.72, 0.87) 0.86 (0.78, 0.93) 0.88 (0.82, 0.94) NR NR NR NR NR

Abbreviations: ICC, intraclass correlation coefficient for interrater reliability; LR, negative likelihood ratio; NR, not reported; +LR, positive likelihood ratio; QUADAS, Quality Assessment of Diagnostic Accuracy Studies; SN, sensitivity; SP, specicity. *Values in parentheses are 95% condence interval, except for QUADAS. Diagnostic accuracy calculations reported by authors of this systematic review.

quality of these studies. The diagnostic accuracy of the tests investigated in this study was quite variable, with SN values ranging from 0.55 (95% CI: 0.46, 0.63) to 1.00 (95% CI: 0.97, 1.00) and SP values ranging from 0.03 (95% CI: 0.00, 0.22) to 1.00 (95% CI: 0.97, 1.00). The paucity of studies precluded meta-analysis. The study by Schneider-Kolsky et al,37 the only investigation that utilized composite clinical testing, had less potential for bias (as demonstrated by a higher QUADAS score) than the Cacchio et al11 and Zeren and Oztekin45 studies. Although this study had the highest quality and least potential for bias of the 3 studies in this review, it had the weakest ability to determine a diagnosis. Although this study had a high SN value of 0.95 (95% CI: 0.83, 0.99), it only altered the posttest probability of a diagnosis to a degree less than minimal, with a +LR of 0.97 and a LR of 1.9. The subjects in this study were examined more acutely than those in the Cacchio et al11 and Zeren and Oztekin45 studies (within 3 days of onset), and the sample size was moderate compared to the other 2 studies. Sample sizes in the Cacchio et al11 and Schneider-Kolsky et al37 studies were much smaller than that in the Zeren and Oztekin study.45 Although Zeren and

Oztekin45 commendably investigated a fairly large sample of 140 male soccer players (using the noninvolved limb as a control), the study scored the lowest (6/14) of the 3 studies on the QUADAS. Despite the fact that the taking-off-theshoe test demonstrated 1.00 (95% CI: 0.97, 1.00) SN and SP, as well as the fact that the +LR and LR values for this test were suggestive of increasing and decreasing, respectively, posttest probability of a hamstring injury diagnosis almost conclusively, the zero cell counts for false positive and false negative resulted in substantially large CIs for both likelihood ratio values. The other tests from this study demonstrated LR values in the 0.4-to-0.5 range. Pretest-to-posttest probability shifts of ruling out a diagnosis of hamstring injury for these other tests would, therefore, be small and of questionable clinical utility. In contrast, the QUADAS scores for the Cacchio et al11 and Schneider-Kolsky et al37 studies were higher, potentially resulting in a lower risk of bias. However, the sample sizes and diagnostic accuracy values in these studies were smaller than those in Zeren and Oztekin.45 As previously mentioned, the time frame from injury onset to examination was variable in all 3 studies. The

Schneider-Kolsky et al37 and Zeren and Oztekin45 studies examined the more acutely injured athlete, compared to the Cacchio et al11 study. Therefore, the acuteness of the injury, like the type of injury, as previously discussed, could be a confounding factor in the diagnosis of hamstring injury. The assessment of posterior thigh pain may be complex on occasion. Once red ags are ruled out (previous history of cancer, age of onset less than 20 or greater than 55 years old, saddle anesthesia, and so on),17 a detailed subjective history can help rule out signs and symptoms inconsistent with hamstring injury. Additionally, examination of the lumbar spine, pelvis, and related nervous system may assist in ruling out these areas as potential pain generators. Lumbar spine contribution to posterior thighrelated pain could appropriately be ruled out (SN, 0.92)18 or ruled in (SP, 0.94)28 with repeated motions for potential lumbar spine radiculopathy. Orthopaedic special tests for this same purpose would include the slump test (SN, 0.83)39 and straight leg raise test (SN, 0.97).42 Sacroiliac joint dysfunction and piriformis syndrome could be ruled out with cluster testing (SN, 0.91)29 and the FAIR test (SN, 0.88-0.97),21 respectively, as

226 | april 2013 | volume 43 | number 4 | journal of orthopaedic & sports physical therapy

potential contributors to posterior thigh pain. Orthopaedic special tests for those patients with suspicion of a hamstring injury would, therefore, be valuable in identifying those patients with a hamstring injury. Hence, the clinical utility of the various orthopaedic special tests investigated in this review requires careful consideration. Future studies should concentrate on investigating a set of tests with good diagnostic accuracy to either rule in or rule out hamstring injury as a potential cause of posterior thigh pain. Such tests, singly or in a cluster, could then complement other tests that have been shown to be useful to identify posterior thigh pain due to the lumbar spine, sacroiliac joint, and piriformis syndrome, as previously mentioned. The diagnostic accuracy of these hamstring injury orthopaedic special tests would better be determined by future investigations with less bias. Future studies controlling for injury acuteness assessment, blinding the results of diagnostic imaging, and reporting the reasons for study participant withdrawal would help limit this bias.

CONCLUSION

terior thigh pain attributable to hamstring injury.

ACKNOWLEDGEMENTS: We would like to thank

Limitations
A limitation of the present study is its use of stratied QUADAS scores to assess study quality. Although previous studies have used QUADAS summary scores,14,22,23 others have cautioned against the use of a dedicated quality score.36 The 1 study ranked as high quality could have potentially been inated, as the QUADAS does not qualitatively score for sample size or a case-control design. The use of different reference standards may be a limitation in this review, although diagnostic US has been suggested to be as useful as, and more cost-effective than, MRI as a reference standard.13,38 Additional limitations include limiting the studies to publication in English and having only 1 author search the grey literature and pull data points, which increases the risk of potential error. However, the other authors did verify the data points.

here are a limited number of studies and, therefore, tests that investigate the diagnostic accuracy of orthopaedic special tests for hamstring injury in the athletic population. The diagnostic accuracy of these orthopaedic special tests is quite variable. The Puranen-Orava, bent-knee stretch, and modied bent-knee stretch tests were found to alter posttest probability of a diagnosis to a small to moderate degree. The taking-off-the-shoe test was found to alter posttest probability to an almost conclusive degree, although the study investigating this test demonstrated the potential for signicant bias. The use of a composite clinical assessment, although demonstrating high SN, only alters posttest probability to a degree less than minimal. Caution should be used when utilizing orthopaedic special tests for the diagnosis of hamstring injury, as diagnostic accuracy of these tests is not well established. A comprehensive clinical examination for diagnosis of posterior thigh pain attributable to hamstring injury that excludes other potential pain generators, versus reliance on these tests alone for decisive diagnostic clinical practice, is suggested. t

Carly Reiman for serving as a model; Holly R. Thompson, BA for her review; and Leila Ledbetter, MLIS for assisting with the literature search for this study.

REFERENCES
1. A  rnason A, Andersen TE, Holme I, Engebretsen L, Bahr R. Prevention of hamstring strains in elite soccer: an intervention study. Scand J Med Sci Sports. 2008;18:40-48. http://dx.doi. org/10.1111/j.1600-0838.2006.00634.x 2.  Askling C, Saartok T, Thorstensson A. Type of acute hamstring strain affects exibility, strength, and time to return to pre-injury level. Br J Sports Med. 2006;40:40-44. http://dx.doi. org/10.1136/bjsm.2005.018879 3.  Askling CM, Nilsson J, Thorstensson A. A new hamstring test to complement the common clinical examination before return to sport after injury. Knee Surg Sports Traumatol Arthrosc. 2010;18:1798-1803. http://dx.doi.org/10.1007/ s00167-010-1265-3 4.  Askling CM, Tengvar M, Saartok T, Thorstensson A. Acute rst-time hamstring strains during high-speed running: a longitudinal study including clinical and magnetic resonance imaging ndings. Am J Sports Med. 2007;35:197-206. http://dx.doi.org/10.1177/0363546506294679 5.  Askling CM, Tengvar M, Saartok T, Thorstensson A. Acute rst-time hamstring strains during slow-speed stretching: clinical, magnetic resonance imaging, and recovery characteristics. Am J Sports Med. 2007;35:1716-1724. http:// dx.doi.org/10.1177/0363546507303563 6.  Askling CM, Tengvar M, Saartok T, Thorstensson A. Proximal hamstring strains of stretching type in different sports: injury situations, clinical and magnetic resonance imaging characteristics, and return to sport. Am J Sports Med. 2008;36:1799-1804. http://dx.doi. org/10.1177/0363546508315892 7.  Blankenbaker DG, Tuite MJ. Temporal changes of muscle injury. Semin Musculoskelet Radiol. 2010;14:176-193. http://dx.doi. org/10.1055/s-0030-1253159 8.  Brooks JH, Fuller CW, Kemp SP, Reddin DB. Epidemiology of injuries in English professional rugby union: part 1 match injuries. Br J Sports Med. 2005;39:757-766. http://dx.doi. org/10.1136/bjsm.2005.018135 9.  Brooks JH, Fuller CW, Kemp SP, Reddin DB. Epidemiology of injuries in English professional rugby union: part 2 training injuries. Br J Sports Med. 2005;39:767-775. http://dx.doi. org/10.1136/bjsm.2005.018408 10.  Brooks JH, Fuller CW, Kemp SP, Reddin DB.

KEY POINTS
FINDINGS: The findings from the few stud-

ies that have looked at clinical diagnosis of hamstring injury suggest that single clinical examination orthopaedic special tests demonstrate stronger diagnostic than screening capability. IMPLICATIONS: Due to a dearth of, and potential bias in, the current literature, it is apparent that there is a need for high-quality diagnostic accuracy studies of clinical orthopaedic special tests for diagnosis of hamstring injury. CAUTION: Very few studies have investigated the application of clinical orthopaedic special tests for hamstring injury. Further investigation is warranted prior to suggesting these tests as unequivocal methods of clinical assessment for pos-

journal of orthopaedic & sports physical therapy | volume 43 | number 4 | april 2013 | 227

[
11. Incidence, risk, and prevention of hamstring muscle injuries in professional rugby union. Am J Sports Med. 2006;34:1297-1306. http://dx.doi. org/10.1177/0363546505286022 Cacchio A, Borra F, Severini G, et al. Reliability  and validity of three pain provocation tests used for the diagnosis of chronic proximal hamstring tendinopathy. Br J Sports Med. 2012;46:883-887. http://dx.doi.org/10.1136/ bjsports-2011-090325 Clanton TO, Coupe KJ. Hamstring strains in  athletes: diagnosis and treatment. J Am Acad Orthop Surg. 1998;6:237-248. Connell DA, Schneider-Kolsky ME, Hoving JL, et  al. Longitudinal study comparing sonographic and MRI assessments of acute and healing hamstring injuries. AJR Am J Roentgenol. 2004;183:975-984. Cook C, Hegedus E. Diagnostic utility of  clinical tests for spinal dysfunction. Man Ther. 2011;16:21-25. http://dx.doi.org/10.1016/j. math.2010.07.004 Cox DR. The Analysis of Binary Data. London,  UK: Methuen; 1970. Croisier JL, Ganteaume S, Binet J, Genty M,  Ferret JM. Strength imbalances and prevention of hamstring injury in professional soccer players: a prospective study. Am J Sports Med. 2008;36:1469-1475. http://dx.doi. org/10.1177/0363546508316764 Deyo RA, Rainville J, Kent DL. What can the his tory and physical examination tell us about low back pain? JAMA. 1992;268:760-765. Donelson R, Aprill C, Medcalf R, Grant W. A pro spective study of centralization of lumbar and referred pain. A predictor of symptomatic discs and anular competence. Spine (Phila Pa 1976). 1997;22:1115-1122. Doust JA, Pietrzak E, Sanders S, Glasziou PP.  Identifying studies for systematic reviews of diagnostic tests was difficult due to the poor sensitivity and precision of methodologic lters and the lack of information in the abstract. J Clin Epidemiol. 2005;58:444-449. http://dx.doi. org/10.1016/j.jclinepi.2004.09.011 Drezner JA. Practical management: ham string muscle injuries. Clin J Sport Med. 2003;13:48-52. Fishman LM, Dombi GW, Michaelsen C, et al.  Piriformis syndrome: diagnosis, treatment, and outcomea 10-year study. Arch Phys Med Rehabil. 2002;83:295-301. Hegedus EJ, Cook C, Hasselblad V, Goode A,  McCrory DC. Physical examination tests for assessing a torn meniscus in the knee: a systematic review with meta-analysis. J Orthop Sports Phys Ther. 2007;37:541-550. http://dx.doi. org/10.2519/jospt.2007.2560

RESEARCH REPORT
23. H  egedus EJ, Goode A, Campbell S, et al. Physical examination tests of the shoulder: a systematic review with meta-analysis of individual tests. Br J Sports Med. 2008;42:80-92. http:// dx.doi.org/10.1136/bjsm.2007.038406 24.  Hoskins W, Pollard H. The management of hamstring injurypart 1: issues in diagnosis. Man Ther. 2005;10:96-107. http://dx.doi. org/10.1016/j.math.2005.03.006 25.  Hunter DG, Speed CA. The assessment and management of chronic hamstring/posterior thigh pain. Best Pract Res Clin Rheumatol. 2007;21:261-277. http://dx.doi.org/10.1016/j. berh.2006.12.002 26.  Jaeschke R, Guyatt GH, Sackett DL. Users guides to the medical literature. III. How to use an article about a diagnostic test. B. What are the results and will they help me in caring for my patients? The Evidence-Based Medicine Working Group. JAMA. 1994;271:703-707. 27.  Landis JR, Koch GG. An application of hierarchical kappa-type statistics in the assessment of majority agreement among multiple observers. Biometrics. 1977;33:363-374. 28.  Laslett M, berg B, Aprill CN, McDonald B. Centralization as a predictor of provocation discography results in chronic low back pain, and the inuence of disability and distress on diagnostic power. Spine J. 2005;5:370-380. http://dx.doi. org/10.1016/j.spinee.2004.11.007 29.  Laslett M, Young SB, Aprill CN, McDonald B. Diagnosing painful sacroiliac joints: a validity study of a McKenzie evaluation and sacroiliac provocation tests. Aust J Physiother. 2003;49:89-97. 30.  Lehance C, Binet J, Bury T, Croisier JL. Muscular strength, functional performances and injury risk in professional and junior elite soccer players. Scand J Med Sci Sports. 2009;19:243-251. http://dx.doi. org/10.1111/j.1600-0838.2008.00780.x 31.  Meeuwisse WH, Sellmer R, Hagel BE. Rates and risks of injury during intercollegiate basketball. Am J Sports Med. 2003;31:379-385. 32.  Mendiguchia J, Brughelli M. A return-to-sport algorithm for acute hamstring injuries. Phys Ther Sport. 2011;12:2-14. http://dx.doi.org/10.1016/j. ptsp.2010.07.003 33.  Moher D, Liberati A, Tetzlaff J, Altman DG. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. Int J Surg. 2010;8:336-341. http://dx.doi. org/10.1016/j.ijsu.2010.02.007 34.  Orchard J, Best TM, Verrall GM. Return to play following muscle strains. Clin J Sport Med. 2005;15:436-441. 35.  Orchard J, Seward H. Epidemiology of injuries in the Australian Football League, seasons 1997-

]
2000. Br J Sports Med. 2002;36:39-44. 36. R  utjes AW, Reitsma JB, Di Nisio M, Smidt N, van Rijn JC, Bossuyt PM. Evidence of bias and variation in diagnostic accuracy studies. CMAJ. 2006;174:469-476. http://dx.doi.org/10.1503/ cmaj.050090 37.  Schneider-Kolsky ME, Hoving JL, Warren P, Connell DA. A comparison between clinical assessment and magnetic resonance imaging of acute hamstring injuries. Am J Sports Med. 2006;34:1008-1015. http://dx.doi. org/10.1177/0363546505283835 38.  Slavotinek JP. Muscle injury: the role of imaging in prognostic assignment and monitoring of muscle repair. Semin Musculoskelet Radiol. 2010;14:194-200. http://dx.doi. org/10.1055/s-0030-1253160 39.  Stankovic R, Johnell O, Maly P, Willner S. Use of lumbar extension, slump test, physical and neurological examination in the evaluation of patients with suspected herniated nucleus pulposus. A prospective clinical study. Man Ther. 1999;4:25-32. 40.  Swartz MK. The PRISMA statement: a guideline for systematic reviews and meta-analyses. J Pediatr Health Care. 2011;25:1-2. http://dx.doi. org/10.1016/j.pedhc.2010.09.006 41.  Verrall GM, Slavotinek JP, Barnes PG, Fon GT. Diagnostic and prognostic value of clinical ndings in 83 athletes with posterior thigh injury: comparison of clinical ndings with magnetic resonance imaging documentation of hamstring muscle strain. Am J Sports Med. 2003;31:969-973. 42.  Vroomen PC, de Krom MC, Wilmink JT, Kester AD, Knottnerus JA. Diagnostic value of history and physical examination in patients suspected of lumbosacral nerve root compression. J Neurol Neurosurg Psychiatry. 2002;72:630-634. 43.  Whiting P, Rutjes AW, Reitsma JB, Bossuyt PM, Kleijnen J. The development of QUADAS: a tool for the quality assessment of studies of diagnostic accuracy included in systematic reviews. BMC Med Res Methodol. 2003;3:25. http:// dx.doi.org/10.1186/1471-2288-3-25 44.  Woods C, Hawkins RD, Maltby S, Hulse M, Thomas A, Hodson A. The Football Association Medical Research Programme: an audit of injuries in professional footballanalysis of hamstring injuries. Br J Sports Med. 2004;38:36-41. 45.  Zeren B, Oztekin HH. A new self-diagnostic test for biceps femoris muscle strains. Clin J Sport Med. 2006;16:166-169.

12.

13.

14.

15. 16.

17.

18.

19.

20.

21.

22.

MORE INFORMATION
WWW.JOSPT.ORG

228 | april 2013 | volume 43 | number 4 | journal of orthopaedic & sports physical therapy

APPENDIX

DESCRIPTION OF THE ORTHOPAEDIC SPECIAL TESTS

Test Description Puranen-Orava test  This test entails actively stretching the hamstring muscles in the standing position with the hip exed at about 90, the knee fully extended, and the foot on a solid support surface. Positive Finding Illustration Exacerbation of the patients symptoms.

Puranen-Orava test. Bent-knee stretch test T  he patient is supine. The hip and knee of the symptomatic limb are maximally exed, and the clinician slowly straightens the knee while keeping the hip exed. Exacerbation of the patients symptoms.

Modied bent-knee  The patient lies in the supine position with the stretch test lower extremities fully extended. The clinician grasps the symptomatic limb behind the heel with one hand and at the knee with the other. The clinician maximally exes the hip and knee, and then rapidly straightens the knee.

Exacerbation of the patients symptoms.

Modied bent-knee stretch test, start position and nish position.

journal of orthopaedic & sports physical therapy | volume 43 | number 4 | april 2013 | 229

RESEARCH REPORT
APPENDIX

Test Description Taking-off-the-shoe  In standing, the patient is asked to take off the test shoe on the affected side with the help of his/ her other shoe. While performing this maneuver, the affected leg hindfoot must press the longitudinal arch of the noninvolved foot. The affected leg during the maneuver is in approximately 90 of external rotation at the hip and 20 to 25 of exion at the knee.

Positive Finding The feeling of a sharp pain over the injured biceps femoris.

Illustration

Taking-off-the-shoe test: posterior view. Active range-ofmotion test H  ip extension: in prone, the patient is asked to actively extend the hip with an extended knee.  Knee exion: in prone, the patient is asked to ex the knee as far as he/she can. Reproduction of patients concordant pain with either test.

Passive range-ofmotion test

P  assive hip exion: the patient is supine, with Reproduction of patients the pelvis stabilized by grasping the iliac crest. concordant pain with As the hip is exed, the knee is allowed to ex either test. from the tension placed on the hamstrings and gravity. With pressure applied proximal to the knee joint, the normal end feel for hip exion is soft owing to the approximation of the quadriceps with the abdomen.  Passive knee extension: the patient is supine with the hip exed to 90, with the knee exed in a relaxed position. The lower leg (below the knee) is passively extended to a rm muscle tension end point.

Resisted range-ofmotion test

H  ip extension with an extended knee: the patient is prone, with the knee extended and the pelvis stabilized with pressure on the iliac crest. An isometric break test is performed at end-range hip extension, with resistance applied to the popliteal fossa.

Reproduction of patients concordant pain with either test.

230 | april 2013 | volume 43 | number 4 | journal of orthopaedic & sports physical therapy

APPENDIX

Test Resisted rangeof-motion test (continued)

Description Positive Finding  Knee exion: the patient is prone with the knee extended. An isometric break test is performed with the knee exed to 10, 45, and 90 unless contraindicated. Clinician provides resistance over the Achilles tendon. P  assive straight leg raise: with the patients Reproduction of patients lower extremity completely relaxed, the cliniconcordant pain/stiffness cian lifts the lower extremity off the plinth with during any 1 of the 3 tests. the knee fully extended. The degree of hip exion is measured with a bubble goniometer.  Active knee extension: the patients thigh is vertical with the posterior distal aspect of the thigh, resting lightly against a frame to keep the thigh perpendicular to the plinth. With the ankle relaxed in plantar exion, the patient is asked to actively extend the knee while maintaining light contact with the horizontal part of the frame. A temporary myoclonus of alternating contraction and relaxation of the quadriceps and hamstring muscle groups tends to occur at the maximum angle of active knee extension. At this point, the patient is instructed not to force the leg past the point of initial mild resistance. The patient is then asked to slightly ex the knee until myoclonus ceases. At the rst point at which the shaking ceases, the angle between the vertical and the tibia is recorded using an inclinometer.  Manual muscle testing: manual muscle testing in the prone position is performed by asking the patient to lift his/her heel by bending his/ her knee to the point at which the toe is off the couch. The patient is asked to hold that position while a gentle, steadily increasing resistance is applied to the heel (about 15 of knee exion).

Illustration

Composite clinical assessment

DOWNLOAD PowerPoint Slides of JOSPT Figures & Tables

JOSPT offers PowerPoint slides of gures and tables to accompany selected articles on the Journals website (www.jospt.org). These slides can be downloaded and saved and include the article title, authors, and full citation. With each article where this feature is available, click View Slides and then right click on the link and select Save Target As.

journal of orthopaedic & sports physical therapy | volume 43 | number 4 | april 2013 | 231