Bäck et al.

BMC Musculoskelet Disord (2021) 22:889

https://doi.org/10.1186/s12891-021-04768-7

RESEARCH Open Access

Return to work after subacromial

decompression, diagnostic arthroscopy,
or exercise therapy for shoulder impingement:
a randomised, placebo-surgery controlled
FIMPACT clinical trial with five-year follow-up
Mathias Bäck1, Mika Paavola2, Pasi Aronen3, Teppo L. N. Järvinen1 and Simo Taimela1*for the Finnish Shoulder
Impingement Arthroscopy Controlled Trial (FIMPACT) Investigators

Abstract
Background: Arthroscopic subacromial decompression is one of the most commonly performed shoulder surgeries
in the world. It is performed to treat patients with suspected shoulder impingement syndrome, i.e., subacromial pain
syndrome. Only few studies have specifically assessed return-to-work rates after subacromial decompression surgery.
All existing evidence comes from open, unblinded study designs and this lack of blinding introduces the potential for
bias. We assessed return to work and its predictors in patients with shoulder impingement syndrome in a secondary
analysis of a placebo-surgery controlled trial.
Methods: One hundred eighty-four patients in a randomised trial had undergone arthroscopic subacromial decom-
pression (n = 57), diagnostic arthroscopy, a placebo surgical intervention, (n = 59), or exercise therapy (n = 68). We
assessed return to work, defined as having returned to work for at least two follow-up visits by the primary 24-month
time point, work status at 24 and 60 months, and trajectories of return to work per follow-up time point. Patients and
outcome assessors were blinded to the assignment regarding the arthroscopic subacromial decompression vs. diag-
nostic arthroscopy comparison. We assessed the treatment effect on the full analysis set as the difference between
the groups in return-to-work rates and work status at 24 months and at 60 months using Chi-square test and the
predictors of return to work with logistic regression analysis.
Results: There was no difference in the trajectories of return to work between the study groups. By 24 months, 50 of
57 patients (88%) had returned to work in the arthroscopic subacromial decompression group, while the respective
figures were 52 of 59 (88%) in the diagnostic arthroscopy group and 61 of 68 (90%) in the exercise therapy group. No
clinically relevant predictors of return to work were found. The proportion of patients at work was 80% (147/184) at
24 months and 73% (124/184) at 60 months, with no difference between the treatment groups (p-values 0.842 and
0.943, respectively).

*Correspondence: simo.taimela@helsinki.fi
1
Finnish Centre for Evidence‑Based Orthopaedics (FICEBO), Department
of Orthopaedics and Traumatology, University of Helsinki and Helsinki
University Hospital, Töölö Hospital, P.O. Box 266, 00029 HUS, Helsinki,
Finland
Full list of author information is available at the end of the article

© The Author(s) 2021. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which
permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the
original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or
other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line
to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory
regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this
licence, visit http://​creat​iveco​mmons.​org/​licen​ses/​by/4.​0/. The Creative Commons Public Domain Dedication waiver (http://​creat​iveco​
mmons.​org/​publi​cdoma​in/​zero/1.​0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.
Bäck et al. BMC Musculoskelet Disord (2021) 22:889 Page 2 of 10

Conclusions: Arthroscopic subacromial decompression provided no benefit over diagnostic arthroscopy or exer-
cise therapy on return to work in patients with shoulder impingement syndrome. We did not find clinically relevant
predictors of return to work either.
Trial registration: ClinicalTrials.gov identifier NCT00​428870.
Keywords: Subacromial pain, Randomised controlled trial, Subacromial decompression, Placebo surgery, Exercise
therapy, Return to work

Background Methods
Subacromial pain, shoulder impingement syndrome Trial design and setting
and rotator cuff disease describe the most common We conducted this superiority trial at three orthopaedic
shoulder conditions. The syndrome is distinctively clinics (Hatanpää hospital, Tampere, Pirkanmaa Hos-
characterised by pain located subacromially while lift- pital District; Jorvi and Herttoniemi hospitals, Helsinki,
ing the arm and has traditionally been attributed to Helsinki University Hospital District) in Finland from
‘impingement’ of the rotator cuff tendons between the February 1, 2005 through October 10, 2018. The study
humeral head and the overlying acromion. Following was registered at ClinicalTrials.gov (NCT00428870, on
to this logic, a surgical procedure called subacromial 30/01/2007). Details of the trial design and conduct have
decompression was developed. It is a surgical proce- been published elsewhere [5, 10]. The patients and those
dure premised on smoothening the undersurface of who collected the data regarding the ASD vs. DA com-
the acromion, which decompresses the rotator cuff parison were blinded to the study-group assignments.
tendons passing through the subacromial space. With The protocol was approved by the Institutional Review
the advent of arthroscopy, arthroscopic subacromial Board of the Pirkanmaa Hospital District (R04200).
decompression (ASD) became a safe, quick and con- The study was conducted in accordance with the Dec-
venient procedure, and thus became one of the most laration of Helsinki. All patients gave written informed
frequently performed orthopaedic surgical procedures consent. On entering the study, they were unequivocally
in the world [1]. informed that they might undergo diagnostic arthroscopy
A Cochrane review [2] and a BMJ Rapid Recommen- and that they would be allowed to consider crossing over
dation [3], two recently published systematic analyses to ASD if they did not have adequate relief of symptoms,
of the evidence [4, 5], summarized that ASD offers no preferably no sooner than 6 months after randomisation.
clinically relevant benefits on shoulder symptoms and
function over nonoperative treatment or placebo sur- Participants
gery. However, some advocates of the surgery and We enrolled patients 35 to 65 years old who had subacro-
some authorities (funders, insurance companies) still mial pain (for more than 3 months) that was unrespon-
defend the appropriateness of the procedure based on sive to conventional conservative treatment and had
its perceived ability to enable people to return to work clinical findings consistent with shoulder impingement
more quickly. Evidence from previous trials on ASD syndrome. Magnetic resonance imaging with intra-artic-
(vs. non-operative treatment, usually exercise therapy) ular contrast was performed on all patients to exclude a
have reported rates of return to work (RTW) that sup- rotator cuff tear. Detailed inclusion and exclusion criteria
port such contention [6, 7], and other trials do not are provided in Table 1.
support it [5, 8, 9]. Given the unblinded, open design
of all these previous studies, the data may be prone to Randomisation and blinding
bias. Accordingly, we conducted a secondary analysis of In an attempt to obtain three balanced study groups
our data from our recently published placebo-surgery of similar group size, we planned a two-fold, sequen-
controlled FIMPACT trial [5, 10, 11], where we com- tial randomization as follows: First, during the base-
pared the rates of RTW in patients who had under- line appointment, patients were randomized to surgical
gone arthroscopic subacromial decompression (ASD), or conservative treatment (Exercise therapy, ET) in 2:1
diagnostic arthroscopy (DA, a placebo surgical inter- ratio. Those randomised to ET started standardised
vention), and exercise therapy (ET, a non-operative physiotherapy within 2 weeks of the baseline appoint-
alternative) among patients with shoulder impingement ment, while those allocated to surgery were scheduled
syndrome. We also assessed whether age, gender, base- for surgery aimed to be performed within 12 weeks of
line work status, and strain at work were potential pre- this first randomisation. In those allocated to surgery, a
dictors of RTW. diagnostic arthroscopy was first carried out to rule out a
Bäck et al. BMC Musculoskelet Disord (2021) 22:889 Page 3 of 10

Table 1 Inclusion and exclusion criteria

Inclusion criteria
1. Adult men or woman ages 35 to 65 years.
2. Subacromial pain for greater than 3 months with no relief from non-operative means (physiotherapy, non-steroidal anti-inflammatory medica-
tion, corticosteroid injections, and rest).
3. Pain provoked by abduction and positive painful arc -sign.
4. Positive impingement test (temporary relief of pain by subacromial injection of lidocaine).
5. Pain in at least 2 out of 3 isometric tests (abduction 0° and 30° or external rotation).
6. Provision of informed consent from the patient.
7. Ability to speak, understand and read in the language of the clinical site.
Exclusion criteria
1. Full thickness tear of the rotator cuff tendons diagnosed on clinical examination (marked weakness in any of the examined muscles) or magnetic
resonance imaging with intra-articular contrast (MRA).
2. Osteoarthritis of the glenohumeral and/or acromioclavicular joint diagnosed on clinical examination or on x-rays.
3. Substantial calcific deposits in the rotator cuff tendons found in the preoperative imaging.
4. Previous surgical procedure on the affected shoulder.
5. Evidence of the shoulder instability (positive apprehension/positive sulcus sign).
6. Symptomatic cervical spine pathology.
7. History of alcoholism, drug abuse, psychological or psychiatric problems that are likely to invalidate informed consent.
8. Patient declined to participate.

rotator cuff tendon tear and other obvious intra-articular Diagnostic arthroscopy (DA)
pathology requiring surgical treatment. If a rotator cuff Arthroscopic examination of the glenohumeral joint and
tear large enough to require repair according to current subacromial space was performed with the patient under
clinical practice guidelines [12] was found, patients were general anaesthesia. We performed an intra-articular
excluded and the tear was repaired. Patients with a par- and subacromial assessment of the rotator cuff integ-
tial tear not requiring repair were included in the study. If rity. If arthroscopic examination revealed any pathology
the eligibility of the patient was confirmed in diagnostic requiring intervention other than ASD, the patient was
arthroscopy, the surgeon asked a research nurse to carry excluded from the trial (Fig. 1). Once the eligibility was
out the second randomisation by opening an envelope confirmed, the participants were randomly assigned to
containing the study-group assignment (Arthroscopic either ASD or DA. For those allocated to the DA group,
subacromial decompression, ASD or Diagnostic arthros- the operation was terminated. To ensure concealment
copy, DA; ratio 1:1). Only the orthopaedic surgeon and of the participants and the staff other than those in the
other staff in the operating room were made aware of the operating theatre, the DA participants were kept in the
surgical group assignment, and they did not participate operating theatre for the time required to perform subac-
in further treatment or follow-up of the patient. romial decompression [5].
Randomisation was carried out using sequentially
numbered sealed opaque envelopes. Separate randomi-
sation lists for each three centres, with blocks varying Arthroscopic subacromial decompression (ASD)
randomly in size, were prepared by a statistician with After arthroscopic examination of the shoulder (i.e., diag-
no involvement in the clinical care of participants in the nostic arthroscopy), the ASD procedure that involved the
trial. debridement of the entire subacromial bursa (bursec-
tomy) and resection of the bony spurs and the projecting
anterolateral undersurface of the acromion, was carried
Study interventions out with a shaver, burr and/or electrocoagulation [14].
Exercise therapy (ET)
Supervised, progressive, individually-designed physi-
otherapy was started within 2 weeks of randomisation Postoperative care
using a standardised protocol that relied primarily on In the ASD and DA groups, the postoperative rehabilita-
daily home exercises as well as 15 visits to an independ- tion was identical, consisting of one visit to an independ-
ent physiotherapist [13]. The details of the ET protocol ent physiotherapist, blind to the group assignment, for
are available elsewhere [10]. guidance and instructions for home exercises.
Bäck et al. BMC Musculoskelet Disord (2021) 22:889 Page 4 of 10

Fig. 1 Flowchart of the Full Analysis Set

Outcome measures advantage of ASD over DA, under the assumption of a

Questionnaires were administered at baseline and 3, 6, two-sided type 1 error rate of 5%, we planned to recruit
12, 24, and 60 months after randomisation. Work status 70 patients per group.
was inquired with the following answering four options: The present study was primarily designed to ascertain
1) At work; 2) On sickness absence; 3) Unemployed; 4) whether ASD is superior to DA in RTW with the two
Retired. Work status at each time point was dichot- outcomes (the primary comparison). We also included a
omised (At work / Not at work). The dichotomous RTW pragmatic comparison of the relative benefits of ASD vs.
(Yes/No) was defined as follows: the study participant ET (the secondary comparison).
was at work at the primary 24-month time point or was We primarily quantified the treatment effect on full
at work at least at two of the three earlier time points. We analysis set basis as the difference between the groups
also calculated the trajectories of return to work using in RTW rates up to 24 months and in work status at 24
dichotomous data (At work/Not at work) per each fol- and 60 months after the primary randomisation. In these
low-up time point. analyses, the participants were included as randomized
and we analysed all available data, the full analysis set. As
Statistical analysis and sample size calculation a sensitivity analysis, we performed on-treatment analy-
We originally powered the study to detect a difference sis of unblinding, in which the participants were included
of at least the minimal clinically important difference as treated in case the participants in the DA and ET
(15 points [15]) in the two primary outcomes, pain vis- groups had received ASD after unblinding.
ual analogue scale (0–100 with 0 denoting no pain and Categorical variables were analysed using Chi-square
100 denoting extreme pain) at rest and with arm activ- test and logistic regression analysis. We compared the
ity between the ASD and DA groups. For the study to rates of RTW between the study groups with Chi-square
have 90% power to show a minimal clinically important test. In the logistic regression model RTW was the
Bäck et al. BMC Musculoskelet Disord (2021) 22:889 Page 5 of 10

independent variable and dependent variables were base- Three subjects had received a pension at the time of
line work status (Yes/No), age (dichotomised by median baseline assessment, one in each study group. Over the
50.6 years), gender (Male/Female), physical work strain course of the 24-month follow-up, two patients in the ET
(dichotomised) and study group (three): Odds ratios with group and two patients in the DA group withdrew from
95% confidence intervals and P-values were calculated. the study and one patient in the DA group was deceased,
We estimated the prediction model accuracy by calculat- while one patient had retired before the 3-month follow-
ing a confusion matrix (actual vs. predicted) and calcu- up time point in the ASD group. They all were excluded
lated the positive and negative predictive values thereof. from the RTW analyses. Altogether, we had RTW data
All analyses are secondary by nature and must be inter- available of 184 patients (95%) during the 24-month fol-
preted as exploratory and supportive only. A p value of low-up and of 170 patients (88%) at 60 months out of the
.05 was considered to indicate statistical significance. The 193 randomised (Fig. 1). Another 14 patients were lost to
estimation of the logistic regression model parameters follow-up between the 24 and 60 time points.
was conducted using R 3.5.2 statistical software.
Return to work outcomes
Results At the primary 24-month time point, 147 patients (80%)
Characteristics of the patients were at work, 23 (13%) on sickness absence, 10 (5%) had
There were 281 eligible patients, of whom 71 were retired, two (1%) were unemployed or students, and data
excluded (Fig. 1). Two hundred ten patients underwent of two patients (1%) was missing. The respective figures
the 1st randomisation; 71 were assigned to ET and 139 to at 60 months were 124 (67%), 26 (14%), 2 (1%), and 14
surgery. Of the 139 patients allocated to surgery, another (8%). Work status at the 24-month and 60-month time
17 were excluded before the 2nd randomisation (Fig. 1), points by group on the full analysis set basis are shown in
leaving 59 patients to receive ASD and 63 to receive DA Table 3. In short, there were no differences between the
[5]. treatment groups in the proportion of patients at work
All baseline characteristics were well balanced among at 24 months (Chi-square p = 0.84) or at 60 months (Chi-
the three study groups (Table 2). square p = 0.94).

Table 2 Patient characteristics at baseline

Characteristics Arthroscopic subacromial Diagnostic Exercise
decompression (n = 59) arthroscopy (n = 63) therapy (n = 71)

Age, years 50.5 (7.3) 50.8 (7.6) 50.4 (6.6)

Female, n (%) 42 (71) 46 (73) 47 (66)
Dominant hand affected, n (%) 35 (59) 36 (57) 46 (65)
Duration of symptoms, months 18 (14) 18 (19) 22 (23)
Able to work normally regardless of shoulder symptoms, 27 (46) 31 (49) 35 (49)
n (%)
Visual analogue scale score, at ­resta 41.3 (25.8) 41.6 (25.5) 41.7 (27.5)
Visual analogue scale score, on arm ­activitya 71.2 (23.6) 72.3 (21.7) 72.4 (20.8)
Constant-Murley ­scoreb 32.2 (15.8) 31.7 (14.0) 35.2 (16.2)
Simple shoulder test ­scorec 4.9 (2.9) 4.9 (2.9) 4.8 (2.7)
15D ­scored 0.89 (0.06) 0.89 (0.07) 0.88 (0.08)
SF-36 score physical ­healthe 74.3 (12.5) 74.1 (13.1) 75.7 (10.1)
SF-36 score mental ­healthe 79.4 (14.2) 77.9 (16.7) 75.6 (18.2)
Data are presented as mean (SD) unless otherwise indicated
a
Shoulder pain at rest and at activity was assessed on a 100 mm visual analogue scale of 0 to 100, with 0 denoting no pain and 100 denoting extreme pain
b
Scoring system for evaluation of various shoulder disorders consisting of both objective (range of motion and strength) and subjective measurements (pain
assessment, workload, and leisure time activities), summarised in a score between 0 and 100; higher score indicates better shoulder function
c
Based on 12 questions with yes (1) or no (0) response options; maximum score is 12, indicating normal shoulder function; minimum score of 0 points indicates
severely diminished shoulder function
d
Generic health related quality of life instrument comprising 15 dimensions; maximum score is 1 (full health), and minimum score is 0 (death)
e
Generic health related quality of life instrument to quantify the physical, functional, and psychological aspects of health-related quality of life. It consists of 36
questions in eight subscales that assess physical, functional, social, and psychological well-being. Score ranges from 0 to 100, where a higher score is associated with
better health
Bäck et al. BMC Musculoskelet Disord (2021) 22:889 Page 6 of 10

Table 3 Work status at 24 and at 60 months by treatment group. N denotes number of patients and % percentage within group
Treatment group ASD DA ET
N % N % N %

Work status at 24 months

At work 47 82% 47 80% 53 78%
Sickness absence 6 11% 8 14% 9 13%
Unemployed/student 1 2% 1 2% 0 0%
Retired 3 5% 2 3% 5 7%
Death/Withdrawn/Loss-to-F-up 0 0% 1 2% 1 1%
TOTAL 57 100% 59 100% 68 100%
Work status at 60 months
At work 38 67% 41 69% 45 66%
Sickness absence 11 19% 8 14% 7 10%
Unemployed/student 0 0% 1 2% 1 1%
Retired 4 7% 5 8% 9 13%
Death/Withdrawn/Loss-to-F-up 4 7% 4 7% 6 9%
TOTAL 57 100% 59 100% 68 100%

Trajectories of RTW were virtually identical by time In the ET group, the overall incidence of surgery due to
points between the study groups (Fig. 2A). Trajectories persistent symptoms was 15/71. Of these, 14 participants
of RTW by baseline work status are shown in Fig. 2B. underwent ASD, and one underwent acromioclavicular
No difference between the three study groups was found resection. Further, three participants experienced persist-
in RTW during the 24-month follow-up period (Chi- ing symptoms after surgery and underwent consequent
square p = 0.93). In the ASD group 50 patients out of 57 reoperations.
(88%) returned to work at least at two time points, while Of the 29 patients who were unblinded, 24 (83%)
the respective figures were 52 out of 59 (88%) in the DA returned to work while the respective figure for those
group and 61 out of 68 (90%) in the ET group. who remained in their assigned study groups were 139
out of 155 (90%) (Chi-square 1.16, p = 0.282).
Predictors of RTW​
Participation in working life (odds ratio 3.48; 95% confi- Discussion
dence interval 1.05 to 11.55) and age below the median Principal findings
(0.92; 0.85 to 0.99) at baseline predicted RTW in the This multicentre, randomised, placebo-controlled surgi-
multivariable logistic regression (Table 4). Accuracy of cal trial involving patients with shoulder impingement
the prediction was 64% (117/183) and total R-square was syndrome shows that arthroscopic subacromial decom-
8%. The predictive value of the negative prediction (not pression (ASD) was not superior to diagnostic arthros-
returning to work) was 0.19 only, while the correspond- copy (DA) or exercise therapy (ET) in enabling patients
ing value of positive prediction (returning to work) was to return to work over the course of up to 60-month
0.94. follow–up. Participation in working life at baseline and
younger age predicted sustainable return to work, but
On‑treatment analysis of unblinding differences in participation in working life by these pre-
After surgery, the incidence of persisting symptoms suf- dictors were minimal at 24-month follow-up and the
ficiently severe to lead to unblinding of the study-group accuracy of the predictive model was unacceptably poor
assignment was similar in the two surgically treated for clinical practice.
groups (6/59 in the ASD group and 9/63 in the DA group,
p = 0.49). In the ASD group, 2 participants underwent a Comparison with other studies
consequent re-operation, manipulation under anaesthe- There are three previous studies that have reported on
sia for one participant while the other first had acromio- RTW after treatment of subacromial pain syndrome, all
clavicular resection and then later manipulation under unblinded, open trials comparing surgery to exercise
anaesthesia. In the DA group, there were seven ASDs therapy [16–18]. The short- to medium-term (3 month
and one ASD coupled with subscapularis tendon repair. and 2-year follow-up) results showed that there was no
Bäck et al. BMC Musculoskelet Disord (2021) 22:889 Page 7 of 10

Fig. 2 Trajectories of RTW. The Y-axis shows the proportion of patients at work (%) at each follow-up time point. The scale of X-axis is months. A
Trajectories by treatment group: ASD denotes arthroscopic subacromial decompression, DA diagnostic arthroscopy, and ET exercise therapy. B
Trajectories by baseline work status

difference in proportion of shoulder-related absence was no difference in return to work, sick leave days nor
from work between the groups [16, 17]. These findings retirement rates due to shoulder problems between the
mirror closely with the results of our ASD vs. ET com- groups [18].
parison (Fig. 2). Two of these trials have also compared The existing literature concerning predictors of RTW
the RTW outcomes between ASD and exercise therapy among patients with subacromial pain syndrome is
at a longer, 5-year follow-up, finding no differences in very limited. Engebretsen at al [20] found that low edu-
self-reported work status or working ability [17, 19]. In cation, probably acting as a proxy for work demands,
an extended 10-year follow-up of the latter trial, there was the most consistent predictor of pain and disability,
and work status at 1-year follow-up. Also, the baseline
Bäck et al. BMC Musculoskelet Disord (2021) 22:889 Page 8 of 10

Table 4 Predictors of return to work in multivariable logistic sickness absence [25]. Our arbitrary definition of RTW
regression: Odds ratios (OR) and 95% confidence intervals (CI) consisting of being at work at the follow-up time points
Predictor OR 95% CI P-value deviates from the common definitions for sustainable
RTW, but we also assessed the trajectories of return to
Study group: Diagnostic arthroscopy 1.13 0.35 to 3.69 0.84 work per follow-up time point per group: either way, no
Study group: Exercise therapy 1.32 0.42 to 4.17 0.64 difference between the study groups was found. Moreo-
Baseline work status: Working 3.48 1.05 to 11.55 0.04 ver, the potential bias does not differ between the AD
Physical strain at work: Yes 0.62 0.20 to 1.93 0.41 and DA study groups since the outcome assessors were
Gender: Male 0.57 0.21 to 1.58 0.28 blinded to the type of operation.
Age (Centred) 0.92 0.85 to 0.99 0.02 In addition to our primary sham-surgery controlled
efficacy comparison between ASD and DA, our study
also included a pragmatic, exploratory secondary com-
parison between surgical and non-operative care (ASD
level of shoulder symptoms, previous shoulder pain and
vs. ET). In apparent contrast to four previous, similar
self-reported health status predicted outcome.
randomised trials that found no benefit of ASD over vari-
ous exercise therapy regimens [8, 9, 13, 26], we observed
Strengths and limitations of study
a small but statistically significant benefit of ASD over ET
As surgery has been shown to produce a profound pla- in patient-rated outcomes on pain and function at 2 years
cebo response [21–23], the actual treatment effect is [5]. However, ASD provided no benefit over DA or ET at
impossible to distinguish from the nonspecific (and 5 years in the FIMPACT trial [11] and the rates of sus-
placebo) effects – such as the patients’ or researchers’ tainable RTW did not differ between the ET and ASD
expectations of benefit – without a placebo compari- study groups.
son group [24]. This factor known as performance bias The absence of a clinically relevant benefit of ASD
is a particularly problematic in trials comparing surgery in the 2- and 5-year follow-up reports of this trial [5,
to non-operative care. In our trial, it concerns primar- 11] has prompted assertions that our trial design is
ily the ASD vs. ET comparison while the ASD vs. DA fraught with certain methodological flaws. Allegedly,
comparison was theoretically less affected, especially as our recruited patients were too young to have shoulder
similar proportion of patients in the ASD and DA groups impingement syndrome or were recruited too early, in
guessed correctly whether they had undergone the actual so called reactive phase of the disease (< 6 months from
surgery or a placebo procedure. We are also pleased with the onset of symptoms). At baseline, the mean age of
our high participation rate during the 5-year follow-up. our patients was 50 years. The median duration of their
Our study had some limitations. The study sample cal- symptoms was 13 months and only nine patients had
culations for the trial were based on continuous variables had symptoms shorter than 6 months. Also, in our
and the present study is apparently underpowered to pre-specified subgroup analysis comparing those with
detect subtle differences in the dichotomous RTW out- symptoms less than 12 months to those with symp-
come variables between the treatment arms. Moreover, toms longer than 12 months, no between-group dif-
the number of participants with RTW data available who ference was found between DA and ASD groups [5].
completed the entire 2-year follow-up was 57 and 59 in Concerns have also been raised that a marginal resec-
the ASD and DA groups, below the pre-specified target tion of subacromial bursal tissue that was carried out in
of 68. However, the lack of any difference between the 30% of the patients randomised to diagnostic arthros-
treatment groups in RTW leads us to conclude that our copy skewed our results. Our pre-specified post-hoc
findings are not based on insufficient evidence, as in an analysis on the potential effect of bursal tissue resec-
underpowered study, but rather on evidence demonstrat- tion on our findings failed to show any statistically sig-
ing the lack of a clinically relevant treatment benefit. nificant differences in the primary outcomes between
Our RTW outcomes were based on self-reports by those who had resection carried out versus those who
questionnaires, which may be considered as a limita- did not. Although underpowered, the marginal differ-
tion. What mitigates this concern is having a study nurse ences observed in this analysis favoured no resection
checked the work status records at each follow-up visit, [5]. Another frequent assertion is that there is possi-
looking for inconsistencies, which likely improved the bly a specific subgroup of patients who clearly benefit
quality of the data. Also, sustainable RTW is typically from the ASD procedure. This hypothesis may well be
defined as a stable full-time or part-time RTW to either true, but it has to be tested in a randomised (placebo-
original job or a modified job for a pre-defined period surgery) controlled study. Finally, even if some “subop-
of, for example, at least 3 months, without reoccurring timal” patients were included in our trial, they should
Bäck et al. BMC Musculoskelet Disord (2021) 22:889 Page 9 of 10

have been equally distributed within the three ran- Declarations

domisation groups, and accordingly, we argue that the
Ethics approval and consent to participate
between-group comparisons remain valid. This study was approved by the Institutional Review Board of the Pirkanmaa
Hospital District (R04200; December 28, 2004) and informed consent was
obtained from all patients.

Conclusions and implications Consent for publication

The results of this randomised controlled trial indicate Not applicable.
that arthroscopic subacromial decompression pro-
Competing interests
vides no benefit over diagnostic arthroscopy or exer- All authors have completed the ICMJE uniform disclosure form at www.​icmje.​
cise therapy on return to work in patients with shoulder org/​coi_​discl​osure.​pdf (available on request from the corresponding author)
and declare: ST reports personal fees from Terveystalo group of companies
impingement syndrome. We also failed to find pre-
and from DBC group of companies, outside the submitted work. Authors not
dictors for return to work that would be applicable to named here have disclosed no conflicts of interest.
clinical practice. Along with previous controlled trials
on the subject, these findings form a credible body of Author details
1
Finnish Centre for Evidence‑Based Orthopaedics (FICEBO), Department
evidence that refutes the notion that patients suffering of Orthopaedics and Traumatology, University of Helsinki and Helsinki
from subacromial pain syndrome will return to work University Hospital, Töölö Hospital, P.O. Box 266, 00029 HUS, Helsinki, Finland.
2
faster if they have surgery. Given this body of evidence, Department of Orthopaedics and Traumatology, Helsinki University Hospital,
Helsinki, Finland. 3 Biostatistics Unit, Faculty of Medicine in the University
patients should be made fully aware of the potential of Helsinki and Helsinki University Hospital, Helsinki, Finland.
benefits and harms related to the various treatment
options for subacromial pain and be invited to agree on Received: 15 April 2021 Accepted: 24 September 2021

a treatment plan as a fully informed participants in the

shared decision-making between clinician and patient.
At a policy level, we strongly encourage decision mak-
References
ers to consider alternatives to subacromial decompres- 1. Vitale MA, Arons RR, Hurwitz S, et al. The rising incidence of acromio-
sion for managing shoulder impingement symptoms. plasty. J Bone Joint Surg Am. 2010;92(9):1842–50.
2. Karjalainen TV, Jain NB, Page CM, et al. Subacromial decompres-
sion surgery for rotator cuff disease. Cochrane Database Syst Rev.
Abbreviations 2019;1:CD005619.
ASD: arthroscopic subacromial decompression; DA: diagnostic arthroscopy; 3. Vandvik PO, Lahdeoja T, Ardern C, et al. Subacromial decompression
ET: exercise therapy; FIMPACT​: Finnish Subacromial Impingement Arthroscopy surgery for adults with shoulder pain: a clinical practice guideline. BMJ.
Controlled Trial; RTW​: return to work. 2019;364:l294.
4. Beard DJ, Rees JL, Cook JA, et al. Arthroscopic subacromial decompres-
Acknowledgements sion for subacromial shoulder pain (CSAW): a multicentre, pragmatic,
We thank the patients for their participation. We would like to thank the entire parallel group, placebo-controlled, three-group, randomised surgical trial.
FIMPACT investigators study group for the execution of the trial. Alan Cassels, Lancet. 2018;391(10118):329–38.
MPA, is acknowledged for his linguistic revisions. 5. Paavola M, Malmivaara A, Taimela S, et al. Subacromial decompression
versus diagnostic arthroscopy for shoulder impingement: randomised,
Authors’ contributions placebo surgery controlled clinical trial. BMJ. 2018;362:k2860.
FIMPACT Investigators collected the data, ST designed the database and 6. Hata Y, Saitoh S, Murakami N, et al. A less invasive surgery for rotator cuff
cleaned the data. ST, MB, and PA contributed to the study analysis plan and tear: mini-open repair. J Shoulder Elb Surg. 2001;10(1):11–6.
data analysis. MB, ST, and TJ participated in the analysis and interpretation of 7. Coghlan JA, Buchbinder R, Green S, et al. Surgery for rotator cuff disease.
the data. MB, ST, and TJ drafted the article. All authors contributed to final data Cochrane Database Syst Rev. 2008;1:CD005619.
interpretation and approved the final draft of the manuscript. 8. Ketola S, Lehtinen J, Arnala I, et al. Does arthroscopic acromioplasty
provide any additional value in the treatment of shoulder impingement
Funding syndrome?: a two-year randomised controlled trial. J Bone Joint Surg Br.
The FIMPACT trial was supported by the Sigrid Juselius Foundation and the 2009;91(10):1326–34.
State funding for university-level health research (Tampere and Helsinki Uni- 9. Haahr JP, Ostergaard S, Dalsgaard J, et al. Exercises versus arthroscopic
versity Hospitals). The funders of the study had no role in study design, data decompression in patients with subacromial impingement: a ran-
collection, data analysis, data interpretation, or writing of the report. Sponsors domised, controlled study in 90 cases with a one year follow up. Ann
had no access to the data and did not perform any of the study analysis. The Rheum Dis. 2005;64(5):760–4.
corresponding author had full access to all the data in the study and had final 10. Paavola M, Malmivaara A, Taimela S, et al. Finnish subacromial impinge-
responsibility for the decision to submit for publication. ment arthroscopy controlled trial (FIMPACT): a protocol for a ran-
domised trial comparing arthroscopic subacromial decompression
Availability of data and materials and diagnostic arthroscopy (placebo control), with an exercise therapy
The datasets generated and/or analysed during the current study are not control, in the treatment of shoulder impingement syndrome. BMJ Open.
publicly available because the informed consent of the FIMPACT trial did not 2017;7(5):e014087.
include a provision for data sharing (trial launched in 2005 when such policy 11. Paavola M, Kanto K, Ranstam J, et al. Subacromial decompression versus
was not endorsed). At present, the dataset cannot thus be shared due to a diagnostic arthroscopy for shoulder impingement: a 5-year follow-up of
potential breach of the Finnish Personal Data Act. We intend to rectify this a randomised, placebo surgery controlled clinical trial. Br J Sports Med.
situation by renewing the informed consents of the trial. The corresponding 2020.
author can be contacted for information related to the study. 12. Ellman H. Diagnosis and treatment of incomplete rotator cuff tears. Clin
Orthop Relat Res. 1990;254:64–74.
Bäck et al. BMC Musculoskelet Disord (2021) 22:889 Page 10 of 10

13. Brox JI, Staff PH, Ljunggren AE, et al. Arthroscopic surgery compared 20. Engebretsen K, Grotle M, Bautz-Holter E, et al. Predictors of shoulder pain
with supervised exercises in patients with rotator cuff disease (stage II and disability index (SPADI) and work status after 1 year in patients with
impingement syndrome). BMJ. 1993;307(6909):899–903. subacromial shoulder pain. BMC Musculoskelet Disord. 2010;11:218.
14. Ellman H. Arthroscopic subacromial decompression: analysis of one- to 21. Buchbinder R, Osborne RH, Ebeling PR, et al. A randomized trial of
three-year results. Arthroscopy. 1987;3(3):173–81. vertebroplasty for painful osteoporotic vertebral fractures. N Engl J Med.
15. Tashjian RZ, Deloach J, Porucznik CA, et al. Minimal clinically important 2009;361(6):557–68.
differences (MCID) and patient acceptable symptomatic state (PASS) for 22. Moseley JB, O’Malley K, Petersen NJ, et al. A controlled trial of arthroscopic
visual analog scales (VAS) measuring pain in patients treated for rotator surgery for osteoarthritis of the knee. N Engl J Med. 2002;347(2):81–8.
cuff disease. J Shoulder Elb Surg. 2009;18(6):927–32. 23. Kallmes DF, Comstock BA, Heagerty PJ, et al. A randomized trial
16. Brox JI, Gjengedal E, Uppheim G, et al. Arthroscopic surgery versus super- of vertebroplasty for osteoporotic spinal fractures. N Engl J Med.
vised exercises in patients with rotator cuff disease (stage II impingement 2009;361(6):569–79.
syndrome): a prospective, randomized, controlled study in 125 patients 24. Buchbinder R. Meniscectomy in patients with knee osteoarthritis and a
with a 2 1/2-year follow-up. J Shoulder Elb Surg. 1999;8(2):102–11. meniscal tear? N Engl J Med 2013;Epub date march 18, 2013.
17. Haahr JP, Andersen JH. Exercises may be as efficient as subacromial 25. Etuknwa A, Daniels K, Eib C. Sustainable return to work: a system-
decompression in patients with subacromial stage II impingement: 4-8- atic review focusing on personal and social factors. J Occup Rehabil.
years’ follow-up in a prospective, randomized study. Scand J Rheumatol. 2019;29(4):679–700.
2006;35(3):224–8. 26. Farfaras S, Sernert N, Hallstrom E, et al. Comparison of open acromio-
18. Ketola S, Lehtinen JT, Arnala I. Arthroscopic decompression not recom- plasty, arthroscopic acromioplasty and physiotherapy in patients with
mended in the treatment of rotator cuff tendinopathy: a final review of subacromial impingement syndrome: a prospective randomised study.
a randomised controlled trial at a minimum follow-up of ten years. Bone Knee Surg Sports Traumatol Arthrosc. 2016;24(7):2181–91.
Joint J. 2017;99-B(6):799–805.
19. Ketola S, Lehtinen J, Rousi T, et al. No evidence of long-term benefits of
arthroscopicacromioplasty in the treatment of shoulder impingement Publisher’s Note
syndrome: five-year results of a randomised controlled trial. Bone Joint Springer Nature remains neutral with regard to jurisdictional claims in pub-
Res. 2013;2(7):132–9. lished maps and institutional affiliations.

Ready to submit your research ? Choose BMC and benefit from:

• fast, convenient online submission

• thorough peer review by experienced researchers in your field
• rapid publication on acceptance
• support for research data, including large and complex data types
• gold Open Access which fosters wider collaboration and increased citations
• maximum visibility for your research: over 100M website views per year

At BMC, research is always in progress.

Learn more biomedcentral.com/submissions