Article, Orthopedics

Point-of-care ultrasound for the diagnosis of shoulder dislocation: A systematic review and meta-analysis

a b s t r a c t

Introduction: Shoulder dislocations are a common injury causing patients to present to the emergency depart- ment. Point-of-care ultrasound has the potential to reduce time, radiation exposure. and healthcare costs among patients presenting with shoulder dislocations. We performed this systematic review and meta- analysis to determine the diagnostic accuracy of ultrasound compared with Plain radiography in the assessment of shoulder dislocations.

Methods: PubMed, Scopus, CINAHL, LILACS, the Cochrane databases, Google Scholar, and bibliographies of se- lected articles were assessed for all prospective and randomized control trials evaluating the accuracy of POCUS for identifying shoulder dislocation. Data were dual extracted into a predefined worksheet and quality analysis was performed with the QUADAS-2 tool. Data were summarized and a meta-analysis was performed with subgroup analyses by technique. Diagnostic accuracy of identifying associated fractures was assessed as a secondary outcome.

Results: Seven studies met our inclusion criteria, comprising 739 assessments with 306 dislocations. Overall, POCUS was 99.1% (95% CI 84.9% to 100%) sensitive and 99.9% (95% CI 88.9% to 100%) specific for the diagnosis

of shoulder dislocation with a LR+ of 796.2 (95% CI 8.0 to 79,086.0) and a LR- of 0.01 (95% CI 0 to 0.17).

There was no statistically significant difference between techniques. POCUS was also 97.9% (95% CI 10.5% to 100%) sensitive and 99.8% (95% CI 28.0% to 100%) specific for the diagnosis of associated fractures.

Conclusions: POCUS is highly sensitive and specific for the identification of shoulder dislocations and reductions, as well as associated fractures. POCUS may be considered as an alternate diagnostic method for the management of shoulder dislocations.

(C) 2019

Introduction

Acute shoulder dislocations are a common injury, affecting between 15 and 40 patients per 100,000 person-years [1-3]. These injuries have been estimated to affect nearly 2% of the population, leading to over 200,000 Emergency Department (ED) visits each year [4,5].

Traditionally, shoulder dislocations are reduced in the ED with ra- diographs obtained initially to identify the dislocation and afterward to confirm the reduction. However, these routine radiographs expose patients to radiation and can result in significant time delays, as well as increased healthcare costs. Moreover, radiographs may miss some posterior dislocations and usually require transfer to a radiology suite for adequate films, which means that some patients may require re- sedation for a repeat Reduction attempt if the initial attempt is unsuc- cessful [5].

* Corresponding author at: 1750 West Harrison Street, Suite 108 Kellogg, Chicago, IL 60612, United States of America.

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

Consequently, there has been increasing interest in the use of point- of-care shoulder sonography to identify dislocations and reductions rapidly at the bedside [6-8]. Point-of-care ultrasound is non- invasive, inexpensive, readily available in most EDs, and can be per- formed while a patient remains sedated after a reduction attempt. While early literature suggested that POCUS may be beneficial for dislo- cations, a number of recent large studies have been published prompting the need for this review [9-16].

The primary objective of this study is to determine the diagnostic ac- curacy of POCUS for identifying shoulder dislocation and reduction when compared with a gold standard of x-rays. An a priori subgroup analysis was planned based upon the ultrasound technique. The sec- ondary objective was to determine the diagnostic accuracy of POCUS for identifying fractures associated with the dislocation.

Methods

Our study conforms to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses – Diagnostic test accuracy (PRISMA-DTA) guidelines for systematic reviews and was performed in accordance

https://doi.org/10.1016/j.ajem.2019.02.024

0735-6757/(C) 2019

Fig. 1. PRISMA flow diagram.

with best practice guidelines [17]. This review was registered with PROSPERO (CRD42019118887). In conjunction with a medical librarian, we conducted a search of PubMed, the Cumulative Index of Nursing and Allied Health (CINAHL), the Latin American and Caribbean Health Sci- ences Literature database (LILACS), Scopus, Google Scholar, the Cochrane Database of Systematic Reviews, and the Cochrane Central Register of Controlled Trials to include citations from inception to Janu- ary 4th, 2019. The authors also hand-searched the conference abstracts for the annual scientific meetings of the American College of Emergency Physicians, the American Institute of Ultrasound in Medicine, and the Society for Academic Emergency Medicine. Details of the search strat- egy are included in the Appendix. We reviewed the bibliographies of identified studies and review articles for potential missed articles. We also consulted with topic experts to help identify any further relevant studies.

Inclusion and exclusion criteria

Inclusion criteria consisted of all prospective or randomized control trials assessing ultrasound for the identification of shoulder dislocation. There were no language, date, or age restrictions. All studies must have

had a confirmatory test (e.g., radiograph or computed tomogram). We excluded case reports, retrospective studies, and cadaver studies.

Two investigators independently assessed studies for eligibility based upon the above criteria. All abstracts meeting initial criteria were reviewed as full manuscripts. Studies determined to meet the eli- gibility criteria on full text review by both extractors were included in the final data analysis. Any discrepancies were resolved by consensus.

Data collection and processing

Two investigators independently extracted data from the included studies. The investigators underwent initial training and extracted data into a pre-designed data collection form. The following information was abstracted: last name of the first author, Publication year, study country, study population size, type of study (e.g., prospective or ran- domized controlled trial), study location (e.g., ED, Intensive Care Unit), study inclusion criteria, study exclusion criteria, mean age of study pa- tients, gender of study patients, percentage of shoulder dislocations, percentage of each dislocation type (i.e., anterior, posterior, inferior), ul- trasound transducer, ultrasound technique, sonographer training, oper- ator specialty, operator experience (i.e., attending or non-attending

Table 1

Characteristics of the included studies.

Study

Study population

Country

Study location

Mean patient age

Male patients (%)

Shoulder dislocations (%)

Ultrasound transducer

Ultrasound technique

Operator experience

Bianchi 1994

10

United States

ND

35.5 years

ND

2 (20%)

Linear

Posterior

ND

Abbasi 2013

142

Iran

ED

31.6 years

126 (91.3%)

71 (50%)

Linear

Anterior + lateral

Attending, resident

Ahmadi 2016

108

Iran

ED

30.1 years

91 (64.1%)

13 (12%)

Linear

Anterior + lateral

Attending

Akyol 2016

197

Turkey

ED

33.9 years

164 (80.6%)

99 (50%)

Linear

Posterior

Attending

Lahham 2016

84

United States

ED

45 years

52 (62%)

19 (22.6%)

Linear

Posterior

Student

Seyedhosseini 2017

163

Iran

ED

35.9 years

140 (83.3%)

79 (48.5%)

Curvilinear

Posterior

Attending, resident

Secko 2018

35

United States

ED

ND

ND

23 (66%)

Linear or curvilinear

Posterior

Attending

ND, not described; ED, emergency department.

Fig. 2. Forest diagram of the overall sensitivity and specificity of ultrasound for identifying shoulder dislocations.

physician), and true positives, false positives, true negatives, and false negatives for both identifying dislocations and identifying fractures. When studies included assessments of both dislocation and reduction confirmation, the data was combined for meta-analytic purposes. Stud- ies were independently assessed for quality by two separate investiga- tors utilizing the Quality Assessment of Diagnostic Accuracy Studies (QUADAS-2) tool [18]. Any discrepancies were resolved by consensus.

Data analysis

Diagnostic test accuracy variables, including sensitivity, specificity, positive likelihood ratio (LR+), and negative likelihood ratio , were calculated using a bivariate random effects model. For subgroup analysis when fewer than four studies were included, a univariate anal- ysis was conducted. All data were calculated with 95% confidence inter- vals (CIs). Heterogeneity was assessed using Chi-square and I2 statistics. A p-value b0.1 or an I2 N50% was considered significant for heterogene- ity [19]. We constructed a summary receiver-operating characteristic (SROC) curve with observed study data and a 95% CI confidence region. A linear regression test of funnel plot asymmetry with a p-value b0.1 for the slope coefficient was considered significant for asymmetry.

Statistical analysis was completed with the MIDAS module for StataMP, version 13 (StataCorp LP, College Station, Texas) to perform analyses, including bivariate random effects analyses, SROC curve anal- ysis, assessment of publication bias, and to construct graphs. The DIAGT module was utilized for univariate analysis. Forest plots were con- structed using RevMan (The Nordic Cochrane Centre, Copenhagen, Denmark), version 5.3.

Fig. 3. Summary receiver operating characteristics of ultrasound for identifying shoulder dislocations.

Results

A total of 800 studies were identified. PubMed yielded 220 studies, Scopus identified 334 studies, CINAHL found 80 studies, LILACS discov- ered 66, the Cochrane Database of Systematic Reviews yielded no stud- ies, and the Cochrane Central Register of Controlled Trials identified no studies. In addition, the initial 100 studies from Google Scholar were also included as recommended by Bramer and colleagues [20]. After re- moving duplicates, 475 original abstracts were reviewed with 16 se- lected for full text review (Fig. 1). No additional papers were identified by the topic experts or through bibliographic review.

Seven studies, comprising 739 total assessments, were selected for

the final analysis (Table 1). All seven studies were prospective, observa- tional trials. Three studies were conducted in Iran, [11,12,15] three were performed in the United States, [10,14,16] and one took place in Turkey [13]. Six studies took place in the ED, [11-16] while one paper did not describe the study location [10]. The majority of POCUS examinations were performed by Emergency Medicine providers [11-13,15,16]. Five studies used a linear transducer, [10-14] while one study used a curvi- linear [15] and one allowed either a linear or curvilinear to be used [16]. The mean age was 35.1 years and 82.3% of patients were male. The overall dislocation rate was 41.4%. All dislocations were anteriorly displaced with the exception of two posterior dislocations.

Overall, POCUS had a 99.1% (95% CI 84.9% to 100%) sensitivity and

99.9% (95% CI 88.9% to 100%) specificity for the diagnosis of shoulder dislocation with a LR+ of 796.2 (95% CI 8.0 to 79,086.0) and a LR- of

0.01 (95% CI 0 to 0.17) (Fig. 2). The area under the SROC curve indicated high accuracy (0.999; 95% CI 0.990 to 1.000) (Fig. 3). Statistical hetero- geneity was high with an I2 of 85%. Funnel plot analysis demonstrated some evidence of publication bias (Fig. 4).

When assessing only the posterior technique, POCUS had a 99.0% (95% CI 92.3% to 99.9%) sensitivity and 99.7% (95% CI 89.5% to 100%) specificity for the diagnosis of shoulder dislocation with a LR+ of

338.7 (95% CI 8.6 to 13,395.8) and a LR- of 0.01 (95% CI 0 to 0.08)

Fig. 4. Funnel plot of studies assessing the accuracy of ultrasound for identifying shoulder dislocations.

Fig. 5. Forest diagram of the overall sensitivity and specificity of ultrasound for identifying fractures.

(Supplemental Fig. 1). When assessing the anterior and lateral tech- nique, POCUS had a 92.9% (95% CI 85.1% to 97.3%) sensitivity and 100% (95% CI 97.8% to 100%) specificity for the diagnosis of shoulder disloca- tion with a LR+ of 308.5 (95% CI 19.4 to 4914.4) and a LR- of 0.08 (95% CI 0.04 to 0.16) (Supplemental Fig. 2).

When assessing for the presence of an associated fracture, POCUS had a 97.9% (95% CI 10.5% to 100%) sensitivity and 99.8% (95% CI

28.0% to 100%) specificity with a LR+ of 399.2 (95% CI 0.4 to 400,138.9) and a LR- of 0.02 (95% CI 0 to 7.37) (Fig. 5). Statistical het- erogeneity was high with an I2 of 80%.

Studies were at overall low risk of bias and applicability concerns for most parameters (Table 2). Six studies were at unclear risk of bias for patient selection due to the use of a convenience sample [10-14,16]. Two studies [10,13] were at unclear risk of bias for the index test and three studies [10,13,15] were at unclear risk of bias for the reference standard due to inadequate description of the blinding technique. One study was at unclear risk of applicability for patient selection due to ex- clusion of posterior dislocations and patients with a body-mass index of N35 [12]. One study was at unclear risk of applicability for the reference standard due to the use of a single-view radiograph for confirming re- ductions [12]. No studies were at high risk for bias or applicability.

Limitations

It is important to consider several limitations with respect to this systematic review and meta-analysis. First, all studies were prospective, observational trials. There were no randomized control trials that were identified in this review. Therefore, it is possible there may be some un- identified confounders within the included studies. However, we did not identify any significant confounders in our review and the consis- tency of results suggests this is less likely. Nevertheless, future studies should consider assessing this modality with randomized control trials. Additionally, providers were not blinded to the physical examination findings, which may have influenced their decisions. However, because POCUS will be performed on patients at the bedside, we believe this is acceptable as it replicates how this would be applied in real-world con- ditions. Moreover, most dislocations were anterior in nature with only two posterior dislocations. While both posterior dislocations were cor- rectly identified, more data are needed to determine the diagnostic ac- curacy in this important population. There was also significant statistical heterogeneity, which may have been due to variations in the training and Experience level of the sonographers. While it was not possible to perform a subgroup analysis on this with the available data, we did not identify a significant difference between the diagnostic accuracy of the studies to suggest a significant influence of either factor.

Future studies are needed to determine the ideal training protocol and requisite experience for this modality. Finally, it is possible that we may have missed some potentially relevant articles for inclusion. How- ever, we performed an extensive search strategy with the assistance of a medical librarian experienced in systematic reviews, as well as searches of bibliographies and discussion with content experts, so we believe the risk of this is minimal.

Discussion

Our systematic review and meta-analysis demonstrates that POCUS is highly sensitive and specific for identifying shoulder dislocation and reductions. Additionally, no significant difference was identified when comparing the assessment techniques.

This data has significant patient and provider implications with re- spect to time, cost, and radiation. One study found that pre-reduction radiographs alone delayed the time to treatment by 30 minutes [21]. This has the potential to worsen muscular spasm, resulting in a more difficult reduction, as well as increase the overall patient length of stay. POCUS could be performed rapidly at the bedside to diagnose the shoulder dislocation, as well as confirm proper relocation after the re- duction attempt. In the latter case, this would avoid the need to re- sedate the patient after the attempt if it was not successfully relocated (by identifying the need for a repeat attempt while the patient remains under the initial sedation). Additionally, POCUS may decrease overall healthcare costs by reducing the total number of radiographs per- formed, especially given the high incidence and prevalence of this con- dition [1-5].

Interestingly, POCUS was also sensitive and specific for identifying associated fractures. Unfortunately, the confidence intervals were rela- tively wide and further studies are needed to better determine the diag- nostic accuracy and clinical utility of POCUS for this modality [21-23].

Two previous systematic review were performed in this topic in 2015 and 2017 [9,24]. The most recent systematic review by Gottlieb and Russell identified four studies comprising 531 total assessments and suggested that POCUS may be a considered as a potential modality for identifying shoulder dislocations but suggested that more studies were needed [24]. Our current review performed a more expanded search and was conducted two years after the most recent review, iden- tifying three additional studies and increasing the total assessments by 30%. Additionally, this the first review to perform a meta-analysis, as well as a subgroup analysis by the assessment technique. Finally, this is the first systematic review to also assess the diagnostic accuracy of POCUS for identifying associated fractures.

Table 2

QUADAS-2 for included studies.

Study

Risk of bias

Applicability concerns

Patient selection

Index test

Reference standard

Flow and timing

Patient selection

Index test

Reference standard

Bianchi 1994

U

U

U

L

L

L

L

Abbasi 2013

U

L

L

L

L

L

L

Ahmadi 2016

U

L

L

L

U

L

U

Akyol 2016

U

U

U

L

L

L

L

Lahham 2016

U

L

L

L

L

L

L

Seyedhosseini 2017

L

L

U

L

L

L

L

Secko 2018

U

L

L

L

L

L

L

L, low risk of bias; U, unclear risk of bias.

As with all ultrasound applications, there can be operator variability. While the consistently high accuracy with a range of sonographers in this review suggests a broader applicability of this skill, future studies should determine the ideal training protocol for shoulder sonography. Additionally, future studies should compare the two primary POCUS techniques to determine which technique is the most accurate.

Conclusion

POCUS is highly sensitive and specific for the identification of shoul- der dislocations and has the potential to decrease time, healthcare costs, and radiation exposure. POCUS may be considered as an alternate diag- nostic method for the management of shoulder dislocations.

Supplementary data to this article can be found online at https://doi. org/10.1016/j.ajem.2019.02.024.

Disclosures/funding

This research did not receive any specific grant from funding agen- cies in the public, commercial, or not-for-profit sectors. We have no dis- closures to declare.

Meetings

None.

IRB

N/A.

Acknowledgements

The authors would like to thank Jennifer C. Westrick, MSLIS for her assistance with the literature search.

Appendix A. Search strategy

PubMed: 220

(“Shoulder Dislocation”[Mesh] or “Shoulder dislocation” or “Shoul- der dislocations” or “shoulder relocation” or “shoulder reduction” or “shoulder reductions” or “dislocation of the shoulder” or “dislocated shoulder” or “dislocated shoulders” or “shoulder instability” or “shoul- der subluxation”)

AND (“Ultrasonics”[Mesh] OR “Ultrasonography”[Mesh] OR ultraso* OR sonogra* or POCUS)

Scopus: 334

(TITLE-ABS-KEY(({Shoulder dislocation} or {Shoulder dislocations} or {shoulder relocation} or {shoulder reduction} or {shoulder reduc- tions} or {dislocation of the shoulder} or {dislocated shoulder} or

{dislocated shoulders} or {shoulder instability} or {shoulder subluxation}))

AND (TITLE-ABS-KEY(ultraso* OR sonogra* or POCUS)) CINAHL: 80

Cochrane Central Register of Controlled clinical trials: 16 Cochrane Database of Systematic Reviews: 0

((MH “Shoulder Dislocation”) OR (“Shoulder dislocation” or “Shoul- der dislocations” or “shoulder relocation” or “shoulder reduction” or “shoulder reductions” or “dislocation of the shoulder” or “dislocated shoulder” or “dislocated shoulders” or “shoulder instability” or “shoul- der subluxation”))

AND ((MH “Ultrasonography”) OR (ultraso* OR sonogra* or POCUS))

LILACS: 66

((“shoulder dislocation” or “dislocated shoulder” or “shoulder re- duction” or “shoulder relocation” or “shoulder subluxation”)) AND (ultraso* OR sonogra* or POCUS)

Google Scholar: First 100

(“Shoulder dislocation” OR “shoulder reduction” OR “shoulder relo- cation” OR “dislocation of the shoulder” OR “dislocated shoulder” OR “shoulder instability”)

AND (ultrasound OR ultrasonography OR sonograph OR sonogra- phy) -guided

References

  1. Shah A, Judge A, Delmestri A, Edwards K, Arden NK, Prieto-Alhambra D, et al. Inci- dence of shoulder dislocations in the UK, 1995-2015: a population-based cohort study. BMJ Open 2017 Nov 14;7(11):e016112.
  2. Leroux T, Wasserstein D, Veillette C, Khoshbin A, Henry P, Chahal J, et al. Epidemiol- ogy of primary Anterior shoulder dislocation requiring closed reduction in Ontario, Canada. Am J Sports Med 2014 Feb;42(2):442-50.
  3. Zacchilli MA, Owens BD. Epidemiology of shoulder dislocations presenting to emer- gency departments in the United States. J Bone Joint Surg Am 2010 Mar;92(3): 542-9.
  4. Youm T, Takemoto R, Park BK. acute management of shoulder dislocations. J Am Acad Orthop Surg 2014 Dec;22(12):761-71.
  5. Gottlieb M. Current approach to the diagnosis and management of shoulder disloca- tion in children. Pediatr Emerg Care 2018 May;34(5):357-62.
  6. Blakeley CJ, Spencer O, Newman-Saunders T, Hashemi K. A novel use of portable ul- trasound in the management of shoulder dislocation. Emerg Med J 2009;26(9): 662-3.
  7. Halberg MJ, Sweeney TW, Owens WB. Bedside ultrasound for verification of shoul- der reduction. Am J Emerg Med 2009;27(1) (134.e5-6).
  8. Mackenzie DC, Liebmann O. Point-of-care ultrasound facilitates diagnosing a poste- rior shoulder dislocation. J Emerg Med 2013;44(5):976-8.
  9. Gottlieb M, Edwards H. BET 1: utility of ultrasound in the diagnosis of shoulder dis- location. Emerg Med J 2016 Sep;33(9):671-2.
  10. Bianchi S, Zwass A, Abdelwahab I. Sonographic evaluation of posterior instability and dislocation of the shoulder: prospective study. J Ultrasound Med 1994 May;13 (5):389-93.
  11. Abbasi S, Molaie H, Hafezimoghadam P, Zare MA, Abbasi M, Rezai M, et al. Diagnostic accuracy of ultrasonographic examination in the management of shoulder disloca- tion in the emergency department. Ann Emerg Med 2013 Aug;62(2):170-5.
  12. Ahmadi K, Hashemian AM, Sineh-Sepehr K, Afzal-Aghaee M, Jafarpour S, Rahimi- Movaghar V. Bedside ultrasonography for verification of shoulder reduction: a long way to go. Chin J Traumatol 2016;19(1):45-8.
  13. Akyol C, Gungor F, Akyol AJ, Kesapli M, Guven R, Cengiz U, et al. Point-of-care ultra- sonography for the management of shoulder dislocation in ED. Am J Emerg Med 2016 May;34(5):866-70.
  14. Lahham S, Becker B, Chiem A, Joseph LM, Anderson CL, Wilson SP, et al. Pilot study to determine accuracy of posterior approach ultrasound for shoulder dislocation by novice sonographers. West J Emerg Med 2016 May;17(3):377-82.
  15. Seyedhosseini J, Saiidian J, Hashemi Taheri A, Vahidi E. Accuracy of point-of-care ul- trasound using low frequency curvilinear transducer in the diagnosis of shoulder dislocation and confirmation of appropriate reduction. Turk J Emerg Med 2017 Jul; 17(4):132-5.
  16. Secko M, Reardon L, Gottlieb M, Adhikari S, Theodoro D, Bailitz J, et al. Musculoskel- etal ultrasound to diagnosis dislocated shoulders study. Ann Emerg Med 2018 Oct; 72(4):S71.
  17. McInnes MDF, Moher D, Thombs BD, McGrath TA, Bossuyt PM, Clifford T, et al. Pre- ferred reporting items for a systematic review and meta-analysis of diagnostic test accuracy studies: the PRISMA-DTA statement. JAMA 2018;319(4):388-96.
  18. Whiting PF, Rutjes AW, Westwood ME, Mallett S, Deeks JJ, Reitsma JB, et al. QUADAS-2: a revised tool for the quality assessment of diagnostic accuracy studies. Ann Intern Med 2011 Oct 18;155(8):529-36.
  19. Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta- analyses. BMJ 2003 Sep 6;327(7414):557-60.
  20. Bramer WM, Rethlefsen ML, Kleijnen J, Franco OH. Optimal database combinations for literature searches in systematic reviews: a prospective exploratory study. Syst Rev 2017 Dec 6;6(1):245.
  21. Shuster M, Abu-Laban RB, Boyd J. Prereduction radiographs in clinically evident an- terior shoulder dislocation. Am J Emerg Med 1999;17(7):653-8.
  22. Hendey GW, Kinlaw K. Clinically significant abnormalities in postreduction radio- graphs after anterior shoulder dislocation. Ann Emerg Med 1996 Oct;28(4): 399-402.
  23. Gottlieb M, Nakitende D, Krass L, Basu A, Christian E, Bailitz J. Frequency of fractures identified on post-reduction radiographs after shoulder dislocation. West J Emerg Med 2016 Jan;17(1):35-8.
  24. Gottlieb M, Russell F. Diagnostic accuracy of ultrasound for identifying shoulder dis- locations and reductions: a systematic review of the literature. West J Emerg Med 2017 Aug;18(5):937-42.