• 
• 

Three Part Question

In [adults with shoulder dislocation] are [post-reduction radiographs] [necessary for discharge from the emergency department]?

Clinical Scenario

A 34-year-old man presents to the emergency department directly from the local gym with sudden-onset shoulder pain and immediate loss of function. He had lifted a weight with shoulders abducted and hyper-extended, feeling his right shoulder immediately give way. Plain radiographs demonstrated anterior dislocation of the humeral head. The shoulder is reduced under conscious sedation with good clinical result and appropriate follow-up is arranged. As you order 'routine' post-reduction radiographs, you wonder if they will really influence management in the emergency department?

Search Strategy

(((shoulder dislocat*).ti,ab OR exp "SHOULDER INJURIES"/ OR (shoulder* fracture*).ti,ab OR exp "SHOULDER INJURIES"/ OR exp "SHOULDER FRACTURES"/ OR (shoulder* injur*).ti,ab) AND ((postreduction radiograph*).ti,ab OR (post reduction radiograph*).ti,ab OR RADIOGRAPHY/ OR "DIAGNOSTIC IMAGING"/ OR (diagnostic imag*).ti,ab)) AND ((emergency service*).ti,ab OR "EMERGENCY MEDICAL SERVICES"/ OR (emergency service*).ti,ab OR (emergency department*).ti,ab)"
MEDLINE, EMBASE, CINHAL, Cochrane trials and the Cochrane Database of Systematic Reviews were searched from their inception to March 2019. Papers not written in English, or those that were case reports or small case-series were excluded. We also searched BestBETs.org.

Search Outcome

We found 143 papers, of which 9 were relevant. Hand-searching reference lists of included papers and suggestive-search functions found an additional 4 papers. Included papers are shown in Table 1. We found one previous BestBET on this topic, published online in 2008, however this was incomplete.

Relevant Paper(s)

Author, date and country	Patient group	Study type (level of evidence)	Outcomes	Key results	Study Weaknesses
Milzman et al. 2018 USA	1,830 patients presenting to ED with anterior shoulder dislocation.	Multi-centre retrospective cohort study.	Do pre-/post-reduction radiographs alter management?	84% had pre-reduction radiographs, 95% had post-reduction radiographs. 4.83% had a new H-S or Bankhart lesion identified on post-reduction radiographs (p<0.001). No change in clinical management reported as the result of any of these findings.	Abstract only. Insufficient patient demographic data. Not possible to determine overlap in pre- and post-reduction imaging groups.
			True fracture presence determined on operative repair, CT, USS or MRI.	450 patients with negative imaging, 39% had a lesion identified on subsequent CT or operative repair.
Solovyova et al. 2017 USA	150 patients with shoulder dislocation of any cause presenting to ED requiring an orthopaedic review. Mean age of 46 (13-90) years.	Multi-centre retrospective cohort study	Fracture pattern pre- and post-reduction.	0/150 post-reduction radiographs demonstrated iatrogenic fractures or displacement of fractures identified on pre-reduction films (n=111 dislocations, n=39 fracture-dislocations). No additional iatrogenic injuries identified following reduction.	Radiographs reviewed by single individual. Retrospective design. Patients referred to orthopaedics unlikely to be representative of an unselected ED population.
Wronka et al. 2017 UK	102 patients with 104 fracture-dislocations of the proximal humerus. Average age 65 (22-69).	Single-centre retrospective cohort study	Safety of reducing fracture dislocations under sedation.	48/51 anterior dislocations with GT fracture underwent successful closed reduction, in no cases was there subsequent fracture propagation of displacement. 6/11 anterior dislocations with surgical neck and GT fractures underwent successful closed reduction, in the five failed cases there was subsequent fracture propagation or displacement.	All reductions performed by orthopaedic trainees, may not be applicable to ED setting.
Gottlieb et al. 2016 USA	185 patients with any shoulder dislocation presenting to ED. Average age 39 (16-85) years.	Single-centre retrospective cohort study	Incidence of new, clinically significant, fractures on post-reduction radiographs (influencing ED management).	0/185 clinically significant fractures identified on post-reduction imaging. 13 new fractures identified that were not considered clinically significant (12 H-S, 4 Bankart).	H-S, Bankart, and GT fractures were not considered clinically significant. Powered to a detect fracture rate of 3%.
Atoun et al. 2013 UK	92 patients. All > 40 years old (mean 66.6), with first-time anterior dislocation presenting to ED.	Single-centre retrospective cohort study	Prevalence of iatrogenic fractures on post-reduction radiographs.	5/92 patients had a post-reduction humeral neck fracture. All of these patients also had a GT fracture on initial radiographs. All had a poor surgical outcome.	Reduction method and number of attempts not included.
Khan J and Mehta S. 2007 USA	73 patients with shoulder dislocation presenting to ED (55 anterior dislocations included in analysis). Median age 39 (15-76).	Single-centre prospective observational study	Discrepancy between pre- and post-reduction radiographs as viewed by ED physician or attending radiologist.	1/55 patients reviewed by ED doctor, 1 new fracture was identified post-reduction (H-S). In n=41 patients reviewed by attending radiologist, 6 new fractures were identified post-reduction (5 HS or Bankhart, 1 glenoid). All patients had relocation confirmed on post-reduction films. Of all fractures, 37.5% were only seen on post-reduction films, none of these were thought to affect ED management.	Clinically significant defined as needing orthopaedic input. There was no long-term follow-up. Only n=41 patients had both films reviewed by attending radiologist.
Hendey et al. 2006 USA	100 patients with suspected shoulder dislocation presenting to ED (94 confirmed dislocations). Mean age of 34. Followed up by phone at day 1 and day 30.	Single-centre prospective clinical rule validation study	Reduction in radiograph use.	Overall reduction in radiograph utilisation (46%).	Post-reduction films were only protocol for fracture-dislocations. 24% of patients could not be followed-up. Underpowered to detect harms. Clinicians free to deviate from protocol.
			Number of missed fractures or dislocations.	31 patients had pre- and post-reduction films. 0/31 missed fractures or persistent dislocations in any group (95% CI 0-4%).
Shuster et al. 2002 Canada	63 patients with confirmed dislocations presenting to ED. Mean age 33 (SD 14.5).	Single-centre prospective clinical rule validation study	Safe reduction in radiograph use.	12/56 patients who had post-reduction radiographs had fractures reported. 94.9% reduction in radiographs.	No pre-reduction radiographs performed. Protocol deviation by clinicians.
Tannenbaum et al. 2001 USA	128 patients with anterior shoulder dislocation in ED.	Single-centre retrospective cohort study	Do postreduction radiographs reveal any new clinically significant fractures?	3 HS fractures identified on post-reduction films. 8 'possible' HS fractures identified and 5 'possible' chip fractures also seen. None impacted on ED management.	Letter to Editor only. Does not fully report outcome measures or patient groups.
			Do postreduction radiographs identify any previously unsuspected persistent dislocations?	1/128 unexpected persisting dislocation.
Hendey G. 2000 USA	104 patients with suspected dislocation presenting to ED (98 with confirmed dislocation).	Single-centre prospective observational study	Prevalence of missed fractures or persistent dislocations on post-reduction radiographs.	In n=76 first-time dislocations or those with a traumatic mechanism: physicians were confident is assessing 84% of joint positions, and were correct in 98% of cases. 1/76 assessed as reduced had a persistent dislocation and required ORIF. 0/76 post-reduction fractures identified.	Post-reduction views obtained may have missed some lesions. Clinical assessments completed by single-observer of different grades.
			Clinical assessment of dislocation.	In n=28 patients with atraumatic, recurrent, dislocations: physicians were confident in assessing 92% of joint positions, and were 100% correct in these. 0/28 persistent dislocations or fractures identified post-reduction.
Shuster et al. 1999 Canada	97 patients with suspicion of shoulder dislocation. Seen in either 'hill-side clinic' or ED. Median age 28 (range 16-75).	Single-centre prospective observational study	Identification of patients not requiring pre-reduction radiographs.	Fractures identified on 7/45 pre-reduction and 11/63 post-reduction radiographs (in cases where these were performed). No new fractures were identified on post-reduction films.	Significant number of patients lost to follow-up. No pre-reduction films in 36 cases.
Hendey G and Kinlaw K. 1996 USA	131 patients with anterior shoulder dislocation (175 dislocations in total), presenting to ED. Average age 29 years.	Single-centre retrospective cohort study	Incidence of new, clinically significant, post-reduction radiological abnormalities.	14/34 new H-S deformities were detected on post-reduction films. In 3/175 cases with new fracture reported post-reduction (2 Bankhart, 1 GT), it was confirmed on pre-reduction film by attending radiologist. Management was not affected by any finding. 1/175 persistent dislocation was noted on post-reduction films (subsequently successfully reduced).	Post-reduction views may have missed some lesions. Initial radiological assessment may have been confounded by subsequent detection of a fracture (not blinded).
Harvey et al. 1992 USA	57 patients with 69 anterior shoulder dislocations presenting to ED. Average age 32 years (range 15-89).	Single-centre retrospective cohort study	Need for post-reduction radiographs in uncomplicated anterior shoulder dislocation.	65/69 post-reduction radiographs did not identify a fracture or dislocation. 2/69 new H-S deformity. 2/69 persistent dislocations (requiring multiple attempts at reduction).	Does not define clinically significant fractures.

Comment(s)

Shoulder dislocations remain a common presentation to the emergency department; concerns of reduction associated injury have led to post-reduction radiographs being commonly performed. In this short-cut review we found very few injuries that were likely to affect emergency department management present uniquely on post-reduction imaging. In the one study that did report five patients with worse outcomes following reduction, all were older and had a glenoid tubercle fracture identified on pre-reduction imaging. Other new lesions identified were managed with outpatient orthopaedic follow-up.

Editor Comment

CF

Clinical Bottom Line

-Few new shoulder fractures are identified on post-reduction imaging, when these do occur they rarely affect emergency department management. -Older patients and those with pre-existing fractures are more often associated with new fractures on post-reduction imaging.

References

1. Milzman et al. Utility of X-ray for shoulder dislocations: It's past time for a change Academic Emergency Medicine 2018; Vol 25.
2. Solovyova et al. Should All Shoulder Dislocations be Closed Reduced? Assessment of Risk of Iatrogenic Injury in 150 Patients The Iowa Orthopaedic Journal 2017 vol: 37 pp: 47-52
3. Wronka et al When is it safe to reduce fracture dislocation of shoulder under sedation? Proposed treatment algorithm European Journal of Orthopaedic Surgery and Traumatology 2017 vol: 27 (3) pp: 335-340
4. Gottlieb et al. Frequency of fractures identified on post-reduction radiographs after shoulder dislocation Western Journal of Emergency Medicine 2016 vol: 17 (1) pp: 35-38
5. Atoun et al. Management of first-time dislocations of the shoulder in patients older than 40 years: the prevalence of iatrogenic fracture. Journal of Orthopaedic Trauma 2013 vol: 27 (4) pp: 190-193
6. Khan J and Mehta S. The Role of Post-Reduction Radiographs After Shoulder Dislocation Journal of Emergency Medicine 2007 vol: 33 (2) pp: 169-173
7. Hendey et al. Selective radiography in 100 patients with suspected shoulder dislocation Journal of Emergency Medicine 2006 vol: 31 (1) pp: 23-28
8. Shuster et al. Prospective evaluation of a guideline for the selective elimination of pre-reduction radiographs in clinically obvious anterior shoulder dislocation Canadian Journal of Emergency Medicine 2002 vol: 4 (4) pp: 257-262
9. Tannenbaum et al. Postreduction Radiographs for Anterior Shoulder Dislocation: A Reappraisal Annals of Emergency Medicine 2002 vol: 37 (4) pp: 417-418
10. Hendey G. Necessity of Radiographs in the Emergency Department Management of Shoulder Dislocations Annals of Emergency Medicine 2000 vol: 36 (2) pp: 108-113
11. Shuster et al. Prereduction radiographs in clinically evident anterior shoulder dislocation American Journal of Emergency Medicine 1999 vol: 17 (7) pp: 653-658
12. Hendey G and Kinlaw K. Clinically significant abnormalities in postreduction radiographs after anterior shoulder dislocation Annals of Emergency Medicine 1996 vol: 28 (4) pp: 399-402
13. Harvey et al. Are postreduction anteroposterior and scapular Y views useful in anterior shoulder dislocations? American Journal of Emergency Medicine 1992 vol: 10 (2) pp: 149-151