Eur J Orthop Surg Traumatol

DOI 10.1007/s00590-016-1839-y

ORIGINAL ARTICLE • SHOULDER - SPORT

Extracorporeal shock wave therapy, ultrasound-guided

percutaneous lavage, corticosteroid injection and combined
treatment for the treatment of rotator cuff calcific tendinopathy:
a network meta-analysis of RCTs
Alisara Arirachakaran1 • Manusuk Boonard2 • Sarunpong Yamaphai3 •
Akom Prommahachai4 • Suraphol Kesprayura1 • Jatupon Kongtharvonskul5

Received: 4 August 2016 / Accepted: 11 August 2016

 Springer-Verlag France 2016

Abstracts Treatment of calcific tendinitis using extracor- compared to placebo. Barbotage plus ESWT significantly
poreal shock wave therapy (ESWT), ultrasound-guided improved CMS, VAS and decreased size of calcium
percutaneous lavage (UGPL or barbotage), subacromial deposit when compared to ESWT, while barbotage plus
corticosteroid injection (SAI) and combined treatment is SAI significantly improved CMS and decreased size of
still controversial. This systematic review and meta-re- calcium deposit when compared to SAI. There have no
gression aimed to compare clinical outcomes between different adverse effects of all treatment groups. Multiple
treatments. Relevant RCTs were identified using PubMed active treatment comparisons indicated that barbotage plus
and Scopus search engines to date of September 23, 2015. SAI significantly improved VAS and size of calcium
Seven of 920 studies identified were eligible. Compared to deposit when compared to other groups, while barbotage
the other treatments, the results of this study indicate that plus SAI improved CMS when compared to other groups.
ESWT significantly improved CMS and VAS when But there was no significant difference. The network meta-
analysis suggested that combined US-guided needling and
subacromial corticosteroid injection significantly decreased
& Jatupon Kongtharvonskul shoulder pain VAS, improved CMS score and decreased
[email protected]
the size of calcium deposits, while also lowering risks of
Alisara Arirachakaran adverse event when compared to barbotage plus ESWT,
[email protected]
ESWT and subacromial corticosteroid injection; therefore,
Manusuk Boonard the evidence points to UGPL as being the treatment of
[email protected]
choice for nonsurgical options of treatment in calcific
Sarunpong Yamaphai tendinitis of the shoulder.
[email protected]
Level of evidence I.
Akom Prommahachai
[email protected]
Keywords Calcific tendinitis
ESWT
UGPL
CS

Suraphol Kesprayura Ultrasound guide
Extracorporeal shock wave therapy
[email protected]
1
Orthopedics Department, Police General Hospital, Bangkok,
Thailand Introduction
2
Orthopedics Department, Srinakarin Hospital, Khonkan,
Thailand Calcific tendinitis of the rotator cuff is one of the most
3
Orthopaedic Department, Saint Louis Hospital, Bangkok, common causes of shoulder pain, characterized by calcium
Thailand phosphate crystal deposition in the rotator cuff tendons and
4
Orthopedic Department, Aek Udon International Hospital, generally affecting people between the ages of 30 and
Udontani, Thailand 50 years [4, 41]. The prevalence of calcium deposits has
5
Section for Clinical Epidemiology and Biostatistics, Faculty been reported as 2–20 % in asymptomatic shoulders and is
of Medicine Ramathibodi Hospital, Bangkok, Thailand estimated to affect 7–17 % of the patients experiencing

123
 Eur J Orthop Surg Traumatol

shoulder pain [14, 15, 32, 33, 39]. It can occasionally using a network meta-analysis. The objective of this study
characterized by intractable pain and morbidity to the was to assess each method of minimally invasive treatment
patient. However, this disease is usually self-limited and of calcific tendinitis, with the aim of comparing relevant
can be treated by conservative treatment methods with clinical outcomes such as visual analog score (VAS),
good results [11, 34]. The most efficient treatment for this Constant–Murley Score (CMS), Disabilities of the Arm,
common disease is still debatable, and no standard treat- Shoulder and Hand (DASH), Western Ontario Rotator Cuff
ment has yet been established [20]. Numerous treatments index (WORC), University of California Los Angeles scale
have been advocated for calcific tendinitis. The initial (UCLA), full resorption rate and adverse effects (pain
treatment is usually conservative and consists of rest, during treatment, soreness, local subcutaneous hematomas
physical therapy and a medication regimen that includes and small petechial hemorrhages) by combining direct and
nonsteroidal anti-inflammatory drugs. In patients with indirect evidence in a network meta-analysis.
severe or persisting symptoms, more invasive therapy is
indicated. Numerous types of conservative treatments have
been described. These include calcium deposit needling Materials and methods
and lavage (barbotage), subacromial corticosteroid injec-
tions (SAI) and extracorporeal shock wave therapy Search strategy
(ESWT) [1, 2, 6, 13, 18, 21, 27, 30, 34, 36, 42]. Never-
theless, failure of these methods may necessitate the need A systematic review and network meta-analysis was
for surgical management [4, 19, 26, 35]. However, as there performed in accordance with Preferred Reporting Items
is a lack of high-level evidence studies comparing these for Systematic Reviews and Meta-Analyses (PRISMA)
modalities, the preferred treatment for calcific tendinitis guidelines [22]. The Medline and Scopus databases were
remains a subject of debate. Minimally invasive treatment used to identify relevant studies published in English
modalities such as SAI, ESWT and barbotage are fre- from the date of inception to September 23, 2015. The
quently performed [5, 8–10, 13, 17, 20, 31, 43] prior to PubMed and Scopus search engines were used to locate
surgical intervention, as both have been reported to be studies using the following search terms: [(calcific AND
efficient nonoperative treatment modalities and may even (tendinosis OR tendinopathy OR tendinitis) AND shoul-
eliminate the need for surgery. SAI is relatively easy to der) AND (ESWT OR ultrasound OR needles OR per-
perform, has a low risk of complications, is low cost and is cutaneous OR extracorporeal shock wave therapies OR
widely available [9]. ESWT has been recommended as a barbotage OR arthroscopes)]. Search strategies for
second-line therapy prior to surgical management [13], as it Medline and Scopus are described in detail in the
is not an invasive procedure and is relatively easy to per- appendix. References from the reference lists of included
form in the outpatient setting. Several studies trials and previous systematic reviews were also
[5, 8, 13, 17, 31] have demonstrated the effectiveness of explored.
ESWT in treating this condition when compared to pla-
cebo. However, this procedure can be extremely painful
Selection of studies
when the patient is in hyperalgesic crisis [27]. Barbotage
treatment is more invasive, requires more skills and
Identified studies were first selected based on titles and
equipment, is time-consuming and can be painful during
abstracts by two independent authors (J.K. and A.A.). Full
and after the procedure [10, 36]. Some studies report that
papers were retrieved if a decision could not be made from
combining barbotage with SAI [36] or ESWT [21] can
the abstracts. Disagreements were resolved by consensus
improve pain and function of shoulders with calcific ten-
and discussion with a third party (S.M.). Reasons for
dinitis when compared to treatment with barbotage alone.
ineligibility or exclusion of studies were recorded and
One systematic review and one meta-analysis recently
described.
published [23, 24] found that ESWT and barbotage have
been proved to be safe and effective treatments. Barbotage
Inclusion criteria
has not been proved to be more effective than a SAI [9].
However, this meta-analysis did not consider sources of
Randomized controlled trials, quasi-experimental designs
heterogeneity (age, sex, disease duration and size of cal-
studies that compared clinical outcomes between ESWT,
cium deposit), publication bias was not assessed, and other
SAI, barbotage with or without SAI or ESWT and placebo
RCTs [10, 20] have since been published. Moreover, none
were eligible if they met the following criteria:
of the reviews compared all the different minimally inva-
sive treatments of the shoulder with calcific tendinitis, and • Compared clinical outcomes between ESWT, SAI,
none compared the effectiveness of the different treatments barbotage with or without SAI and placebo.

123
Eur J Orthop Surg Traumatol

• Compared at least one of the following outcomes: Outcomes

visual analog score (VAS), Constant–Murley Score
(CMS), Disabilities of the Arm, Shoulder and Hand The outcomes of interest were Constant–Murley score
(DASH), Western Ontario Rotator Cuff index (CMS), pain VAS, size of calcium deposit and adverse
(WORC), University of California Los Angeles scale effect. Methods of measure for these outcomes were used
(UCLA), full resorption rate and adverse effects (pain according to the original studies. Briefly, this includes
during treatment, soreness, local subcutaneous hema- Constant–Murley score is a 100-point scale composed of a
tomas and small petechial hemorrhages). number of individual parameters [7]. These parameters
• Had sufficient data to extract and pool, i.e., the reported define the level of pain and the ability to carry out the
mean, standard deviation (SD), the number of subjects normal daily activities of the patient. The test is divided
according to treatments for continuous outcomes and into four subscales: pain (0–15 points), activities of daily
the number of patients according to treatment for living (0–20 points), strength (0–25 points) and range of
dichotomous outcomes. motion: forward elevation, external rotation, abduction and
internal rotation of the shoulder (0–40 points). The higher
When eligible papers had insufficient information, we
the score, the higher the quality of the function [16]. The
contacted authors by email (up to three consecutive emails
VAS pain scale (0–10), size of calcium deposit in mil-
if there was no reply) for additional information. If authors
limeter and adverse effect (local pain, infection, skin
did not provide additional data, the study was excluded
reaction and bleeding) was considered.
from the review.
Statistical analysis
Data extraction
Direct comparisons of continuous outcomes measured at
Two reviewers (J.K. and A.A.) independently performed
the end of each study between ESWT versus placebo,
data extraction using standardized data extraction forms.
ESWT versus barbotage, barbotage versus SAI and barb-
General characteristics of the study were age, sex, disease
otage versus placebo were pooled using an unstandardized
duration and size of calcium deposit. The number of sub-
mean difference (UMD). Heterogeneity of the mean dif-
jects, mean and SD of continuous outcomes [i.e., visual
ference across studies was checked using the Q-statistic
analog score (VAS), Constant–Murley Score (CMS), Dis-
and the degree of heterogeneity was quantified using the I2
abilities of the Arm, Shoulder and Hand (DASH), Western
statistic. If heterogeneity was present as determined by a
Ontario Rotator Cuff index (WORC), University of Cali-
statistically significant Q-statistic or by I2 [ 25 %, the
fornia Los Angeles scale (UCLA)] between the groups
UMD was estimated using a random-effects model;
were extracted. Cross-tabulated frequencies between
otherwise, a fixed-effects model was applied.
treatment and all dichotomous outcomes [full resorption
For dichotomous outcomes, a relative risk (RR) of full
rate and adverse effects (pain during treatment, soreness,
resorption and adverse effects of treatment comparisons at
local subcutaneous hematomas and small petechial hem-
the end of each study was estimated and pooled. Hetero-
orrhages)] were also extracted. Any disagreements were
geneity was assessed using the previous method. If
resolved by discussion and consensus with a third party
heterogeneity was present, the Dersimonian and Laird
(M.B.).
method [3] was applied for pooling. If not, the fixed-effects
model by inverse variance method was applied. Meta-re-
Risk of bias assessment
gression was applied to explore the source of heterogeneity
(e.g., mean age, percentage of females, disease duration
Two authors (J.K. and A.A.) independently assessed risk
and size of calcium deposit) if data were available. Publi-
of bias for each study following suggestions in the
cation bias was assessed using contour-enhanced funnel
PRISMA guidelines [22]. Six domains were assessed,
plots [28, 29] and Egger tests [12].
which included sequence generation, allocation conceal-
For indirect comparisons, network meta-analyses were
ment, blinding (participant, personnel and outcome
applied to assess all possible effects of treatment if sum-
assessors), incomplete outcome data, selective outcome
mary data were available for pooling [25, 37, 40]. A linear
reporting and other sources of bias. Disagreements
regression model, weighted by inverse variance, was
between two authors were resolved by consensus and
applied to assess the treatment effects for continuous out-
discussion with a third party (M.B.). Level of agreement
comes. For adverse effects, a mixed-effect Poisson
for each domain and the overall domains were assessed
regression was applied to assess treatment effects [25].
using the Kappa statistics.
Summary data were expanded to individual patient data

123
 Eur J Orthop Surg Traumatol

CMS, VAS, size of calcium, adverse effect

using the ‘‘expand’’ command in STATA. Treatment was

VAS visual analog score, KSS Knee Society Score, KFS knee function score, AKP anterior knee pain, PB patella problem, PF patella fracture, PI patella fracture, RP reoperation patella
considered as a fixed effect, whereas the study variable was

CMS, size of calcium, adverse effect

CMS, size of calcium, adverse effect

CMS, size of calcium, adverse effect
VAS, size of calcium, adverse effect

VAS, size of calcium, adverse effect

considered as a random effect in a mixed-effect model. The
pooled RR and its 95 % confidence intervals (CIs) were
estimated by exponential coefficients of treatments. All
analyses were performed using STATA version 14.0 [40].
A p value \0.05 was considered statistically significant,
except for the test of heterogeneity where \0.10 was used.

Outcome
Results

CMS
In total, 132 and 880 studies from Medline and Scopus

symptoms (months)
were identified, respectively; 92 studies were duplicates,

Mean duration of
leaving 895 studies for review of titles and abstracts. Of
these, 7 full papers were reviewed and extracted data.
Characteristics of the 7 studies are described in Table 1.

14.75
Among 6 ESWT studies, the comparators included placebo

33.4

13.5
11.7
42
in 4 studies, barbotage and ESWT in one study and barb-

–
otage and SAI in one study. One study compares combined

Mean size of calcium

barbotage and SAI with SAI. Most studies have a follow-
up period within 12 months. Only one study has a follow-
up time approximately 2 years. Mean age, size of calcium deposit (mm)
deposit and duration of symptom varies from 48.4 to
12.4

17.9

20.5
11.2
14.4

12.9
56 years, 11.2 to 20.5 mm and 11.7 to 33.4 months. Per-
–
centage of females ranged from 38.7 to 90.2 %. Various
Female (%)

outcomes were compared between treatment groups

(Fig. 1).
57.3
67.1
51.3

38.9
58.7

90.1
52
Risk of bias in included studies
48.4

55.4
52.1

55.7
Age

52

56
–

Risk of bias assessment is described in Table 2.

Barbotage and

Barbotage and
Comparator

Direct comparisons
ESWT
Placebo
Placebo

Placebo
Placebo

SAI
SAI

Data for direct comparisons of all treatments and outcomes

measured at the end of each study are described in Table 1.
Intervention

Pooling according to outcomes was performed if there

Barbotage
and SAI

were at least two studies for each comparison, as clearly

ESWT
ESWT
ESWT

ESWT
ESWT

ESWT

described below. There was no evidence of publication bias

by Egger’s test for both pooled effects of all outcomes from
Table 1 Characteristics of included studies
Follow-up

direct comparison.
(months)

4.1

23.2

Constant–Murley score
6

6
12

12
12
Type of

Among five studies, the UMD of 23.3 (95 % CI 9.8, 17.6)

study

RCT
RCT
RCT

RCT
RCT
RCT

RCT

showed that there was significantly higher CMS in ESWT

than placebo (Table 3). Heterogeneity was present
Gerdesmeyer [13]

(I2 = 92.4). ESWT was statistical significantly lower CMS

Cosentino [8]

of -9.5 (95 % CI -18.5, -0.5) when compared to com-

de Witte [9]
Krasny [21]

Cacchio [5]
References

Kim [20]

bined barbotage and ESWT. SAI was statistically signifi-

Hsu [17]

cantly lower CMS of -12.1 (95 % CI -20.6, -3.6) when

compared to combined barbotage and SAI.

123
Eur J Orthop Surg Traumatol

880 studies
132 studies
retrieved from
retrieved from
Scopus
Medline

895 left after removed

duplicates

888 studies were ineligible:

• 853 non-RCTs

• 22 other diseases

• 13 other interventions

7 studies left for full paper

4 studies:ESWT VS placebo
1 study : ESWT VS barbotage and SAI
1 study : ESWT VS barbotage and ESWT
1 study : SAI VS barbotage and SAI

Constant-Murley Score: 5 studies

Visual Analog Score: 5 studies
Size of calcium deposit: 4 studies
Adverse effect: 7 studies

Fig. 1 Flow of study selection

Table 2 Risk of bias assessment

References Adequate sequence Adequate allocation Blinding address incomplete Selective Free of Description of
generation concealment outcome data outcome report other bias other bias

Gerdesmeyer Y Y Y Y Y Y –
[13]
Cosentino [8] U N N Y Y Y –
Krasny [21] Y Y Y Y Y Y –
Cacchio [5] Y Y Y Y Y Y –
Hsu [17] Y N Y Y Y Y –
de Witte [9] Y Y Y Y Y Y –
Kim [20] Y Y Y N Y N Per protocol
analysis

Visual analog score Size of calcium deposit

Among five studies, the UMD of -4.4 (95 % CI -6.3, Among four studies, the UMD of -11.3 (95 % CI -24.7,
-2.3) showed that there was significantly lower VAS in 2.2) mm showed that there was lower the size of calcium
ESWT than placebo (Table 3). Heterogeneity was present deposit after ESWT when compared to placebo, but this
(I2 = 95.8). ESWT was statistically significantly higher was not significant (Table 3). Heterogeneity was present
VAS of 1.3 (95 % CI 0.9, 1.7) when compared to combined (I2 = 97.4). ESWT was statistically significant higher than
barbotage and ESWT. SAI was lower of -0.2 (95 % CI the size of calcium deposit of 5.2 (95 % CI 4.8, 5.5) mm
-1.5, 1.1) when compared to combined barbotage and SAI, when compared to combined barbotage and ESWT. SAI
but this was not significant (Table 3). was statistically significant higher than the size of calcium

123
 Eur J Orthop Surg Traumatol

Table 3 Summary of results of

Clinical outcomes No. studies I2 No. subjects UMD (95 % CI)
direct comparisons according to
type of interventions Constant Murray Scale
ESWT vs. placebo 3 92.4 116 vs. 96 23.3 (9.8, 17.6)*
ESWT vs. (barbotage and ESWT) 1 – 40 vs. 40 -9.5 (-18.5, -0.5)*
SAI vs. (barbotage and SAI) 1 – 25 vs. 23 -12.1 (-20.6, -3.6)*
Visual analog score
ESWT vs. placebo 3 95.8 126 vs. 106 -4.4 (-6.3, -2.3)*
ESWT vs. (barbotage and SAI) 1 – 29 vs. 25 1.3 (0.9, 1.7)*
SAI vs. (barbotage and SAI) 1 – 25 vs. 23 -0.2 (-1.5, 1.1)
Size of calcium deposit
ESWT vs. placebo 2 97.4 78 vs. 58 -11.3 (-24.7, 2.2)
ESWT vs. (barbotage and SAI) 1 – 29 vs. 25 5.2 (4.8, 5.5)*
SAI vs. (barbotage and SAI) 1 – 25 vs. 23 11.8 (6.3, 17.3)*
Adverse effect
ESWT vs. placebo 4 0 161 vs. 141 1.21 (0.91, 1.58)
ESWT vs. (barbotage and SAI) 1 – 29 vs. 25 0.9 (0.02, 42.16)
ESWT vs. (barbotage and ESWT) 1 – 40 vs. 40 1 (0.02, 49.20)
SAI vs. (barbotage and SAI) 1 – 25 vs. 23 0.2 (0.009, 3.65)
* Statistically significant difference (p \ 0.05)

deposit of 11.8 (95 % CI 6.3, 17.3) mm when compared to indicated that there was no significant difference between
combined barbotage and SAI. ESWT and SAI or combined groups (barbotage and SAI or
barbotage and ESWT).
Adverse effects
Visual analog score
Among 7 studies, the pooled RR of ESWT was homoge-
neous (I2 = 0) with a value of 1.21 (95 % CI 0.91, 1.58) The regression analysis from five studies suggested that the
which showed a significantly higher risk of adverse effects mean VAS of ESWT, combined barbotage and SAI and
when compared with placebo (Table 3). The RR of ESWT SAI were -2.91 (95 % CI -5.51, -0.30), -3.14 (95 % CI
showed value of 0.9 (95 % CI 0.02, 42.16) and 1 (95 % CI -5.64, -0.64), -2.38 (95 % CI -4.79, -0.03) showing
0.02, 49.20) showed insignificantly lower risk of adverse that the VAS was statistically significantly lower than
effects when compared with combined barbotage and SAI placebo (as seen in Table 4; Fig. 3). Multiple comparisons
and combined barbotage and ESWT. SAI was insignifi- indicated that combined barbotage and SAI resulted in a
cantly lower RR of 0.2 (95 % CI 0.009, 3.65) when com- value -0.76 (95 % CI -1.42, -0.11) significantly differ-
pared with Combined barbotage and SAI. ent lower than SAI.

Network meta-analysis Size of calcium deposit

Constant–Murley score The regression analysis from four studies suggested that the
mean size of calcium deposit of combined barbotage and
The regression analysis from five studies suggested that the SAI was -12.6 (95 % CI -24.8, -0.4) showing that the
mean CMS of ESWT was 22.36 (95 % CI 0.35. 44.37) size of calcium deposit was statistically significantly lower
showing that the CMS was statistically significantly higher than placebo, whereas ESWT was -9.6 (95 % CI -24.0,
than placebo, whereas combined barbotage and SAI, 4.8) lower and SAI was 0.8 (95 % CI -11.2, 12.9) higher
combined barbotage and ESWT and SAI were 23.12 (95 % than placebo, but these were not significantly different (as
CI -4.16. 50.40), 13.92 (95 % CI -13.36, 41.20) and seen in Table 4; Fig. 4). Multiple comparisons indicated
11.02 (95 % CI -16.26, 38.30) showing that the CMS was that combined barbotage and SAI and ESWT resulted in a
higher than placebo, but these were not significantly dif- value -13.4 (95 % CI -14.9, -11.9) and -10.5 (95 % CI
ferent (as seen in Table 4; Fig. 2). Multiple comparisons -14.1, -6.8) significantly different lower than SAI.

123
Eur J Orthop Surg Traumatol

Table 4 Comparisons of
Constants–Murray score
treatment effects: a network
meta-analysis N b p value 95 % CI

Treatment
ESWT vs placebo 252 22.36 0.048 0.35, 44.37*
(Barbotage and SAI) vs. placebo 119 23.12 0.078 -4.16, 50.40
(Barbotage and ESWT) vs. placebo 136 13.92 0.229 -13.36, 41.20
SAI vs. placebo 121 11.02 0.325 -16.26, 38.30
ESWT vs. (barbotage and SAI) 179 -0.76 0.873 -13.13, 11.62
ESWT vs. (barbotage and ESWT) 196 8.44 0.131 -3.93, 20.82
ESWT vs. SAI 181 11.34 0.064 -1.03, 23.72
Visual analog score
ESWT vs. placebo 261 -2.91 0.036 -5.51, -0.30*
(Barbotage and SAI) vs. placebo 154 -3.14 0.025 -5.64, -0.64*
SAI vs. placebo 131 -2.38 0.052 -4.79, -0.03*
ESWT vs. (barbotage and SAI) 203 0.24 0.738 -1.60, 2.07
ESWT vs. SAI 180 -0.52 0.346 -1.89, 0.84
(Barbotage and SAI) vs. SAI 73 -0.76 \0.032 -1.42, -0.11*
N B p value 95 % CI

Calcium deposit size

ESWT vs. placebo 286 -9.6 0.137 -24.0, 4.8
(Barbotage and SAI) vs. placebo 228 -12.6 0.046 -24.8, -0.4*
SAI vs. placebo 83 0.8 0.856 -11.2, 12.9
ESWT vs. (barbotage and SAI) 398 2.9 0.141 -1.5, 7.4
ESWT vs. SAI 253 -10.5 0.001 -14.1, -6.8*
(Barbotage and SAI) vs. SAI 195 -13.4 \0.001 -14.9, -11.9*
N RR p value 95 % CI

Adverse effect
ESWT vs. placebo 371 1.18 0.490 0.74, 1.87
(Barbotage and SAI) vs. placebo 189 0.78 0.807 0.11, 5.64
(Barbotage and ESWT) vs. placebo 181 0.75 0.826 0.06, 9.40
SAI vs. placebo 166 0.33 0.437 0.02, 5.39
ESWT vs. (barbotage and SAI) 278 1.50 0.680 0.22, 10.48
ESWT vs. (barbotage and ESWT) 270 1.56 0.725 0.13, 18.68
ESWT vs. SAI 255 3.55 0.369 0.22, 56.57
(Barbotage and SAI) vs. SAI 73 2.36 0.471 0.22, 24.39
(Barbotage and ESWT) vs. SAI 65 2.27 0.662 0.06, 90.39
(Barbotage and SAI) vs. (barbotage 88 1.03 0.981 0.05, 23.31
and EWST)
* Statistically significant difference (p \ 0.05)

Adverse effects Discussion

Data from seven studies revealed that compared to placebo, Systematic review and a network meta-analysis was per-
combined barbotage and SAI, barbotage and ESWT and SAI formed to comparing effect of ESWT, SAI, barbotage and
had 0.78 (95 % CI 0.11, 5.64), 0.75 (95 % CI 0.06, 9.40) and SAI, barbotage and ESWT and placebo for treatment
0.33 (95 % CI 0.02, 10.48) lower occurrences of adverse rotator cuff calcific tendinopathy. Relevant clinical out-
effects, whereas ESWT had an approximately 18 % comes included CMS, VAS, size of calcium deposit and
[RR = 1.18 (95 % CI 0.74, 1.87)] statistically insignificantly adverse effect were pooled. The results of this study indi-
higher risk than placebo groups (as seen in Table 4; Fig. 5). cate that ESWT significantly improved CMS and VAS

123
 Eur J Orthop Surg Traumatol

ESWT
n = 228
ESWT 9.5* Barbotage and ESWT
n = 156 n = 40
-11.3

-10.5 Placebo 5.2*

23.3* 13.92 n = 58
Placebo
11.34 -9.2 0.8 12.6
n = 96
23.12
11.02

SAI Barbotage and SAI

11.8*
12.1* n = 25 n = 170
SAI Barbotage and SAI
n = 25 n = 23
Fig. 4 Network meta-analysis of treatment effects on size of calcium
deposit. A line in the figure represents treatment comparisons, with
Fig. 2 Network meta-analysis of treatment effects on CMS. A line in
arrows and tails referring to intervention and comparators, respec-
the figure represents treatment comparisons, with arrows and tails
tively. Bold and dashed lines refer to direct and indirect comparisons,
referring to intervention and comparators, respectively. Bold and
respectively. The number at the line indicates chance of treatment
dashed lines refer to direct and indirect comparisons, respectively.
responsiveness, in which \0 indicates favors intervention vs the
The number at the line indicates chance of treatment responsiveness,
comparator. *p \ 0.05 with Bonferroni correction
in which \0 indicates favors intervention versus the comparator.
*p \ 0.05 with Bonferroni correction
improved CMS when compared to other groups. But there
was no significant difference.
ESWT
n = 155
The results of this study were consistent with previous
systematic review and meta-analyses [23, 24] that high-
-4.4*

energy ESWT and US-guided needling are the treatment

1.3* option in the short term to midterm and have proved to be a
-0.52
Placebo safe and effective treatment. We, however, have added
n = 106
more evidences that combined US-guided needling and
subacromial corticosteroid injection is safe and has been
-2.38* -3.14*
proved to be more effective (CMS, VAS and size of cal-
SAI 0.2* Barbotage and SAI
n = 25 n = 48 cium deposit) than other groups.
This study has a number of strengths. We have applied a
network meta-analysis to increase the power of the tests
Fig. 3 Network meta-analysis of treatment effects on VAS. A line in
the figure represents treatment comparisons, with arrows and tails and reduce type I errors [25, 37, 38]. We applied a
referring to intervention and comparators, respectively. Bold and regression model taking into account study effects to assess
dashed lines refer to direct and indirect comparisons, respectively. treatment effects. The network meta-analysis ‘‘borrows’’
The number at the line indicates chance of treatment responsiveness,
treatment information from other studies and increases the
in which \0 indicates favors intervention vs the comparator.
*p \ 0.05 with Bonferroni correction total sample size. As a result, treatment effects that could
not be detected in direct meta-analysis could be identified.
All possible treatment comparisons are mapped and dis-
when compared to placebo. Barbotage plus ESWT signif- played (see Figs. 2, 3, 4, 5). Although our pooled estimates
icantly improved CMS, VAS and decreased size of calcium were heterogeneous, the regression model with cluster
deposit when compared to ESWT, while barbotage plus effect takes into account variations at the study level. None
SAI significantly improved CMS and decreased size of of the RCTs compared ESWT, SAI, barbotage and SAI,
calcium deposit when compared to SAI. There have no barbotage and ESWT and placebo in the treatment of
different adverse effects of all treatment groups. Perform- rotator cuff calcific tendinopathy.
ing a direct meta-analysis is limited by the small number of Although all studies were RCTs, 85.7 % of the studies
studies that evaluated each particular pair of treatments; but were unclear in the randomization sequence generations
a network meta-analysis circumvents this problem by cre- and allocation concealment; hence, selection bias or con-
ating indirect comparisons between active treatments that founding factors may be present. Some pooled results were
can identify the most effective therapy. Multiple active heterogeneous, but we were unable to explore the source of
treatment comparisons indicated that barbotage plus SAI heterogeneity due to limitations of the reported data.
significantly improved VAS and size of calcium deposit In conclusion, the network meta-analysis suggested that
when compared to other groups, while barbotage plus SAI combined US-guided needling and subacromial

123
Eur J Orthop Surg Traumatol

ESWT Barbotage and ESWT

1.21
n = 230 n 40

1.2
1
0.75
Placebo
3.55 n = 141 1.03

0.78
0.33 0.9 2.27

SAI Barbotage and SAI

0.2
n = 25 n = 23

Fig. 5 Network meta-analysis of treatment effects on adverse effect. respectively. The number at the line indicates chance of treatment
A line in the figure represents treatment comparisons, with arrows responsiveness, in which \1 indicates favors intervention vs the
and tails referring to intervention and comparators, respectively. Bold comparator. *p \ 0.05 with Bonferroni correction
and dashed lines refer to direct and indirect comparisons,

corticosteroid injection significantly decreased shoulder Effectiveness of radial shock-wave therapy for calcific tendinitis
pain VAS, improved CMS score and decreased the size of of the shoulder: single-blind, randomized clinical study. Phys
Ther 86(5):672–682
calcium deposits, while also lowering risks of adverse 6. Comfort TH, Arafiles RP (1978) Barbotage of the shoulder with
event when compared to barbotage plus ESWT, ESWT image-intensified fluoroscopic control of needle placement for
and subacromial corticosteroid injection; therefore, the calcific tendinitis. Clin Orthop Relat Res 135:171–178
evidence points to UGPL as being the treatment of choice 7. Conboy VB, Morris RW, Kiss J, Carr AJ (1996) An evaluation of
the Constant–Murley shoulder assessment. J Bone Joint Surg Br
for nonsurgical options of treatment in calcific tendinitis 78(2):229–232
of the shoulder. 8. Cosentino R, De Stefano R, Selvi E, Frati E, Manca S, Frediani
B, Marcolongo R (2003) Extracorporeal shock wave therapy for
Acknowledgments All authors declare no funding source or sponsor chronic calcific tendinitis of the shoulder: single blind study. Ann
involvement in the study design, collection, analysis and interpreta- Rheum Dis 62(3):248–250
tion of the data, in writing the manuscript and in submission of the 9. de Witte PB, Selten JW, Navas A, Nagels J, Visser CP, Nelissen
manuscript for publication. RG, Reijnierse M (2013) Calcific tendinitis of the rotator cuff: a
randomized controlled trial of ultrasound-guided needling and
Compliance with ethical standards lavage versus subacromial corticosteroids. Am J Sports Med
41(7):1665–1673. doi:10.1177/0363546513487066
Conflict of interest All authors declare that they have no conflicts of 10. Del Castillo-Gonzalez F, Ramos-Alvarez JJ, Rodriguez-Fabian
interests. G, Gonzalez-Perez J, Jimenez-Herranz E, Varela E (2016)
Extracorporeal shockwaves versus ultrasound-guided percuta-
Ethical standards This article does not contain any studies with neous lavage for the treatment of rotator cuff calcific
human participants performed by any of the authors. tendinopathy: a randomised controlled trial. Eur J Phys Rehabil
Med 52(2):145–151
11. Depalma AF, Kruper JS (1961) Long-term study of shoulder
joints afflicted with and treated for calcific tendinitis. Clin Orthop
References 20:61–72
12. Egger M, Davey Smith G, Schneider M, Minder C (1997) Bias in
1. Aina R, Cardinal E, Bureau NJ, Aubin B, Brassard P (2001) meta-analysis detected by a simple, graphical test. BMJ
Calcific shoulder tendinitis: treatment with modified US-guided 315(7109):629–634
fine-needle technique. Radiology 221(2):455–461. doi:10.1148/ 13. Gerdesmeyer L, Wagenpfeil S, Haake M, Maier M, Loew M,
radiol.2212000830 Wortler K, Lampe R, Seil R, Handle G, Gassel S, Rompe JD (2003)
2. Albert JD, Meadeb J, Guggenbuhl P, Marin F, Benkalfate T, Extracorporeal shock wave therapy for the treatment of chronic
Thomazeau H, Chales G (2007) High-energy extracorporeal calcifying tendonitis of the rotator cuff: a randomized controlled
shock-wave therapy for calcifying tendinitis of the rotator cuff: a trial. JAMA 290(19):2573–2580. doi:10.1001/jama.290.19.2573
randomised trial. J Bone Joint Surg Br 89(3):335–341. doi:10. 14. Harniman E, Carette S, Kennedy C, Beaton D (2004) Extracor-
1302/0301-620x.89b3.18249 poreal shock wave therapy for calcific and noncalcific tendonitis
3. Altman DG, Bland JM (2003) Interaction revisited: the difference of the rotator cuff: a systematic review. J Hand Ther
between two estimates. BMJ 326(7382):219 17(2):132–151. doi:10.1197/j.jht.2004.02.003
4. Ark JW, Flock TJ, Flatow EL, Bigliani LU (1992) Arthroscopic 15. Harvie P, Pollard TC, Carr AJ (2007) Calcific tendinitis: natural
treatment of calcific tendinitis of the shoulder. Arthroscopy history and association with endocrine disorders. J Shoulder
8(2):183–188 Elbow Surg 16(2):169–173. doi:10.1016/j.jse.2006.06.007
5. Cacchio A, Paoloni M, Barile A, Don R, de Paulis F, Calvisi V, 16. Hirschmann MT, Wind B, Amsler F, Gross T (2010) Reliability
Ranavolo A, Frascarelli M, Santilli V, Spacca G (2006) of shoulder abduction strength measure for the Constant–Murley

123
 Eur J Orthop Surg Traumatol

score. Clin Orthop Relat Res 468(6):1565–1571. doi:10.1007/ publication bias from other causes of asymmetry. J Clin Epi-
s11999-009-1007-3 demiol 61(10):991–996
17. Hsu CJ, Wang DY, Tseng KF, Fong YC, Hsu HC, Jim YF (2008) 30. Re LP Jr, Karzel RP (1993) Management of rotator cuff calcifi-
Extracorporeal shock wave therapy for calcifying tendinitis of the cations. Orthop Clin N Am 24(1):125–132
shoulder. J Shoulder Elbow Surg 17(1):55–59. doi:10.1016/j.jse. 31. Rebuzzi E, Coletti N, Schiavetti S, Giusto F (2008) Arthroscopy
2007.03.023 surgery versus shock wave therapy for chronic calcifying ten-
18. Hurt G, Baker CL Jr (2003) Calcific tendinitis of the shoulder. dinitis of the shoulder. J Orthop Traumatol 9(4):179–185. doi:10.
Orthop Clin N Am 34(4):567–575 1007/s10195-008-0024-4
19. Jacobs R, Debeer P (2006) Calcifying tendinitis of the rotator 32. Rompe JD, Burger R, Hopf C, Eysel P (1998) Shoulder function
cuff: functional outcome after arthroscopic treatment. Acta after extracorporal shock wave therapy for calcific tendinitis.
Orthop Belg 72(3):276–281 J Shoulder Elbow Surg 7(5):505–509
20. Kim YS, Lee HJ, Kim YV, Kong CG (2014) Which method is 33. Rompe JD, Zoellner J, Nafe B (2001) Shock wave therapy versus
more effective in treatment of calcific tendinitis in the shoulder? conventional surgery in the treatment of calcifying tendinitis of
Prospective randomized comparison between ultrasound-guided the shoulder. Clin Orthop Relat Res 387:72–82
needling and extracorporeal shock wave therapy. J Shoulder 34. Rowe CR (1988) Injection technique for the shoulder and elbow.
Elbow Surg 23(11):1640–1646. doi:10.1016/j.jse.2014.06.036 Orthop Clin N Am 19(4):773–777
21. Krasny C, Enenkel M, Aigner N, Wlk M, Landsiedl F (2005) 35. Seil R, Litzenburger H, Kohn D, Rupp S (2006) Arthroscopic
Ultrasound-guided needling combined with shock-wave therapy treatment of chronically painful calcifying tendinitis of the
for the treatment of calcifying tendonitis of the shoulder. J Bone supraspinatus tendon. Arthroscopy 22(5):521–527. doi:10.1016/j.
Joint Surg Br 87(4):501–507. doi:10.1302/0301-620x.87b4.15769 arthro.2006.01.012
22. Liberati A, Altman DG, Tetzlaff J, Mulrow C, Gotzsche PC, 36. Serafini G, Sconfienza LM, Lacelli F, Silvestri E, Aliprandi A,
Ioannidis JP, Clarke M, Devereaux PJ, Kleijnen J, Moher D Sardanelli F (2009) Rotator cuff calcific tendonitis: short-term
(2009) The PRISMA statement for reporting systematic reviews and 10-year outcomes after two-needle us-guided percutaneous
and meta-analyses of studies that evaluate health care interven- treatment–nonrandomized controlled trial. Radiology
tions: explanation and elaboration. PLoS Med 6(7):e1000100. 252(1):157–164. doi:10.1148/radiol.2521081816
doi:10.1001371/journal.pmed.1000100 37. Song F, Altman DG, Glenny AM, Deeks JJ (2003) Validity of
23. Louwerens JK, Sierevelt IN, van Noort A, van den Bekerom MP indirect comparison for estimating efficacy of competing inter-
(2014) Evidence for minimally invasive therapies in the man- ventions: empirical evidence from published meta-analyses. BMJ
agement of chronic calcific tendinopathy of the rotator cuff: a 326(7387):472. doi:10.1136/bmj.326.7387.472
systematic review and meta-analysis. J Shoulder Elbow Surg 38. Song F, Harvey I, Lilford R (2008) Adjusted indirect comparison
23(8):1240–1249. doi:10.1016/j.jse.2014.02.002 may be less biased than direct comparison for evaluating new
24. Louwerens JK, Veltman ES, van Noort A, van den Bekerom MP pharmaceutical interventions. J Clin Epidemiol 61(5):455–463.
(2015) The effectiveness of high-energy extracorporeal shock- doi:10.1016/j.jclinepi.2007.06.006
wave therapy versus ultrasound-guided needling versus arthro- 39. Speed CA, Hazleman BL (1999) Calcific tendinitis of the
scopic surgery in the management of chronic calcific rotator cuff shoulder. N Engl J Med 340(20):1582–1584. doi:10.1056/
tendinopathy: a systematic review. Arthroscopy. doi:10.1016/j. nejm199905203402011
arthro.2015.06.049 40. StataCorp (2014) Stata 14 base reference manual. Stata Press,
25. Lu G, Ades AE (2004) Combination of direct and indirect evi- College Station
dence in mixed treatment comparisons. Stat Med 41. Wainner RS, Hasz M (1998) Management of acute calcific ten-
23(20):3105–3124. doi:10.1002/sim.1875 dinitis of the shoulder. J Orthop Sports Phys Ther 27(3):231–237.
26. Mole D, Kempf JF, Gleyze P, Rio B, Bonnomet F, Walch G doi:10.2519/jospt.1998.27.3.231
(1993) [Results of endoscopic treatment of non-broken tendino- 42. Wang CJ, Yang KD, Wang FS, Chen HH, Wang JW (2003)
pathies of the rotator cuff. 2. Calcifications of the rotator cuff]. Shock wave therapy for calcific tendinitis of the shoulder: a
Rev Chir Orthop Reparatrice Appar Mot 79(7):532–541 prospective clinical study with two-year follow-up. Am J Sports
27. Mouzopoulos G, Stamatakos M, Mouzopoulos D, Tzurbakis M Med 31(3):425–430
(2007) Extracorporeal shock wave treatment for shoulder calcific 43. Yoo JC, Koh KH, Park WH, Park JC, Kim SM, Yoon YC (2010)
tendonitis: a systematic review. Skeletal Radiol 36(9):803–811. The outcome of ultrasound-guided needle decompression and
doi:10.1007/s00256-007-0297-3 steroid injection in calcific tendinitis. J Shoulder Elbow Surg
28. Palmer TMPJ, Sutton AJ, Moreno SG (2008) Contour-enhanced 19(4):596–600. doi:10.1016/j.jse.2009.09.002
funnel plots in meta-analysis. STATA J 8(2):242–254
29. Peters JL, Sutton AJ, Jones DR, Abrams KR, Rushton L (2008)
Contour-enhanced meta-analysis funnel plots help distinguish

123