Accepted Manuscript

Static and Dynamic Shoulder Imaging to Predict Initial Effectiveness and Recurrence
After Ultrasound-guided Subacromial Corticosteroid Injections

Ke-Vin Chang, MD, PhD, Wei-Ting Wu, MD, Der-Sheng Han, MD, PhD, Levent
Özçakar, MD, Professor

PII: S0003-9993(17)30095-3
DOI: 10.1016/j.apmr.2017.01.022
Reference: YAPMR 56804

To appear in: ARCHIVES OF PHYSICAL MEDICINE AND REHABILITATION

Received Date: 6 December 2016

Revised Date: 22 January 2017
Accepted Date: 26 January 2017

Please cite this article as: Chang K-V, Wu W-T, Han D-S, Özçakar L, Static and Dynamic Shoulder
Imaging to Predict Initial Effectiveness and Recurrence After Ultrasound-guided Subacromial
Corticosteroid Injections, ARCHIVES OF PHYSICAL MEDICINE AND REHABILITATION (2017), doi:
10.1016/j.apmr.2017.01.022.

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.
 ACCEPTED MANUSCRIPT
Running head: Ultrasound-guided Subacromial Injections

Static and Dynamic Shoulder Imaging to Predict Initial Effectiveness

and Recurrence After Ultrasound-guided Subacromial Corticosteroid

PT
Injections

RI
Ke-Vin Chang, MD, PhD 1, Wei-Ting Wu, MD 1, Der-Sheng Han, MD, PhD 1

SC
Levent Özçakar, MD, Professor 2
1
Department of Physical Medicine and Rehabilitation, National Taiwan University

U
Hospital, BeiHu Branch and National Taiwan University College of Medicine,
AN
Taipei, Taiwan
2
Department of Physical and Rehabilitation Medicine Hacettepe University Medical
M

School, Ankara, Turkey

D
TE

Funding:

1. This research was supported by grants from the National Science Council
EP

(104-2314-B-002-022-MY2).
C

2. Research funding of National Taiwan University Hospital, Bei-Hu branch

AC

Disclosure of conflict of interest:

The authors have declared that no competing interests exist.

Corresponding author:

Ke-Vin Chang, MD, PhD. Department of Physical Medicine and Rehabilitation,

 ACCEPTED MANUSCRIPT
National Taiwan University Hospital, Bei-Hu Branch and National Taiwan University

College of Medicine, Taipei, Taiwan

Email: pattap@pchome.com.tw

IRB approval

PT
The current research was approved by the institutional review board of National

RI
Taiwan University Hospital (IRB number: 201606061RINB)

U SC
AN
M
D
TE
C EP
AC
 1
ACCEPTED MANUSCRIPT
1 Running head: Ultrasound-guided Subacromial Injections

3 Static and Dynamic Shoulder Imaging to Predict Initial Effectiveness

4 and Recurrence After Ultrasound-guided Subacromial Corticosteroid

PT
5 Injections

RI
7 Abstract

SC
8 Objective

9 To explore factors contributing to initial effectiveness and recurrence after

10
U
ultrasound (US)-guided subacromial corticosteroid injections by assessing clinical
AN
11 measurements as well as static and dynamic shoulder US images.
M

12 Design

13 Retrospective cohort study.

D

14 Setting
TE

15 Rehabilitation outpatient clinic.

EP

16 Participants

17 Adults with shoulder pain referred for injection therapy.

C

18 Interventions
AC

19 US-guided subacromial corticosteroid injection.

20 Main Outcome Measures

21 The association of initial effectiveness (defined as a more than 50% decrease in any

22 of the three pain subdomains after the first injection) and recurrent shoulder pain

23 that required repeated intervention with record-based clinical measurements and

 2
ACCEPTED MANUSCRIPT
24 static/dynamic shoulder US.

25 Results

26 The present study included 164 patients, 106 of whom were responsive to a first

27 injection. Among the 106 participants, 42 received a second injection due to

28 recurrent shoulder pain. By using the multivariate logistic regression analysis, initial

PT
29 effectiveness was positively associated with right handedness, grade 2 subacromial

RI
30 impingement during the dynamic US examination and bicipitial groove tenderness.

SC
31 However, these patients had a negative association with subdeltoid bursitis, grade 3

32 subacromial impingement and shoulder stiffness. Subdeltoid bursitis and a positive

U
33 painful arc test were predictors of recurrent shoulder pain that necessitated a
AN
34 repeated injection in the cox proportional hazards model.

35 Conclusion
M

36 The initial effectiveness and recurrence after US-guided subacromial corticosteroid

D

37 injection were associated with certain clinical measurements and static and dynamic
TE

38 shoulder US, which should be carefully evaluated (and can be used) to guide the best

39 treatment outcomes.
EP

40 Keywords
C

41 Shoulder pain, rotator cuff, subacromial impingement, ultrasound, corticosteroid

AC

42

43

44

45

46

47
 3
ACCEPTED MANUSCRIPT
48 Subacromial impingement syndrome, one of the major causes of chronic

49 shoulder pain, generally ensues due to pathologies or uncoordinated movements of

50 structures in the subacromial space.1, 2 Recent meta-analyses revealed that

51 corticosteroid injection was effective in treating subacromical impingement

52 syndrome and that ultrasound (US) guidance was superior to palpation guidance in

PT
53 shoulder pain relief, functional restoration and range of motion improvement.3-5 In a

RI
54 study employing US-guided injection of 40 mg of triamcinolone into the subacromial

SC
55 bursa, 42% of the patients with clinical diagnosis of subacromial impingement

56 syndrome had a > 50% improvement in pain.6 Performing a subacromial lidocaine

U
57 injection test prior to treatment helps to determine whether corticosteroid should be
AN
58 administered to the subacromial bursa and could improve the success rate (defined

59 as > 50% in pain reduction) by up to 64.1%.6 However, any injection carries the risk of
M

60 local infection and allergic reaction, and the use of a precedent lidocaine test
D

61 indicates at least two injections before an effective treatment.7 For this reason,
TE

62 identifying non-invasive measurements for predicting the efficacy of US-guided

63 subacromial injection is important.

EP

64 In 2014, another study evaluated US appearance of the subacromial bursa and

C

65 coracoacromial ligament deformation during dynamic US examination before a

AC

66 subacromial injection.8 No association was identified between US presentations and

67 effectiveness of the local anesthetic injection. However, in the abovementioned study,

68 only 5 mL of 1% lidocaine without corticosteroid was used and the data was analyzed

69 through univariate analysis without adjusting for relevant confounding factors.8

70 Another crucial issue is the duration of effectiveness after the injection, which has

71 been rarely investigated in this context. Accordingly, the present study sought to
 4
ACCEPTED MANUSCRIPT
72 explore factors contributing to initial effectiveness of US-guided subacromial

73 corticosteroid injection as well as measurements associated with recurrent shoulder

74 pain that necessitated a repeat intervention.

75

76 Methods

PT
77 Subjects and Study design

RI
78

SC
79 All participants were selected from a retrospective cohort of registered patients

80 with shoulder pain referred for an US examination. Since July 2012, an itemized

U
81 check-list of shoulder symptoms and physical findings had been completed before US
AN
82 scanning. Sonographic presentations of affected shoulder joints had been

83 documented in a standardized reporting form. The target population was adult

M

84 patients who had undergone US-guided subacromial corticosteroid injection. The

D

85 treatment indication was pain located at the shoulder girdle with poor response to
TE

86 prior oral medical and physical therapy. The exclusion criteria were based upon

87 previous studies 9-11 and included: (1) a previous injection within the last 6 months,
EP

88 (2) prior surgery on the affected shoulder, (3) inflammatory diseases (e.g.
C

89 rheumatoid arthritis, systemic lupus erythematosus, ankylosing spondylitis, etc.), and

AC

90 (4) concomitant intervention in the same shoulder (e.g. suprascapular nerve block).

91 The study was approved by the institutional review board and the patients were

92 subject to waivers of informed consent due to the retrospective nature of the

93 research.

94

95 Definition of study outcome

 5
ACCEPTED MANUSCRIPT
96

97 From our shoulder US registry, we extracted data of the patients who had

98 received US-guided subacromial injections. Their images had been obtained using

99 linear probes of 10-14 MHz a or 5-18 MHz b. Another senior physiatrist (with at least 5

100 years of musculoskeletal US experience) scrutinized the retrieved images and

PT
101 validated all findings in the reports independent of the physician conducting the

RI
102 scanning.

SC
103 In our center, all patients were asked to attend the follow-up clinic 2 weeks

104 after the intervention. Initial effectiveness was considered a more than 50 %

U
105 decrease in the visual analogue scales in any of the three pain subdomains (resting
AN
106 pain, night pain and pain during overhead activity) after the US guided injection.

107 Those who had not returned for the follow-up were excluded from this study.
M

108 Recurrence was defined when a repeat subacromial injection (due to recurrent
D

109 shoulder pain) was performed on the patients who had achieved initial effectiveness.
TE

110 The research assistant contacted the participants who did not attend our hospital for

111 a second injection to ensure the arrest of shoulder pain or to determine if they had
EP

112 undergone another injection in other centers.

C

113
AC

114 Clinical and sonographic assessment

115

116 The clinical registration form for shoulder US referral included pain duration,

117 visual analogue scales for shoulder pain (at rest, at night and during overhead

118 activity), bicipital groove tenderness using the contralateral shoulder as the reference,

119 Speed test, Yergason’s test, Empty can test, Neer test, Hawkins-Kennedy test, painful
 6
ACCEPTED MANUSCRIPT
120 arc test and shoulder stiffness. Shoulder stiffness was defined as more than 50%

121 limitation in shoulder range of motion in any direction.12

122 The US examinations were performed in the seated position. Biceps

123 peritendinous effusion was considered pathologic when the thickness of anechoic

124 fluid was more than 1 mm in the short axis view. A bicep tendon tear was diagnosed

PT
125 when the tendon was invisible inside or besides the bicipital groove. The criteria for

RI
126 rotator cuff tendon full thickness tear included total absence of the tendon or a focal

SC
127 deficit extending from the bursal side to the humeral side. Meanwhile, a partial

128 thickness tear was diagnosed as a noticeable cleavage inside the tendon without

U
129 extending to either bursal or articular sides. Subdeltoid bursitis was defined as fluid
AN
130 or hypertrophic synovium filling the bursa with a thickness of more than 2 mm. The

131 diagnostic criteria were similar to previous similar studies.9-11

M

132 We followed the criteria specified by Bureau NJ et al. for dynamic testing of
D

133 subacromial impingement.13 The patient was seated with the arm resting beside the
TE

134 trunk. The patient was then asked to raise their arm to the horizontal level in the

135 scapular plane with the transducer placed in the coronal section lateral to the acromion
EP

136 (Fig. 1A-B). Practice before the test was allowed until the patient was familiar with the

137 examination process. The severity was graded as follows: grade 0, no impingement;
C

138 grade 1, pain during shoulder elevation without soft tissue impingement; grade 2, pain
AC

139 with soft tissue bulging during movement; grade 3, upward migration of the humeral

140 head with an inability to rotate the humeral head beneath the acromion.

141

142 Ultrasound guided subacromial corticosteroid injection

143
 7
ACCEPTED MANUSCRIPT
144 The patient was seated with the hand placed on the buttock and the shoulder

145 externally rotated. The transducer was put along the short axis of the supraspinatus

146 tendon, where the subdeltoid bursa, appearing as a hypoechoic linear structure, was

147 clearly visualized between the overlying deltoid muscle and the underlying

148 supraspinatus tendon. The injectate (40 mg triamcinolone acetonide (1 ml) and 1 %

PT
149 lidocaine (3 ml)) was introduced via a lateral to medial approach. The needle was

RI
150 advanced into the bursa to evenly distribute the medication (Fig. 1C-D).

SC
151

152 Statistical analysis

U
153
AN
154 Using a univariate analysis, binary and categorical variables were expressed as

155 proportions and were compared using the Chi-square test or Fisher's exact test (in
M

156 case of skewed and sparse data). Continuous variables are expressed as the mean
D

157 and standard deviation and were compared using independent Student’s t-test for
TE

158 data with a normal distribution. If the data were not normally distributed, the

159 Mann-Whitney U test was used.

EP

160 Multivariate logistic regression analysis was performed to assess associations of

C

161 initial effectiveness with clinical and sonographic findings, expressed by an odds ratio
AC

162 (OR) and 95% confidence interval (CI). The group without more than 50% pain relief

163 in any of the three subdomains was set as the reference group. The analysis was

164 conducted stepwise in three models, all of which were adjusted for age, gender and

165 affected sides of shoulders. Model 1 was adjusted for static US presentations, while

166 US dynamic subacromial impingement testing was included in Model 2. Model 3 was

167 the full model that adjusted for all US findings and clinical measurements.
 8
ACCEPTED MANUSCRIPT
168 The cox proportional hazards model was employed to analyze the association

169 of recurrence with the clinical and sonographic findings, as shown by a hazard ratio

170 (HR) and 95% CI. The group without recurrence was set as the reference group. The

171 steps of including variables for adjustment were the same as those in the logistic

PT
172 regression analysis. The Kaplan-Meier curves were used to depict differences in

173 pain-free time between the subgroups. All statistical tests were performed using

RI
174 Stata version 11 C with a level of significance of α = 5 %.

SC
175

176 Result

U
177 Subjects
AN
178

179 From July 2012 to November 2016, 213 patients had been referred for a
M

180 US-guided subacromial injection. A follow-up record at 2 weeks following injection

D

181 was identified in 164 patients, 106 of whom were categorized in the group of initial
TE

182 effectiveness. Among the abovementioned 106 patients, 42 had undergone a second

183 injection and were treated as the group of recurrence. There were 64 patients who
EP

184 were defined as non-recurrence, 9 of whom having censored data (unable to be

C

185 contacted by telephone) (Fig. 2).

AC

186

187 Initial effectiveness

188

189 Compared with the treatment failure group by using the univariate analysis,

190 initially effectively treated patients had a higher rate of right handedness and a grade

191 2 US dynamic subacromial impingement, but a lower percentage of a bicep long head
 9
ACCEPTED MANUSCRIPT
192 tendon tear, shoulder stiffness and grade 3 dynamic subacromial impingement (Table

193 1). A multivariate analysis indicated that right handedness was positively associated

194 with initial effectiveness in Model 1 (OR: 2.30; 95% CI: 1.10-4.82) and Model 2 (OR:

195 2.73; 95% CI: 1.17-6.34), but not in Model 3 (borderline statistical significance, OR:

196 2.71; 95% CI: 0.98-7.50). In terms of static US findings, subdeltoid bursitis was a

PT
197 negative prognosticator of initial effectiveness in Model 2 (OR: 0.31; 95% CI:

RI
198 0.10-0.90) and Model 3 (OR: 0.27; 95% CI: 0.07-0.99). As regards the US dynamic

SC
199 examination, grade 2 subacromial impingement had a positive association with initial

200 effectiveness in Model 2 (OR: 8.93; 95% CI: 1.61-49.53) and Model 3 (OR: 9.01; 95%

U
201 CI: 1.35-59.91). Grade 3 subacromial impingement was negatively associated with
AN
202 initial effectiveness in Model 2 (OR: 0.17; 95% CI: 0.05-0.55), but became

203 insignificant in Model 3 (OR: 0.62, 95% CI: 0.14-2.73). In terms of clinical
M

204 measurements in Model 3, initial effectiveness was positively associated with

D

205 bicipital groove tenderness (OR: 3.27; 95% CI: 1.05-10.12), but was negatively
TE

206 associated with shoulder stiffness (OR: 0.10; 95% CI: 0.03-0.34) (Table 2).

207
EP

208 Recurrence after an antecedent US-guided subacromial injection

C

209
AC

210 The median pain free interval for all participants with initial effectiveness was

211 632 days and the duration was shortened to 147 days in the recurrence group. The

212 univariate analysis indicated that the recurrence group had a higher number of

213 subdeltoid bursitis cases (Table 3). Regarding the static and dynamic US

214 presentations, the multivariate logistic regression revealed that subdeltoid bursitis

215 was associated with recurrence in all three models (HR: 3.56, 95% CI: 1.54-8.24 in
 10
ACCEPTED MANUSCRIPT
216 Model 1; HR: 3.55, 95% CI: 1.46-8.64 in Model 2; HR: 3.40, 95% CI: 1.30-8.85 in

217 Model 3). In terms of physical findings, the painful arc test had a positive association

218 with recurrence in Model 3 (HR: 3.32; 95% CI: 1.14-9.65) (Table 4). The Kaplan-Meier

219 curves for pain-free time following the first injection are presented in Figure 3.

220

PT
221 Discussion

RI
222

SC
223 Our study sought to explore the factors pertinent to initial effectiveness after a

224 US-guided subacromial injection as well as recurrence following prior treatment

U
225 success. We found that initial effectiveness was positively associated with right
AN
226 handedness, grade 2 subacromial impingement and bicipitial groove tenderness and

227 was negatively associated with subdeltoid bursitis and shoulder stiffness.
M

228 Furthermore, subdeltoid bursitis and a positive painful arc test were predictors of
D

229 recurrent shoulder pain requiring a repeat intervention.

TE

230 Right shoulders exhibited better response to subacromial injections than left

231 shoulders. Although our registration form did not document hand dominance, we
EP

232 considered that most of our patients had right hand-dominance similar to the

general population.14 A previous study clearly demonstrated that shoulders at the

C

233
AC

234 dominant side were more likely to present with pain and that the activity level of the

235 affected shoulder was related to perception of pain.15 Therefore, we speculate that

236 patients with injections in the right shoulders were more likely to experience the

237 benefits of treatment during daily activities and reflected the improvement at

238 follow-up visits.

239 In our study, initial effectiveness of US-guided subacromial injection was

 11
ACCEPTED MANUSCRIPT
240 associated with dynamic US subacromial impingement tests. Grade 2 impingement

241 was found to be a predictor of treatment success. Since subacromial impingement is

242 multifactorial with possible causes, such as swollen rotator cuff tendons, thickened

243 subacromial bursa or a narrow acromiohumeral space,16 a dynamic US test seems to

244 be invaluable in exploring any functional abnormality during shoulder motions.

PT
245 Therefore, a beneficial outcome of injection from patients with grade 2 impingement

RI
246 reflected that their shoulder pain originated from uncoordinated excursion of

SC
247 subacromial components, which was theoretically more responsive to the treatment.

248 In contrast, grade 3 impingement was a disadvantageous prognosticator of

U
249 US-guided subacromial injections and the association became insignificant after
AN
250 including the confounder (i.e., shoulder stiffness) in the regression model. The most

251 plausible explanation was multicollinearity between both factors, which could be
M

252 simultaneously observed in patients with a frozen shoulder. Our results implied that
D

253 subacromial injection was not a preferable choice for pathologies derived from the
TE

254 glenohumeral joints, such as shoulder adhesive capsulitis.

255 Another finding was the positive association between initial effectiveness and
EP

256 bicipital groove tenderness. The long head of the biceps tendon, along with the
C

257 supraspinatus tendon, acts as a humeral head depressor to prevent subacromial

AC

258 impingement during arm elevation. Bicipital groove tenderness might signify overuse

259 injury of the superior labrum-biceps tendon complex due to subacromial

260 impingement, which was more susceptible to subacromial corticosteroid injection.

261 Currently, there are few articles investigating the duration of the efficacy of

262 US-guided subacromial injections. Our findings indicated that subdeltoid bursitis

263 predicted an unfavorable outcome of initial effectiveness as well as recurrence after

 12
ACCEPTED MANUSCRIPT
264 a previous injection. In the logistic regression analysis, the significance of subdeltoid

265 bursitis became recognizable after considering the US dynamic testing in Models 2

266 and 3 (Table 2). Since the directions of associations were different in grades 2 and 3

267 subacromial impingement, adjusting dynamic findings provided further clarification

268 of the predictive capability of the initial effectiveness of static US imaging. The

PT
269 observation was noteworthy because a thickened subdeltoid bursa was the most

RI
270 common indication for subacromial injection. Since the subdeltoid bursa serves as a

SC
271 cushion pad to reduce the friction force between the rotator cuff and the acromion,

272 an increase in thickness usually implies a chronic impingement process, originating

U
273 from uncoordinated movements of reciprocal structures.17 Therefore, one dose
AN
274 might not be effective in pain relief for a long-term condition and the pain might

275 recur if the improper shoulder dynamics are not corrected.

M

276 The painful arc test is a common physical examination for subacromial
D

277 impingement. It was reported to be the most specific test for partial- and
TE

278 full-thickness rotator cuff tendon tears using the sonographic findings.18 Tendon tear,

279 compared with other shoulder pathologies, leads to more pain and disability. In our
EP

280 study, a positive painful arc test might reflect a worse underlying pathology in
C

281 addition to clinical subacromial impingement and was associated with recurrent
AC

282 shoulder pain after a previously performed injection.

283 There are several clinical implications of this study. First and foremost, a

284 dynamic US subacromial impingement test should be included during routine

285 examination because different grades of impingement were shown to be associated

286 with distinct initial efficacies of US-guided subacromial injections. Second, grade 3

287 subacromial impingement and shoulder stiffness were poor prognostic factors of
 13
ACCEPTED MANUSCRIPT
288 initial treatment success. Therefore, a glenohumeral joint injection or suprascapular

289 nerve block should be a higher priority than subacromial injection in the frozen

290 shoulder.19, 20 Third, in patients with subaromial bursitis, post-injection exercise for

291 the restoration of normal shoulder kinematics should be implemented for enhancing

292 the initial effectiveness and prevention of recurrent shoulder pain.

PT
293

RI
294 Study Limitations

SC
295

296 There are several limitations to this study. First, we only tested one injectate: 40

U
297 mg of triamcinolone acetonide in 3 mL of 1% lidocaine. The outcome might vary
AN
298 according to different regimens or dosages and this needs to be explored in future

299 trials. Second, the overall success rate of the first injection could be underestimated
M

300 since patients with an improvement after the first injection may have ignored the
D

301 follow-up, thereby resulting in their exclusion from our study. Third, there were 9
TE

302 patients with initial treatment success who never received a repeat injection and

303 could not be contacted by telephone. Their data were categorized as censored data,
EP

304 which might cause some uncertainty in our results. Fourth, our research did not
C

305 include patients referred for glenohumeral joint injection for comparison. Since our
AC

306 result had revealed shoulder stiffness, a predominant symptom of adhesive capsulitis,

307 as a poor prognostic factor of subacromial injection, it is worthwhile to compare

308 clinical effectiveness between both injection techniques in patients with frozen

309 shoulder, which will be the aim of our future study.

310

311 Conclusion
 14
ACCEPTED MANUSCRIPT
312

313 The initial effectiveness of US-guided subacromial corticosteroid injection was

314 associated with certain clinical measurements and shoulder sonography.

315 Furthermore, this association varied across different grades of US dynamic

316 subcromial impingement. Subdeltoid bursitis was associated with an initial

PT
317 unfavorable outcome and pain recurrence after a prior injection. The

RI
318 post-intervention exercise for normalizing shoulder kinematics should be

SC
319 implemented in case of existing negative prognosticators of US-guided subacromial

320 corticosteroid injection.

U
321
AN
322 Reference

323 1. Greenberg DL. Evaluation and treatment of shoulder pain. Med Clin North Am
M

324 2014;98(3):487-504.
D

325 2. van der Sande R, Rinkel WD, Gebremariam L, Hay EM, Koes BW, Huisstede BM.
TE

326 Subacromial impingement syndrome: effectiveness of pharmaceutical

327 interventions-nonsteroidal anti-inflammatory drugs, corticosteroid, or other

EP

328 injections: a systematic review. Arch Phys Med Rehabil 2013;94(5):961-76.

C

329 3. Arroll B, Goodyear-Smith F. Corticosteroid injections for painful shoulder: a

AC

330 meta-analysis. Br J Gen Pract 2005;55(512):224-8.

331 4. Wu T, Song HX, Dong Y, Li JH. Ultrasound-guided versus blind

332 subacromial-subdeltoid bursa injection in adults with shoulder pain: A systematic

333 review and meta-analysis. Semin Arthritis Rheum 2015;45(3):374-8.

334 5. Mohamadi A, Chan JJ, Claessen FM, Ring D, Chen NC. Corticosteroid Injections

335 Give Small and Transient Pain Relief in Rotator Cuff Tendinosis: A Meta-analysis. Clin
 15
ACCEPTED MANUSCRIPT
336 Orthop Relat Res 2016.

337 6. Kim SJ, Lee HS. Lidocaine Test Increases the Success Rates of Corticosteroid

338 Injection in Impingement Syndrome. Pain Med 2016.

339 7. Brinks A, Koes BW, Volkers AC, Verhaar JA, Bierma-Zeinstra SM. Adverse effects

340 of extra-articular corticosteroid injections: a systematic review. BMC Musculoskelet

PT
341 Disord 2010;11:206.

RI
342 8. Bouju Y, Bouilleau L, Dubois de Montmarin G, Bacle G, Favard L. Do subacromial

SC
343 ultrasonography findings predict efficacy of intra-bursal injection? Prospective study

344 in 39 patients. Orthop Traumatol Surg Res 2014;100(8 Suppl):S361-4.

U
345 9. Hung CY, Chang KV, Ozcakar L, Wang TG, Chen WS. Can Quantification of Biceps
AN
346 Peritendinous Effusion Predict Rotator Cuff Pathologies?: A Retrospective Analysis of

347 1352 Shoulder Ultrasound. Am J Phys Med Rehabil 2016;95(3):161-8.

M

348 10. Chang KV, Chen WS, Wang TG, Hung CY, Chien KL. Quantitative ultrasound
D

349 facilitates the exploration of morphological association of the long head biceps
TE

350 tendon with supraspinatus tendon full thickness tear. PloS one 2014;9(11):e113803.

351 11. Chang KV, Chen WS, Wang TG, Hung CY, Chien KL. Associations of sonographic
EP

352 abnormalities of the shoulder with various grades of biceps peritendinous effusion
C

353 (BPE). Ultrasound Med Biol 2014;40(2):313-21.

AC

354 12. Hsieh LF, Hsu WC, Lin YJ, Chang HL, Chen CC, Huang V. Addition of intra-articular

355 hyaluronate injection to physical therapy program produces no extra benefits in

356 patients with adhesive capsulitis of the shoulder: a randomized controlled trial. Arch

357 Phys Med Rehabil 2012;93(6):957-64.

358 13. Bureau NJ, Beauchamp M, Cardinal E, Brassard P. Dynamic sonography

359 evaluation of shoulder impingement syndrome. AJR Am J Roentgenol

 16
ACCEPTED MANUSCRIPT
360 2006;187(1):216-20.

361 14. Reiss M, Reiss G. Current aspects of handedness. Wien Klin Wochenschr

362 1999;111(24):1009-18.

363 15. Keener JD, Steger-May K, Stobbs G, Yamaguchi K. Asymptomatic rotator cuff

364 tears: patient demographics and baseline shoulder function. J Shoulder Elbow Surg

PT
365 2010;19(8):1191-8.

RI
366 16. Tagg CE, Campbell AS, McNally EG. Shoulder impingement. Semin Musculoskelet

SC
367 Radiol 2013;17(1):3-11.

368 17. Daghir AA, Sookur PA, Shah S, Watson M. Dynamic ultrasound of the

U
369 subacromial-subdeltoid bursa in patients with shoulder impingement: a comparison
AN
370 with normal volunteers. Skeletal Radiol 2012;41(9):1047-53.

371 18. Kelly SM, Brittle N, Allen GM. The value of physical tests for subacromial
M

372 impingement syndrome: a study of diagnostic accuracy. Clin Rehabil

D

373 2010;24(2):149-58.
TE

374 19. Chang KV, Hung CY, Wu WT, Han DS, Yang RS, Lin CP. Comparison of the

375 Effectiveness of Suprascapular Nerve Block With Physical Therapy, Placebo, and
EP

376 Intra-Articular Injection in Management of Chronic Shoulder Pain: A Meta-Analysis of

C

377 Randomized Controlled Trials. Arch Phys Med Rehabil 2016;97(8):1366-80.

AC

378 20. Chang KV, Wu WT, Hung CY, Han DS, Yang RS, Chang CH et al. Comparative

379 Effectiveness of Suprascapular Nerve Block in the Relief of Acute Post-Operative

380 Shoulder Pain: A Systematic Review and Meta-analysis. Pain Physician

381 2016;19(7):445-56.

382

383
 17
ACCEPTED MANUSCRIPT
384 Supplier

385 a. Toshiba America Medical Systems, 2441 Michelle Dr, Tustin, CA 92780

386 b. HI VISION Ascendus, Hitachi, Tokyo, Japan

387 c. StataCorp. 2009. Stata Statistical Software: Release 11. College Station, TX:

388 StataCorp LP.

PT
389

RI
390 Figure legends

SC
391 Fig. 1 Dynamic US testing for subacromial impingement in neutral (A) and abducted

392 (B) positions (Acr, acromion; Del, deltoid; SS, supraspinatus; GT, great tuberosity).

U
393 Demonstration (C) and imaging (D) of US-guided subacromial injection (arrowhead,
AN
394 needle).

395
M

396 Fig. 2 Study flow chart: identification, screening and enrolment of participants for
D

397 study of initial effectiveness and recurrence after ultrasound (US)-guided

TE

398 subacromial injection.

399
EP

400 Fig. 3 Kaplan-Meier curve of recurrence of shoulder pain after a previously

C

401 performed ultrasound-guided subacromial injection based on the presence of

AC

402 sonographically-diagnosed subdeltoid bursitis (SD) (A) and a positive painful arc test

403 (PART) (B).

404

405
 ACCEPTED MANUSCRIPT

Table 1. Baseline characteristics, sonographic presentations and clinical measurements of patients with and without initial effectiveness to ultrasound-guided
subacromial corticosteroid injection

Initial effectiveness (+) Initial effectiveness (-) P value

(N=106) (N=58)

PT
Baseline characteristics
Age (years) 59.5±11.0 58.9±10.2 0.74

RI
Sex
Female (number, %) 65 (61.3%) 36 (62.0%)

SC
Male (number, %) 41 (38.6%) 22 (37.9%) 0.92
Affected side

U
Left 35 (33.0%) 30 (51.7%)

AN
Right 71 (66.9%) 28 (48.2%) 0.02*
Sonographic presentations (static images)

M
Biceps long head tendon tear (number, %) 2 (1.8%) 5 (8.6%) 0.04*
Bicipital peritendinous effusion (number, %) 48 (45.2%) 31 (53.4%) 0.32

D
Subscapularis tendon tear (number, %) 4 (3.7%) 5 (8.6%) 0.19

TE
Subscapularis tendinopathy (number, %) 16 (15.0%) 10 (17.2%) 0.72
Subscapularis tendon calcification (number, %) 29 (27.3%) 16 (27.5%) 0.98
EP
Subdeltoid bursitis (number, %) 18 (16.8%) 16 (27.5%) 0.10
Supraspinatus tendon full thickness tear (number, %) 14 (13.2%) 5 (8.6%) 0.38
C

Supraspinatus tendon bursal-sided partial thickness tear

2 (1.8%) 2 (3.4%) 0.53
AC

(number, %)
Supraspinatus tendon articular-sided partial thickness tear
13 (12.2%) 6 (10.3%) 0.71
(number, %)
Supraspinatus tendinopathy (number, %) 56 (51.8%) 32 (55.1%) 0.77
Supraspinatus tendon calcification (number, %) 23 (21.6%) 13 (22.4%) 0.91
Infraspinatus tendon tear (number, %) 3 (2.8%) 1 (1.7%) 0.66
Infraspinatus tendinopathy (number, %) 4 (3.7%) 1 (1.7%) 0.47
Infraspinatus tendon calcification (number, %) 8 (7.5%) 1 (1.7%) 0.12
 ACCEPTED MANUSCRIPT

Sonographic presentations (dynamic impingement test)

Grade 0, no impingement (number, %) 48 (45.2%) 21 (36.2%)
Grade 1, pain during shoulder elevation without soft tissue
23 (21.6%) 15 (25.8%)
impingement (number, %)
Grade 2, pain during shoulder elevation with soft tissue
26 (24.5%) 2 (3.4%)

PT
impingement (number, %)
Grade 3, pain during shoulder elevation with humeral head
9 (8.4%) 20 (34.4%) <0.01

RI
upward migration (number, %)
Physical findings before injection

SC
Pain duration (month) 7.2±14.6 4.9±6.7 0.26
Resting pain before injection (cm of VAS) 3.4±2.0 3.0±2.4 0.31

U
Night pain before injection(cm of VAS) 5.3±1.8 4.7±1.9 0.07

AN
Pain during overhead activities before injection (cm of VAS) 5.7±1.5 5.6±2.0 0.85
Bicipital groove tenderness (number, %) 69 (65.0%) 30 (51.7%) 0.09

M
Speed test (number, %) 51 (48.1%) 30 (86.2%) 0.65
Yergason’s test (number, %) 25 (23.5%) 14 (24.1%) 0.93

D
Empty can test (number, %) 67 (63.2%) 36 (62.0%) 0.88

TE
Neer test (number, %) 63 (59.4%) 37 (63.7%) 0.58
Hawkins-Kennedy test (number, %) 73 (68.8%) 42 (72.4%) 0.63
EP
Painful arc test (number, %) 69 (65.0%) 41 (70.6%) 0.46
Shoulder stiffness (number, %) 26 (24.5%) 41 (70.6%) <0.01*
C

Assessment of pain 2 weeks after injection

AC

Resting pain after injection (cm of VAS) 1.3±1.1 2.6±2.1 <0.01

Night pain after injection(cm of VAS) 1.7±1.2 4.0±1.7 <0.01
Pain during overhead activities after injection (cm of VAS) 1.8±1.1 4.8±2.0 <0.01
Abbreviation: VAS, visual analogue scale.
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
EP
C
AC
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
EP
C
AC
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
EP
C
AC