Article, Orthopedics

Preservation of active range of motion after acute elbow trauma predicts absence of elbow fracture

Original Contribution

Preservation of active range of motion after acute elbow trauma predicts absence of elbow fracture

Michael A. Darracq MDa,?, David R. Vinson MDc, Edward A. Panacek MD, MPHb

aDepartment of Emergency Medicine, Naval Medical Center, San Diego, CA 92134, USA

bDepartment of Emergency Medicine, University of California, Davis, CA 95824, USA

cDepartment of Emergency Medicine, the Permanente medical group, Sacramento, CA 95817, USA

Received 13 October 2007; revised 2 November 2007; accepted 5 November 2007

Abstract

Objective: Previous studies indicate that an inability to fully extend the elbow after elbow trauma is indicative of fracture. We hypothesized that maintenance of active range of motion (ROM) of the elbow in flexion, extension, pronation, and supination after elbow trauma is very specific for the absence of fracture or effusion, and limitation of ROM is sensitive for fracture or effusion.

Methods: This was a prospective observational study with convenience sampling of a carefully selected patient population. This would allow certain types of patients to be excluded from the study. Patients with elbow injury receiving radiographs and meeting enrollment criteria were enrolled between June 2006 and March 2007 at 4 emergency departments in the local region. Demographics, active ROM, and presence of point tenderness at the olecranon, epicondyles, and radial head were recorded by enrolling clinicians. All enrolled patients received standard elbow radiographs. Radiographs were reviewed by blinded radiologists for the presence of fracture and effusion. Sensitivity, specificity, and 95% confidence intervals (CIs) of examination findings were calculated.

Results: One hundred thirteen patients were enrolled. Limitation of active ROM was 100% (95% CI, 0.93-1.00) sensitive for fracture or effusion. Preservation of active ROM was 97% (95% CI, 0.89-1.00) specific for the absence of fracture. Point tenderness was also highly sensitive but not specific.

Conclusion: Individuals with preservation of active ROM after acute elbow trauma have a Very low risk of associated fracture and may not require radiographic studies.

(C) 2008

Introduction

Background

Elbow injuries are a relatively common presenting complaint to the emergency department (ED) in both adults and children. One study reported that 2% to 3% of presenting

* Corresponding author.

E-mail address: [email protected] (M.A. Darracq).

complaints to the ED involved elbow injury. In addition, a missed fracture rate of 10.8% despite radiographic studies was observed in one study [1].

Importance

Rules for clinical decision making such as the Ottawa ankle and knee rules have been validated in clinical practice for adults [2-4] as well as children [5,6]. In addition to being highly sensitive, these rules have led to a decrease in the

0735-6757/$ – see front matter (C) 2008 doi:10.1016/j.ajem.2007.11.005

number of unnecessary radiographs [2-7]. However, no such rules have been developed or validated in upper extremity injury, specifically the elbow.

Goals

Three previous groups have evaluated the use of limitations in elbow range of motion (ROM) as a predictor of clinically significant injury [8-10]. Two previous groups [8,9] analyzed limitations in elbow extension, whereas a third group [10] derived a rule using logistic regression analysis. We hypothesized that individuals who maintain active ROM of the elbow in flexion, extension, supination, and pronation have a very low risk of associated fracture and therefore may not require Radiographic investigation. In a selected patient population, we evaluated active ROM and association with fracture. In addition, we sought to evaluate whether point tenderness at the olecranon, epicondyles, and radial head improved the sensitivity or specificity of the active ROM maneuvers in the detection of elbow fractures.

Methods

Study design and setting

This was a prospective observational study using convenience sampling. Enrollment took place between June 2006 and March 2007 at four EDs in the local region. Enrollment sites included one academic ED supporting an emergency medicine residency (approximately 60 000 annual visits) and 3 community EDs (each approximately 30000 annual visits) with variable university affiliation. Approval from institution-specific human subjects commit- tees was obtained before enrollment.

Selection of participants

Patients had to be old enough to follow directions during an examination (>=5 years of age). Less than 24 hours had to have elapsed since time of injury. Patients could not have an alteration in mental status or inability to perform examination secondary to concomitant central nervous system injury/ disease or intoxication. Patients with obvious deformities suggesting fracture or dislocation or evidence of vascular/ Neurologic dysfunction were excluded. Patients with other distracting injuries, such as long bone fractures or antecedent elbow injury or medical condition limiting baseline mobility of the elbow (eg, palsy) were also excluded.

Data collection and processing

Baseline demographics including age and sex were prospectively recorded on a standardized enrollment form. Before obtaining radiographs, examining physicians deter-

mined and recorded ability to actively flex to at least 90?, actively extend to full locked position (0?), and actively pronate and supinate to full (180?) with the patient sitting or standing in anatomical palm-forward position. These results were recorded on the standard enrollment form. Examination was performed before administration of any pain medication. Participating physicians were instructed on how to perform the ROM manuevers before beginning enrollment through demonstration, and instructions were also provided on the enrollment form. Determination of point tenderness at the olecranon, radial head, and epicondyles was also recorded. In individuals who were unable to flex to at least 90? or extend to locked position, a comparison between the degree of flexion and extension was made with the uninjured arm. In such cases, inability to flex or extend equal to the unaffected limb was considered a positive abnormal finding.

All enrolled individuals received standard 3-view radio- graphs of the injured elbow after examination. Determination of the need for radiographs was left to the examining clinician. No patient was enrolled in the study if they did not receive radiographs.

Outcome measures

Radiographs were reviewed by Board-certified radiologists for the presence of fracture and effusion (anterior sail sign, posterior fat pad sign, or both). Radiographic results were obtained through investigator review of radiographic report. Individuals without obvious radiographic fracture but with the presence of an effusion were considered positive for occult fracture [11-13]. Radiologists were blinded to the results of physical examination. Comparison between radiographic and physical examination findings was later performed to deter- mine sensitivity and specificity of examination findings for radiographic outcomes of interest (fracture or occult fracture). All patients were followed for a period of 3 months through chart review for subsequent visit with an elbow complaint. No patients in the nonfracture population returned complaining of elbow pain during the 3-month follow-up period.

Primary data analysis

Descriptive statistics, sensitivity, specificity, and 95% confidence intervals (CIs) were calculated using Graph Pad (GraphPad Software Inc, San Diego, Calif) software by a study author (MAD) who did not participate in examination of enrolled participants.

Results

Characteristics of study subjects

One hundred thirteen individuals were enrolled according to specified inclusion criteria. Fifty-three enrolled individuals

had an obvious fracture or an occult fracture, as defined by the presence of effusion without obvious fracture. Demographics are summarized in Table 1. No significant differences were observed between sexes as to prevalence of fracture.

Main results

Any overall limitation in active ROM was 100% (95% CI, 0.93-1.00) sensitive for the presence of fracture. Inability to extend to full-locked position was also 100% (95% CI, 0.93- 1.00) sensitive, whereas flexion, supination, and pronation were each of significantly lower sensitivity. (see Table 2). Point tenderness at the olecranon, epicondyles, and radial head was 100% (95% CI, 0.93-1.00) sensitive for the

presence of fracture but only 67% (95% CI, 0.53-0.78) specific. The addition of point tenderness at the olecranon, epicondyles, and radial head did not improve the sensitivity or specificity of the active ROM maneuvers in detection of elbow fractures. Preservation of overall active ROM was 97% (95% CI, 0.89-1.00) specific for the absence of fracture. Extension to full locked position (0?), flexion to at least 90?, and pronation were each 100% (95% CI, 0.94-1.00) specific, whereas supination was 97% (95% CI, 0.89-1.00) specific.

Discussion

This was a prospective observational study using structured observations and recording of data. The inclusion criteria for enrollment included an age greater than or equal to 5 years, an elapsed time since injury of less than 24 hours, and an absence of alteration in mental status or inability to perform examination secondary to concomitant central nervous system injury/disease or intoxication. Patients with obvious deformities suggesting fracture or

Sensitivity (95% CI) Specificity (95% CI)

Overall ROM

100% (0.93-1.00)

97%

(0.885-1.00)

Extension

100% (0.93-1.00)

100%

(0.94-1.00)

Flexion

64% (0.50-0.69)

100%

(0.94-1.00)

Supination

43% (0.30-0.58)

97%

(0.885-1.00)

Pronation

34% (0.22-0.48)

100%

(0.94-1.00)

Point tenderness

100% (0.93-1.00)

67%

(0.53-0.78)

dislocation or evidence of vascular/neurologic dysfunction were excluded. Patients with other distracting injuries, such as Long bone fractures or antecedent elbow injury or medical condition limiting baseline mobility of the elbow (eg, palsy) were also excluded.

Table 2 Active range of motion sensitivity and specificity

Selection bias may have been introduced because of convenience sampling. Not all patients who met inclusion criteria were enrolled due to clinician nonparticipation/ nonenrollment, patient nonparticipation, or other factors. Attempts at limiting the effect of selection bias were made through the use of research assistant screening of patients with elbow trauma presenting to the academic ED for most hours on nearly all of the study days. Most patients were enrolled at this institution. Spectrum bias may have occurred because we enrolled patients only when the clinician felt a radiograph was indicated. However, these sites generally practiced a liberal ordering of radiographs. We did not evaluate interobserver examination variability. The sample size is small and contributes to relatively wide 95% CIs for both sensitivity and specificity. However, even with this small size, lack of full ROM of the elbow joint appears highly sensitive for fracture. We did not evaluate the effect of patient expectation of radiographic studies, nor did we evaluate mechanism of injury or patients who presented later than 24 hours from time of

Table 1 Patient demographic

All participants Age <=18 years Age N18 years

Fracture

No fracture

Fracture

No fracture

Fracture

No fracture

Number (n)

53

60

27

15

26

45

Male (%)

59%

55%

59%

60%

58%

53%

Mean age (+-SD)

24.1 (17.1)

34.2 (20.2)

9.2 (3.3)

12 (4.1)

39.5 (2.4)

41.6 (2.7)

25-75IQR

9-51

18-50

6-13

9-14

29.5-49

25-54

Radiographic abnormality

53

None

27

None

26

None

Effusion only (occult fracture)

4

1

3

Supracondylar fracture

8

7

1

Medial condyle/epicondyle fracture

5

3

2

Lateral condyle/epicondyle fracture

7

6

1

Coronoid process fracture

3

1

2

Radial head fracture

17

5

12

Olecranon fracture

7

3

4

Capitellum fracture

1

1

0

Galleazzi fracture

1

0

1

injury. Individuals with medical conditions that contribute to elbow effusions (ie, rheumatoid arthritis) and, thus, potential limitation in ROM were not included in this study. We did not study nonfracture elbow injuries such as severe sprains. Individuals with alteration in mental status or central nervous system disease/intoxication were considered to have an unreliable examination and, therefore, were excluded from the study. Individuals with obvious deformities, vascular, or neurologic dysfunction were felt to require radiographs regardless of examination findings.

In this prospective study, all of the patients with radiographically confirmed or occult fracture (presence of effusion without fracture) had limitations in active ROM and, more specifically, in active full extension. In addition, all patients with radiographically confirmed or occult fracture also had point tenderness on examination, but lack of point tenderness was non specific. Preservation of overall active ROM and more specifically in flexion, extension, pronation and supination are of significantly high specificity as to predict the absence of fracture. We found that individuals with preservation of full active ROM after acute elbow trauma have a very low risk of associated fracture and may not require further radiographic investiga- tion. Further studies should be performed to validate this finding. Ours is the first study to prospectively evaluate the sensitivity and specificity of flexion, pronation, and supination in addition to extension. This is also the first study to evaluate the sensitivity and specificity of point tenderness on examination.

Hawskworth and Freeland [8] and Docherty et al [9] both evaluated inability to extend the elbow and found relatively high sensitivity–90.7% and 97.3%, respectively–for fracture or occult fracture. Each of these studies was limited by small enrollment, which contributed to very wide CIs. Multiple differences in enrollment criteria exist between these two studies and our own. In the Hawksworth and Freeland study, the manner of examination was not specified (active or passive). In addition, Hawksworth and Freeland included all patients who presented with an elbow injury regardless of time since injury, whereas we enrolled only patients within 24 hours of injury. Neither the Hawskworth and Freeland nor Docherty et al studies considered antecedent limitations in ROM nor comparison between the affected and unaffected limb. This potentially may have contributed to the lower specificity–69.4 and 69.5 respec- tively–for absence of fracture (confirmed or occult) observed in these studies. Each study also had several patients with nondisplaced radial head fractures who were able to fully extend, a finding which may have contributed to a reduction in sensitivity. None of our 17 patients with a radial head fracture (nondisplaced or otherwise) were able to fully extend. Given the minimal treatment required for

nondisplaced radial head fracture (limited immobility for comfort), detection of this fracture may be less important than other fractures of the elbow.

Our results are consistent with the findings of earlier studies that the inability to extend the elbow after trauma is highly sensitive for radiographic fractures, either confirmed or occult. Most recently, Lennon et al [10] derived a decision rule based on logistic regression analysis. They found a very highly specificity for normal flexion, extension, and supina- tion in detecting normal radiographs. After logistic regression analysis, Lennon et al proposed that patients with normal extension, flexion, and supination had a very low prevalence of fracture and were thought to not require radiographs. However, they did not validate their derived rule.

Our study confirms the high specificity of normal active ROM in predicting normal radiographs and the high sensitivity of limited ROM in predicting fractures of the elbow. The addition of point tenderness did not improve the sensitivity or specificity performance of the test.

References

  1. Freed HA, Shields NN. Most frequently overlooked radiographically apparent fractures in a teaching hospital emergency department. Ann Emerg Med 1984;13:900-4.
  2. Stiell IG, Greenberg GH, McKnight RD, et al. Decision rules for the use of radiography in acute ankle injuries. Refinement and prospective validation. JAMA 1993;269:1127-32.
  3. 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;275:611-5.
  4. 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;38:364-8.
  5. Plint AC, Bulloch B, Osmond MH, et al. Validation of the Ottawa Ankle Rules in children with ankle injuries. Acad Emerg Med 1999;6:1005-9.
  6. 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.
  7. Bachmann LM, Kolb E, Koller MT, et al. Accuracy of Ottawa ankle rules to exclude fractures of the ankle and mid-foot: systematic review. BMJ 2003;326:417.
  8. Hawksworth CRE, Freeland P. Inability to fully extend the injured elbow: an indicator of significant injury. Arch Emerg Med 1991;8: 253-6.
  9. Docherty MA, Schwab RA, John O. Can elbow extension be used as a test of clinically significant injury? South Med J 2002;95:539-41.
  10. Lennon RI, Riyat MS, Hilliam R, et al. Can a normal range of elbow movement predict a normal elbow xray? Emerg Med J 2007;24:86-8.
  11. Corbett RH. Displaced fat pads in trauma to the elbow. Injury 1978;9:297-8.
  12. O’Dwyer H, O’Sullivan P. The fat pad sign following elbow trauma in adults: usefulness and reliability in suspected occult fracture. J Comput Assist Tomogr 2004;28:562-5.
  13. Skaggs DL, Mirzayan R. Posterior fat pad sign in association with occult fracture at the elbow in trauma. J Bone Joint Surg Am 1999;81: 1429-33.