Clinical question.

Is radiologic examination necessary for a 9-year-old girl with a knee injury?

The purpose of "Evidence in Practice" is to illustrate the literature search process to obtain evidence that can guide clinical decision making. This article is not a case report. The examination, evaluation, and intervention sections are purposely abbreviated.

The father of a 9-year-old girl contacted our clinic to schedule an examination. He reported that his daughter injured her left knee during gymnastics practice the previous evening. He stated that, while attempting to perform a forward flip maneuver, she heard a "pop" in her left knee as she initiated the jump. According to witnesses, the girl landed awkwardly on her heels, with her knees bent, and she subsequently fell backward into a sitting position. The girl reported immediate anterior left knee pain following the maneuver. Because of her pain, she had been unable to walk since the time of the injury. Her father spoke to a physical therapist (Smith) at our clinic to relay this information and to express urgency in the case. The patient had not seen another medical professional for the injury. The father was a former client at our clinic and, therefore, sought our opinion about his daughter's injury.

The girl had not yet undergone an examination or diagnostic imaging; therefore, we would be serving as a point of entry to the health care system. We planned to assess the extent of the injury in order to determine whether referral to another medical specialist was warranted. The patient's history of trauma during a high-velocity maneuver (ie, forward flip) and her inability to walk the following day indicated the possibility of a fracture, a growth plate injury, or some other serious injury (eg, ligamentous or meniscal injury). The prevalence of ligamentous or meniscal injury in knee trauma is much larger than the prevalence of fractures; less than 7% of traumatic injuries to the knee result in a fracture," (1,2) and the prevalence of knee fractures in a primary care clinic is only 1.2% of all patients with knee injuries. (3) The frequency of patients presenting with knee fractures in an outpatient physical therapy clinic is probably even lower; however, we decided to investigate the literature before she arrived for her examination to identify evidence that could aid our clinical decision making and help us decide whether radiologic examination was warranted in this case.

Although the incidence of knee fracture is low, we needed to ascertain whether the patient had a fracture prior to implementing a physical therapy plan of care. Therefore, we were interested in determining the factors or conditions that would necessitate taking radiographs of the knee to rule out the possibility of a fracture in the patient.

* Database used for search: Because our clinical question dealt specifically with the necessity of radiologic examination for a specific patient scenario (ie, children with knee trauma), we decided to search MEDLINE. Although other databases may contain a greater number of articles and journals related Io physical therapy, MEDLINE, the National Library of Medicine's computerized database, contains more than 11 million citations and more than 4,600 biomedical journals, the majority of which are related to medicine. Our clinical question focused on determining the need for diagnostic imaging, and we felt that searching a database that primarily focused on medicine was most appropriate. We accessed MEDLINE through PubMed (www.ncbi.nlm.nih.gov/PubMed/). The search was performed on August 23, 2004.

* Initial keywords: knee injury, diagnostic imaging, knee fractures, sensitivity

We typed knee injury AND diagnostic imaging into the query box. We used these terms because we believed that this search would result in articles that would help identify the signs and symptoms that indicate a need for radiographs. A search using "AND" will identify citations using both phrases, whereas "OR" will identify citations with either of the phrases in the search. The search revealed 1,317 citations. In order to narrow this number of results, we performed a second search using only the keyword phrase knee fractures because we specifically wanted to determine signs and symptoms that might be related to the presence of knee fractures. This particular search produced 4,182 citations.

We now wanted to combine our searches to reveal citations that included all previously used keywords. On the main PubMed search screen, we clicked on Preview/Index, located in the Features bar underneath the query box. (For more information on the Preview/Index page, see the sidebar on p. 1094.) The Preview/Index page then displayed our most recent searches under the heading "Most Recent Queries" (Fig. 1). We combined the 2 searches by typing #1 AND #2, which combined search lines 1 and 2 in our search history, into the query box at the top of the page. We then clicked on the Preview button. This search identified 204 citations.

[FIGURE 1 OMITTED]

Because our clinical question focused on identifying factors associated with a knee fracture, we wanted to determine the relevant diagnostic properties associated with findings that may indicate a knee fracture. We decided to use "sensitivity" as a search term. A test with high sensitivity is useful as a screening procedure because high sensitivity indicates that the test has a low rate of false negative results (ie, that a negative test would be useful for excluding the condition): If a finding, or duster of findings, was found to have high sensitivity for detecting a knee fracture and these findings were absent in our patient (ie, a negative test result), we might be able to rule out the presence of a fracture. (4) If the findings were present in our patient, however, a referral for radiographs would be indicated.

We added the keyword "sensitivity" to the previous search by typing sensitivity into the query box located under the heading "Add Term(s) to Query or View Index." We lot the Fields dropdown menu to the left of the box on its default setting of All Fields. (This dropdown menu allows users to limit the search of a keyword to a specific field in MEDLINE records, such as "Author" "Journal," "MESH [Medical Subject Headings] Terms," and "Title.") We then clicked on the AND button, which added the term to our search string. Our final search string was #1 AND #2 AND sensitivity. We then clicked on the Preview button next to the main query box at the top of the Preview/Index screen. This resulted in 23 citations (Fig. 2). We clicked on the Go button at the top of the page, and PubMed displayed a list of these citations. We read all the titles in the list of citations to decide which articles might provide evidence for the diagnostic accuracy of signs and symptoms associated with a knee fracture.

* Selection of articles for review: Several of the 23 citations referred to the Ottawa Knee Rule (OKR). The OKR is a clinical prediction rule (CPR) that has been developed to determine the need for radiographs in patients with traumatic knee injuries or, more specifically, to reduce the costs and morbidity from unnecessary radiographs. Laupacis and colleagues (5) have defined clinical prediction rules (sometimes known as clinical decision rules) as the identification of predictor variables established from the patient history and physical examination that help clinicians determine a diagnosis, prognosis, or appropriate treatment strategies. Clinical prediction rules combine the diagnostic properties of sensitivity and specificity and are used to increase the accuracy of clinical decision making in situations where an incorrect decision can have serious consequences. (5,6)

The development of a CPR is a 3-step process: deriving predictor variables, validating the CPR, and analyzing the impact of the CPR on clinician behavior. The first step is identifying the possible predictor variables that can increase the likelihood of identifying the outcome of interest. (6) The OKR was originally derived by Stiell and colleagues (7) and identified 5 variables that predicted the ordering of radiographs with patients with acute knee injuries. These 5 predictor variables were:

(1) the patient is 55 years of age or older,

(2)the patient has isolated tenderness of the patella (only the patella),

(3) the patient has tenderness of the head of the fibula,

(4)the patient is unable to flex the knee more than 90 degrees, and

(5) the patient is unable to bear weight for 4 steps.

Once a CPR has been derived, it must be validated before it can be incorporated into clinical practice, because the predictor variables identified during the derivation study may have occurred simply by chance: Validation of the rule should occur in different settings, with heterogeneous patient populations, and should be carried out by different examiners to improve the rule's generalizability. (6) The OKR has been validated in a number of studies. (2,8-10) Once a study has been validated for widespread clinical use, it must be subjected to the scrutiny of impact analysis to determine whether the rule changes clinician behaviors or results in improved efficiency and patient care. (6)

As we read through our list of 23 citations, we selected 2 articles that reported evidence on the diagnostic accuracy of the OKR. We were aware of the potential usefulness of CPRs in determining a diagnosis and felt the OKR might be most beneficial in determining appropriate evidence to guide our clinical decision making. Both articles were obtained from a local medical library.

The first article of interest and the one we expected to be of greatest value was Bachmann et al (11) (Fig. 2, citation 1) because it was directly related to our clinical question and because it was a systematic review. Systematic reviews rank high on the hierarchy of evidence proposed by Sackett and colleagues. (12) The creators of a systematic review attempt to limit the inherent selection bias often associated with review articles by following a strict methodology to enhance the review's validity. This is accomplished by performing an exhaustive search of the literature using explicit inclusion and exclusion criteria and by analyzing and scoring the identified articles to determine their methodological quality and strength of the evidence for the benefit and harm of an intervention or, in this instance, the diagnostic accuracy of a test. (13) The citation and abstract for the systematic review by Bachmann et al (11) are reproduced below.

Bachmann LM, Haberzeth S, Steurer J, ter Riet G. The accuracy of the Ottawa knee rule to rule out knee fractures: a systematic review. Ann Intern Med. 2004 Jan 20;140(2):121-4.

BACKGROUND: The Ottawa knee rule is a clinical decision aid that helps rule out fractures and avoid unnecessary radiography. PURPOSE: To summarize evidence about the accuracy of the Ottawa knee rule. DATA SOURCES: Relevant English- and non-English-language articles were identified from PreMEDLINE and MEDLINE (1966-2003), EMBASE (1980-2003), CINAHL (1982-2003), BIOSIS (1990-2003), the Cochrane Library (2002, Issue 3), the Science Citation Index database, reference lists of included studies, and experts. STUDY SELECTION: Articles were included if they reported enough information to determine the sensitivity and specificity of the Ottawa knee rule for detecting fractures confirmed either radiologically or in combination with follow-up. DATA EXTRACTION: Two reviewers independently extracted data on study samples, the ways that the Ottawa knee rule was used, and methodologic characteristics of studies. DATA SYNTHESIS: Of 11 identified studies, 6 involving 4249 adult patients were considered appropriate for pooled analysis. The pooled negative likelihood ratio was 0.05 (95% CI, 0.02 to 0.23), the pooled sensitivity was 98.5% (CI, 93.2% to 100%), and the pooled specificity was 48.6% (CI, 43.4% to 51.0%). CONCLUSION: A negative result on an Ottawa knee rule test accurately excluded knee fractures after acute knee injury. However, because the rule is calibrated toward 100% sensitivity and actual fracture prevalences are usually low, large-scale, multicentered studies are still needed to establish the cost-effectiveness of routinely implementing the rule.

[[c] 2004 American College of Physicians. Abstract reprinted with permission of the American College of Physicians.]

This article was a systematic review analyzing the accuracy of the OKR in determining the need to order radiographs fop lowing an acute knee injury. The systematic review initially identified 104 articles; however; only 11 met the review's strict inclusion criteria. The articles included in the review were cohort studies comparing the OKR with the reference standard of plain film radiography. The authors ranked a study as high quality (level 1) if patient enrollment was consecutive, if all participants received the reference standard, and if radiographs were assessed by experts blinded to the OKR results. (A level 2 ranking was given if 2 of 3 criteria were met, a level 3 ranking was given if only 1 of 3 criteria was met, and a very low ranking [level 4] was given if none of the criteria were met.) Five studies were excluded from the pooled data because of low methodological quality. The pooled results of the remaining 6 studies revealed a sensitivity of 98.5% (95% confidence interval [CI]=93.2%-100%), a specificity of 48.6% (95% CI=43.4%-51%), and a negative likelihood ratio of 0.05 (95% CI=0.02-0.23). In the remaining 6 studies reviewed, the probability of a patient presenting with a patellar fracture after a negative result on the OKR was 0.37% (95% CI=0.15%-1.48%). As we read through the article, however; we realized that studies investigating the accuracy of the OKR in children were excluded from the review. The results of the systematic review, therefore, could not be generalized to our adolescent patient.

The second article of interest from the initial search was Cohen et al (14) (Fig. 2, citation 15) because it dealt specifically with clinical criteria for using radiographs with children with acute knee injuries.

Cohen DM, Jasser JW, Kean JR, Smith GA. Clinical criteria for using radiography for children with acute knee injuries. Pediatr Emerg Care. 1998 Jun;14(3):185-7.

OBJECTIVE: To evaluate clinical criteria for selective radiography for knee injuries in children. DESIGN: Retrospective chart review. SETTING: Emergency department (ED) of a children's hospital. PARTICIPANTS: All patients evaluated by radiography for an isolated, acute knee injury during 12 months. Patients were excluded for injuries: >1 week; isolated to superficial lacerations/abrasions; with prior knee surgery; being reassessed. RESULTS: Two hundred fifty-four patients (60% male; 12.7 years median age) were included. Twelve patients (4.7%) sustained a fracture. Evaluated criteria were point tenderness, inability to bear weight in the ED, and inability to flex the knee to 90 degrees. Point tenderness was not statistically associated with fracture, P = 0.7. Inability to bear weight in the ED (37% fracture rate, P = 0.001) and inability to flex to 90 degrees (52% fracture rate, P < 0.001) were associated with the presence of fracture. [table in text] Applying a rule combining nobearwt and noflex90 would decrease the number of x-rays by 73%, with no missed fractures. CONCLUSIONS: Point tenderness was not a good predictor of knee fracture in children. Using the clinical criteria to select patients requiring knee radiography may greatly reduce the number of unnecessary x-rays.

[[c] 1998 Lippincott Williams & Wilkins Inc. Abstract reprinted with permission of Lippincott Williams & Wilkins Inc.]

This study revealed 2 variables that were associated with the presence of fractures in children--the inability to bear weight and the inability to flex the knee to 90 degrees. The study design, however, was a retrospective chart review, which presents a number of inherent limitations, including missing data points and the lack of standardized methods of patient examination. Considering this, we were not confident that the study by Cohen et al (14) provided sufficient evidence to guide our decision making in this particular case, so we decided to expand our search.

* Second keywords: knee radiographs, rule*

The second search was specifically focused on identifying CPRs related to the ordering of radiographs in children following acute knee trauma. Once again we performed a MEDLINE search through the PubMed database. We typed the following search string knee radiographs AND rule* into the query box of PubMed's main search screen. We wanted our search to include all CPRs related to the knee, and the word "rule" was selected as a keyword because we expected that this term would identify citations containing the phrases "clinical prediction rules," "decision rules," and the "Ottawa Knee Rule." We truncated the word "rule" using an asterisk in an attempt to include singular and plural versions--"rule" and "roles." (15) In addition, Ingul and Rogers (15) have demonstrated that combining search terms with the keyword "rule*," when searching specifically for CPRs, results in a MEDLINE, search with a sensitivity above 90%. The sensitivity of a search strategy is determined by comparing the results of the search to a reference standard, which most often entails hand searching numerous journals referenced within a particular database. The sensitivity is the proportion of retrieved citations to the number of relevant citations. (16) Because our primary focus was to ensure that we retrieved all articles that investigated the OKR, we did not truncate the keyword "radiographs."

The results of our search using the search string knee radiographs AND rule* produced 38 citations. We further refined the search by clicking on the Limits button and selecting the All Child: 0-18 years option in the Ages dropdown menu. This further narrowed the number of citations retrieved to 14 (Fig. 3). Of the 14 citations identified, 2 (Fig. 3, citations 1 and 5) appeared relevant to our clinical question because they both contained "Ottawa Knee Rule" and "children" in their titles. Based on the abstracts of the 2 citations, we decided to retrieve both articles from our local medical library. The first article we read was the study by Khine et al (17) (Fig. 3, citation 5).

Khine H, Dorfman DH, Avner JR. Applicability of Ottawa knee rule for knee injury in children. Pediatr Emerg Care. 2001 Dec;17(6):401-4.

OBJECTIVE: Previous studies have shown that the application of the Ottawa knee rule (OKR) reduces the need for radiographs in adults with acute knee injuries. Our objectives were to describe the epidemiology and incidence of knee injuries in children with acute knee trauma and to validate the OKR in a pediatric population. DESIGN: A prospective, consecutive study. SETTINGS: Two urban pediatric emergency departments. METHODS: All children 18 years of age and under who presented with acute traumatic knee injury of less than 1 week's duration, excluding patients with a normal knee examination, superficial skin injuries, prior history of knee injury, underlying bone disease, serious injuries involving two or more organ systems, or altered mental status were enrolled. Physicians assessed each patient for 22 standardized clinical findings prior to radiography. The OKR was applied to each patient by the investigating physician. RESULTS: All 234 patients eligible for the study had radiographs of the affected knee. The median age was 13 years with a range of 2 to 18 years. Using the OKR criteria for obtaining knee radiographs, 12 of 13 patients with fractures were identified (sensitivity 92%; 95% CI= 64-99). The missed case was an 8-year-old male who had sustained a nondisplaced fracture of the proximal tibia after a fall. If the OKR were applied to the pediatric population, it would have reduced the need for radiography in 46% of children. CONCLUSIONS: In the pediatric population studied, the OKR did not identify all patients with knee fractures. Future studies may consider modifying the OKR to accommodate the differences between pediatric and adult patients to improve the sensitivity of the rule while maintaining its specificity, before it can be applied routinely in clinical practice.

[[c] 2001 Lippincott Williams & Wilkins Inc. Abstract reprinted with permission of Lippincott Williams & Wilkins Inc.]

Khine and colleagues (17) investigated the accuracy of the OKR in 234 patients under 18 years of age (mean age=13 years) who were seen in 1 of 2 emergency departments following a traumatic knee injury. The prevalence of fractures in this sample was 6%. According to the results of the study, the OKR exhibited a sensitivity of 92% (95% CI=64%-99%) and a specificity of 49% (95% CI=42-56). The OKR failed to identify one child (8%)with a knee fracture. The small sample size and limited prevalence of patients with knee fractures resulted in wide CIs associated with the specificity and sensitivity. Considering these factors, we were not confident that this study provided adequate evidence to guide our clinical decision making. The authors acknowledged the limitations of their study and suggested that further validation is necessary before the OKR can routinely be used in a pediatric population.

Bulloch et al (18) (Fig. 3, citation 1) was the second OKR validation study identified in our latest database search.

Bulloch B, Neto G, Plint A, Lim R, Lidman P, Reed M, Nijssen-Jordan C, Tenenbein M, Klassen TP; Pediatric Emergency Researchers of Canada. Validation of the Ottawa Knee Rule in children: a multicenter study. Ann Emerg Med. 2003 Jul;42(1):48-55.

STUDY OBJECTIVE: The main objective of this study was to determine the sensitivity and specificity of the Ottawa Knee Rules when they were applied to children. The secondary objective was to determine post hoc whether use of the rules would reduce the number of knee radiographs ordered. METHODS: This prospective, multicenter validation study included children aged 2 to 16 years who presented to the emergency department with a knee injury sustained in the preceding 7 days. Children were assessed for the variables comprising the Ottawa Knee Rules, and physicians ordered radiographs at their discretion. A positive outcome was defined as any fracture. A negative outcome was defined as children who did not have a fracture on radiograph or, if no radiograph was obtained, were asymptomatic after 14 days. RESULTS: A total of 750 children were enrolled. The mean age was 11.8+/-3.1 years, and 443 (58.7%) were male patients. Seventy children had fractures. Radiography was performed for 670 children, whereas 80 children had only a structured telephone interview. The Ottawa Knee Rules were 100% sensitive (95% confidence interval [CI] 94.9% to 100%), with a specificity of 42.8% (95% CI 39.1% to 46.5%). Only 460 children would have required a radiograph if radiographs had been performed according to the Ottawa Knee Rules, which would have resulted in an absolute reduction of 209 (31.2%) radiographs. CONCLUSION: The Ottawa Knee Rules are valid in children and have the potential to decrease the use of radiography in children with knee injuries.

[[c] 2003 American College of Emergency Physicians. Abstract reprinted with the permission of the American College of Emergency Physicians.]

This validation study included a greater sample size (750 children who were less than 18 years of age) and larger prevalence (9%) of fractures than the study by Khine et al. (17) The average age of patients was 11.8 years, and data were collected from 5 pediatric emergency departments. The sensitivity and specificity of the OKR in this group of patients were 100% (95% CI=94.9%-100%) and 42.8% (95% CI= 39.1%-46.5%) respectively. These results gave us increased confidence in using the OKR to guide our decision making with our patient, considering the sensitivity of 100% (no missed fractures) and the narrower confidence intervals than the study by Khine et al. (17)

Bulloch et al (18) further classified their findings on the accuracy of the OKR into age groups, which allowed us to generalize the findings to our patient more adequately. In the group of patients 6 to 12 years of age (n=375), the sensitivity was determined to be 100% (95% CI-88.8%-100%), and specificity was reported as 47.1% (95% CI=41.8%-52.4%). In addition, the study reported the interrater reliability of the application of the OKR in a pediatric population. In the 6-12 age group, the kappa statistic was .81 (95% CI=.55-1.0).

* Clinical decision: Based on the validation of the OKR in a broad spectrum of pediatric patients from multiple emergency departments, the study by Bulloch et al (18) represents level 2 evidence according to the hierarchy of evidence for" clinical decision rules described by McGinn and colleagues. (6) According to this hierarchy, level 2 CPRs "can be used in various settings with confidence in their accuracy" because these CPRs have "demonstrated accuracy in either 1 large prospective study including a broad spectrum of patients and clinicians or validated in several smaller settings that differ from one another." (6(p81)) Considering the high level of evidence and the ability to relate the findings to a narrower age group similar to our patient, we were confident that we could use the OKR to guide our clinical decision making about referring our patient for radiographic examination. If a highly sensitive test (100% in this case) is negative, then it is valuable for ruling out a disorder. (12) Given the 100% sensitivity of the OKR in children, (18) we could be confident that if the patient was negative for all of the 5 criteria identified by the OKR, then she would not need radiographic examination.

* Examination: The girl, carried by her father, came to our clinic the day after he first contacted us. No visible swelling was noted surrounding the tibiolfemoral joint; however, she had a small raised (1 x 1/2 cm) area on the distal third of the patella, which was extremely tender to palpation. She held her knee in a position of 20 degrees of flexion and reported an increase in pain with active flexion and extension. Active contraction of quadriceps temoris muscle also elicited pain at the anterior patella. When she relaxed, the physical therapist (Smith) was able to fully extend and flex the knee to greater than 125 degrees with little or no pain. She had a negative score on the Lachman test, which has been demonstrated to accurately rule out anterior cruciate ligament tears when the test is negative. (19) Our patient had no tenderness to palpation at the medial collateral ligament or the lateral collateral ligament, with varus and valgus stress testing deferred because of patient anxiety with movement in the coronal plane. She was reluctant to stand on her left leg and was unable to take any steps, but she was able to fully bear weight without pain when the therapist manually stabilized her knee.

* Applying the CPR: The examination revealed that she was positive for 2 of the 5 items in the OKR. She could not walk or shift weight to the involved leg, and she had tenderness only on the patella. The OKR states that only 1 of the 5 criteria needs to be positive to suggest ordering radiographs. Given the current level of evidence and the identification of 2 positive variables, and recognizing the possible consequences associated with not identifying a fracture, we decided that a referral for radiographs was the most appropriate clinical decision in this case.

* Results: Following the consultation, the girl's parents drove her to the emergency department at a local hospital. The attending physician concurred that radiographs should be taken. The radiograph revealed a horizontal fracture across the inferior aspect of her left patella (Fig. 4). She was placed on crutches with a straight leg splint and given an appointment for the following Monday with an orthopedist. The physician placed her in a long leg cast for 2 weeks, and then recasted the leg in a walking cast that she wore for 4 more weeks.

[FIGURE 4 OMITTED]

In summation, in this particular case, the need for radiographs might appear obvious; however, we could not identify the presence of a fracture based solely on her clinical presentation. The OKR was initially developed to reduce the number of unnecessary radiographs. (7) Based on the evidence we obtained during the search, if our patient was negative on the OKR, we would have been confident (as a result of the 100% sensitivity (18)) that she did not have a fracture, and we would not have referred her for radiographs. This case provides an example of how clinicians can utilize search strategies in more complex clinical scenarios where a diagnosis is less certain.

Preview/Index Page

PubMed's Preview/Index page has two functions. First, the Preview function allows users to build and refine search strategies by adding keywords one at a time using a query box at the bottom of the page under the heading "Add Term(s) to Query at View index. Once the user clicks on the Preview button. PubMed displays the number of citations retrieved and a search line number rather than a list of citations, allowing the user to combine separate searches (Fig. 1). Second, the Index function allows users to Build a search in the same entry box using Medical Subject Heading (MeSH) Terms. To do this, the user selects MeSH Terms in the Fields dropdown menu next to the lower query box, types in the keyword and clicks on the Index button. PubMed then displays an index of MeSH terms which the user can scroll through to select the proper MeSH term. The user can then preview the results or combine the term with other terms using the AND, OR, or NOT buttons.

Figure 2. Citations retrieved by the search of PubMed using the search
string #1 AND #2 AND sensitivity. #1=a search using the search string
knee surgery AND diagnostic imaging, #2=a search using the keyword
"knee fractures." Citations that were selected for further examination
are highlighted in red.

1. Bachmann LM, Haberzeth S, Steurer J, ter Riet G.
The accuracy of the Ottawa knee rule to rule out knee
fractures: a systematic review. Ann Intern Med.
2004; 140(2):121-4. Review.

2. Remplik P, Stabler A, Merl T, et al. Diagnosis of acute
fractures of the extremities: comparison of low-field
MRI and conventional radiography. Eur Radiol.
2004;14(4):625-30.

3. Jackson JL, O'Malley PG, Kroenke K. Evaluation of
acute knee pain in primary care. Ann Intern Med.
2003;139(7):575-88. Review.

4. Romaneehsen B, Oberholzer K, Muller LP, Kreitner
KF. Rapid musculoskeletal magnetic resonance imaging
using integrated parallel acquisition techniques
(IPAT)--initial experiences. Rofo Fortschr Geb Rontgenstr
Neuen Bildgeb Verfahr. 2003;175(9):1193-7.

5. Lavy D, Morin O, Venet G, et al. Pseudallescheria
boydii knee arthritis in a young immunocompetent
adult two years after a compound patellar fracture.
Joint Bone Spine. 2001;68(6):517-20.

6. Kotsianos D, Rock C, Wirth S, et al. [Detection of tibial
condylar fractures using 3D imaging with a mobile
image amplifier (Siemens ISO-C-3D): Comparison with
plain films and spiral CT] Rofo Fortschr Geb Rontgenstr
Neuen Bildgeb Verfahr. 2002;174(1):82-7. German.

7. Emparanza JI, Aginaga JR; Estudio Multicentro en
Urgencias de Osakidetza: Reglas de Ottawa
(EMUORO) Group. Validation of the Ottawa Knee
Rules. Ann Emerg Med. 2001;38(4):364-8.

8. Kotsianos D, Rock C, Euler E, et al. [3-D imaging with
a mobile surgical image enhancement equipment (ISO-C-3D).
Initial examples of fracture diagnosis of peripheral
joints in comparison with spiral CT and conventional radiography]
Unfallchirurg. 2001;104(9):834-8. German.

9. Prokop A, Fischbach R, Burger C, et al. [Diagnosis of
intra-articular fracture of the head of the tibia. A prospective
comparative study] Unfallchirurg. 2001;104(2):131-7.
German.

10. Hackl W, Riedl J, Reichkendler M, et al. [Preoperative
computerized tomography diagnosis of fractures of the tibial
plateau] Unfallchirurg. 2001;104(6):519-23. German.

11. Pavlov H, Dalinka MK, Alazraki N, et al. Acute
trauma to the knee. American College of Radiology.
ACR Appropriateness Criteria. Radiology. 2000;215
Suppl:365-73. No abstract available.

12. Wicky S, Blaser PF, Blanc CH, et al. Comparison
between standard radiography and spiral CT with 3D
reconstruction in the evaluation, classification and
management of tibial plateau fractures. Eur Radial.
2000;10(8):1227-32.

13. Cabitza P, Tamim H. Occult fractures of tibial
plateau detected employing magnetic resonance imaging.
Arch Orthop Trauma Surg. 2000;120(5-6):355-7.

14. Neitzschman HR, McCarthy K, Eelkema J. Radiology
case of the month. Acute trauma to the knee.
Schatzker type 5 tibial plateau fracture. J La State Med
Soc. 1998;150(7):297-8. No abstract available.

15. Cohen DM, Jasser JW, Kean JR, Smith GA. Clinical
criteria for using radiography for children with acute knee
inquiries. Pediatr Emerg Care. 1998;14(3):185-7.

16. Tomburrini O, Bianchi D, Capparelli G, et el.
["Occult" posttraumatic lesions of the knee: con magnetic
resonance substitute for diagnostic arthroscopy?]
Radial Med (Torino). 1997;94(5):433-9. Italian.

17. Stiell IG, Wells GA, Haag RH, et al. Implementation
of the Ottawa Knee Rule for the use of radiography in
acute knee injuries. JAMA. 1997;278(23):2075-9.

18. Jurgensen I, Bachmann G, Leiers S, Cassens J.
[Traumatic hemarthrosis of the knee joint--clinical
significance of nuclear magnetic resonance tomography]
Unfallchirurg. 1996;99(9):665-70. German.

19. Bellelli A, Nardis P. [Occult or unknown traumatic
osteochondral lesions of the knee. Assessment of 19 cases
studied with conventional radiology and magnetic resonance]
Radial Med (Torino). 1996;91(6):700-4. Italian.

20. Stiell IG, Greenberg GH, Wells GA, et al. Prospective
validation of a decision rule for the use of radiography
in acute knee injuries. JAMA. 1996;275(8):611-5.

21. Giammarile F, Masciocchi C, Barile A, et al. Three-phase
radionuclide bone imaging and magnetic resonance
imaging detection of occult knee fractures in
athletes. Eur J Nucl Med. 1994;21(6):493-6.

22. Murray IP, Dixon J, Kohan L. SPECT for acute knee
pain. Clin Nucl Med. 1990;15(11):828-40.

23. Mink JH, Deutsch AL. Magnetic resonance imaging
of the knee. Clin Orthop. 1989;(244):29-47.

Figure 3. Citations retrieved by the search of PubMed using the
keywords "knee radiograph" and "rule*" with the limit "All Child:
0-18 years" from the Ages dropdown menu. Citations that were
selected for further examination are highlighted in red.

1. Bulloch B, Neto G, Plint A, Lim R, Lidman P, Reed
M, Nijssen-Jordan C, Tenenbein M, Klassen TP; Pediatric
Emergency Researchers of Canada. Validation of
the Ottawa Knee Rule in children: a multicenter study.
Ann Emerg Med. 2003 Jul;42(1):48-55.

2. Matteucci MJ, Roos JA. Ottawa Knee Rule: a comparison
of physician and triage-nurse utilization of a
decision rule for knee injury radiography. J Emerg
Med. 2003 Feb;24(2):147-50.

3. Thacher TD, Fischer PR, Pettifor JM. The usefulness
of clinical features to identify active rickets. Ann Trop
Paediatr. 2002 Sep;22(3):229-37.

4. Meyer S, Reinhard H, Graf N, Puschel W, Ziegler K,
Schneider G. [The importance of conventional radiographs
in the diagnosis of osteosarcoma] Klin Pediatr.
2002 Mar-Apr;214(2):58-61, German.

5. Khine H, Dorfman DH, Avner JR. Applicability of
Ottawa knee rule for knee injury in children. Pediatr
Emerg Care. 2001 Dec;17(6):401-4.

6. Pecina M, Bajok I, Pecina HI. Tuberculum intercondylare
tibiae tedium as a predictive factor for
anterior cruciate ligament injury. Am J Sports Med.
2001 Nov-Dec;29(6):709-11.

7. Szucs PA, Richman PB, Mandell M. Triage nurse
application of the Ottawa knee rule. Acad Emerg
Med. 2001 Feb;8(2):112-6.

8. Seaberg DC, Yealy DM, Lukens T, Auble T, Mathias
S. Multicenter comparison of two clinical decision
rules for the use of radiography in acute, high-risk
knee injuries. Ann Emerg Med 1998 Jul;32(1):8-13.

9. Richman PB, McCuskey CF, Nashed A, Fuchs S,
Petrik R, Imperato M, Hollander JE. Performance of
two clinical decision rules for knee radiography. J
Emerg Med. 1997 Jul-Aug;15(4):459-63.

10. Stiell IG, Wells GA, McDowell I, Greenberg GH,
McKnight RD, Cwinn AA, Quinn JV, Yeats A. Use of
radiography in acute knee inluries: need for clinical
decision rules. Acad Emerg Med. 1995
Nov;2(11):966-73.

11. Weber JE, Jackson RE, Peacock WF, Swor RA, Carley
R, Larkin GL. Clinical decision rules discriminate
between fractures and nonfractures in acute isolated
knee trauma. Ann Emerg Med. 1995 Oct;26(4):429-33.

12. Bauer SJ, Hollander JE, Fuchs SH, Thode HC Jr. A
clinical decision rule in the evaluation of acute knee
injuries J Emerg Med. 1995 Sep-Oct;13(5):611-5.

13. Seaberg DC, Jackson R. Clinical decision rule for
knee radiographs. Am J Emerg Med. 1994
Sep;12(5):541-3.

14. Jones AC, Ledingham J, McAlindon T, Regan M,
Hart D, MacMillan PJ, Doherty M. Radiographic
assessment of patellofemoral osteoarthritis. Ann
Rheum Dis. 1993 Sep;52(9):655-B.

References

(1) Gratton MC, Salomone J III, Watson WA. Clinically significant radiograph misinterpretations at an emergency medicine residency program. Ann Emerg Med. 1990; 19:497-502.

(2) Stiell IG, Wells GA, Hoag RH, et al. Implementation of the Ottawa Knee Rule for the use of radiography in acute knee injuries. JAMA. 1997;278:2075-2079.

(3) Jackson JL, O'Malley PG, Koenke K. Evaluation of acute knee pain in primary care. Ann Intern Med. 2003; 139:575-588.

(4) Sackett DL. The rational clinical examination: a primer on the precision and accuracy of the clinical examination. JAMA. 1992;267:2638-2644.

(5) Laupacis A, Sekar N, Stiell IG. Clinical prediction rules: a review and suggested modifications of methodological standards. JAMA. 1997;277:488-494.

(6) McGinn TG, Guyatt GH, Wyer PC, et al. Users' guides to the medical literature, XXII: how to use articles about clinical decision rules. JAMA. 2000;284:79-84.

(7) Stiell IG, Greenberg GH, Wells CA, et al. Derivation of a decision rule for the use of radiography in acute knee injuries. Ann Emerg Med. 1995;26:405-413.

(8) Emparanza JI, Aginaga JR. Validation of the Ottawa Knee Rules. Ann Emerg Med. 2001;38:364-368.

(9) Szucs PA, Richman PA, Mandell M. Triage nurse application of the Ottawa Knee Rule. Acad Emerg Med. 2001;8:112-116.

(10) Ketelslegers E, Collard X, Vande Berg B, et al. Validation of the Ottawa Knee Rules in an emergency teaching center. Eur Radiol. 2002;12:1218-1220.

(11) Bachmann LM, Haberzeth S, Stuerer J, ter Riet G. The accuracy of the Ottawa Knee Rule to rule out knee fractures: a systematic review. Ann Intern Med. 2004; 140:121-124.

(12) Sackett DL, Straus SE, Richardson WS, et al. Evidence-Based Medicine: How to Practice and Teach EBM. 2nd ed. Edinburgh, United Kingdom: Churchill Livingstone Inc; 2000.

(13) Oxman A, Guyatt G, Cook D, Montori V. Summarizing the evidence. to: Guyatt G, Drummond R, eds. User's Guides to the Medical Literature: A Manual for Evidence-Based Clinical Practice. Chicago, Ill: AMA Press; 2002:155-172.

(14) Cohen DM, Jasser JW, Kean JR, Smith GA. Clinical criteria for using radiography for children with acute knee injuries. Pediatr Emerg Care. 1998;14:185-187.

(15) Ingul B, Rogers M. Searching for clinical prediction rules in MEDLINE. J Am Med Inform Assoc. 2001;8:391-397.

(16) Bachmann LM, Coray R, Estermann P, ter Riet G. Identifying diagnostic studies in MEDLINE: reducing the number needed to read. J Am Med Inform Assoc. 2002;9:653-658.

(17) Khine H, Dorfman DH, Avner JR. Applicability of Ottawa Knee Rule for knee injury in children. Pediatr Emerg Care. 2001;17:401-404.

(18) Bulloch B, Neto G, Plint A, et al. Validation of the Ottawa Knee Rule in children: a multicenter study. Ann Emerg Med. 2003;42:48-55.

(19) Lee JK, Yao L, Phelps CT, et al. Anterior cruciate ligament tears: MR imaging compared with arthroscopy and clinical tests. Radiology. 1988;166:861-864.

Britt Smith, PT, MSPT, OCS, FAA OMPT, is a private practice physical therapist, S.O.A.R. Physical Therapy, Grand Junction, Colo.

Joshua A Cleland, PT, DPT, OCS, is Assistant Professor, Physical Therapy Program, Franklin Pierce College; Physical Therapist, Rehabilitation Services of Concord Hospital, Concord, NH; and Fellow, Manual Therapy Program, Regis University, Denver, Colo.