The impact of shoulder pathologies on job discontinuation and return to work: a pilot ultrasonographic investigation

Abstract

BACKGROUND

Ultrasound imaging has emerged as one of the most useful tools for evaluating shoulder disorders. To date, the association between shoulder ultrasonography and a patient’s work status has rarely been explored by antecedent studies.

AIM

This study aimed to investigate the association between sonographically diagnosed shoulder pathologies and job discontinuation and return to work.

DESIGN

A cross-sectional study.

SETTING

Outpatient clinic in the university hospital.

POPULATION

Fifty-nine patients who were older than 20 years of age and had worked in a full-time job within the past three years.

METHODS

All participants underwent clinical evaluation using the visual analog scale (for pain), Shoulder Pain and Disability Index, Pittsburgh Sleep Quality Index, and shoulder ultrasound examination. The work-related ergonomic risks, including dealing with heavy objects, repeated use and requiring forceful motion of the affected upper extremity, were assessed. The ultrasound-identified shoulder pathologies associated with job discontinuation, that is, sick leave due to painful shoulder for more than two consecutive months, were considered as the primary outcome. In the job discontinuation subgroup, we further investigated the association between return to work and the clinical/sonographic findings.

RESULTS

Univariate analysis revealed a positive association between job discontinuation and shoulder surgery or work types requiring forceful upper-limb movements. Multivariate analysis demonstrated that job discontinuation was positively associated with supraspinatus tendon full-thickness tears (risk ratio, 8.80; 95% CI, 1.77-10.56; P=0.018). Of the patients who received shoulder surgery, 46.6% had recurrent rotator cuff tears. Return to work was likely to be related to pain scores during overhead activities and shoulder function impairment but not to sonographic findings.

CONCLUSIONS

Job discontinuation is associated with shoulder surgery, work that necessitates forceful upper-extremity movements and supraspinatus tendon full-thickness tears detected by ultrasound.

CLINICAL REHABILITATION IMPACT

Sonographic findings should not be used as the only standard for evaluating the patient’s work capability.

In patients with painful shoulders, pathologies (e.g., effusion, bursitis, calcification, tendinopathy, and tears) pertaining to the rotator cuff and adjacent structures are common.¹ Among those, rotator cuff tendon tears appear to be the most detrimental, leading to pain, disturbed sleep, limited movement, and reduced muscle strength.² Complementing physical examination, ultrasound imaging have emerged as the most useful tool for evaluating shoulder disorders. Compared with plain radiography, computed tomography, and magnetic resonance imaging (MRI), ultrasound excels in its portability, zero radiation, resolution for superficial soft tissues, and allowance for dynamic assessment and guided intervention.^3-5 Furthermore, the accuracy of ultrasonography in comparison with MRI and arthroscopic findings has been validated by numerous studies. A previous meta-analysis⁶ revealed that the overall estimated sensitivity and specificity of ultrasonography for detecting full-thickness rotator cuff tendon tears exceeded 0.90. In terms of partial-thickness tears, the pooled specificity was also greater than 0.90, although the sensitivity was lower (0.67-0.83). Considering its convenience and diagnostic performance, ultrasound has been regarded as the first-line tool for evaluating shoulder disorders in different clinical and research settings.

Looking into the problem from a social and occupational perspective, rotator cuff injuries hamper work ability, especially for work requiring repeated movements or significant upper limb muscle strength. Moreover, the job itself can be a hazard for chronic shoulder tendon pathologies. In 2020, a retrospective observational study highlighted that the use of hand tools, exposure to mechanic stress on the upper limbs, and non-ergonomic postures were prevalent risk factors in workers of a car assembly company.⁷ A cross-sectional study investigating workers (mean age: 38.3 years) in the electronics enterprise showed a high prevalence of shoulder symptoms (30.5%), which was associated with repetitive use, posture, and force during work.⁸ Needless to say, workers with severe shoulder disorders opt to discontinue their jobs to pursue rehabilitation for recovery. Another important issue is the return to work after job discontinuation, which is related to employment status, type of work, nature of tendon injury, or the applied technique for those who undergo surgery.⁹

To date, the possible impact of shoulder ultrasonography on a patient’s work status has rarely been explored by antecedent studies. Therefore, the present study aimed to investigate the association between shoulder sonographic findings and job discontinuation as well as return to work. In addition, the relationships among work status, clinical evaluation (including visual analog scale of pain, Shoulder Pain and Disability Index [SPADI], and Pittsburgh Sleep Quality Index [PSQI]), and work-related factors were also examined. To address this gap, we hypothesized that shoulder sonographic imaging would be helpful in differentiating the status of job discontinuation as well as return to work in the painful shoulder population.

Materials and methods

Participants

Participants had to be older than 20 years of age and worked in a full-time job in the past three years. Between May 2019 and October 2020, patients consecutively referred from the orthopedics department were included in our study. They were initially assessed by orthopedic surgeons, who determined that an ultrasound examination was necessary to clarify the cause of their shoulder pain or to monitor for rotator cuff retear following repair surgery. Once the patients were scheduled in the ultrasonographic examination laboratory, we invited them to participate in the study. Prior to their ultrasound examination, we asked them to complete informed consent forms and clinical evaluations/questionnaires to gather additional data. Participants also had to be capable of answering a questionnaire.

The exclusion criteria were as follows: 1) impaired cognitive function; 2) uncontrolled medical conditions (e.g., diabetic ketoacidosis and unstable angina); 3) neurological disorders that might affect muscle strength (e.g., stroke and spinal cord injury); 4) inflammatory diseases (e.g., rheumatoid arthritis and systemic lupus erythematosus).

Clinical evaluation

The demographic features of the participants (age, sex, body weight, body height, smoking habit, history of previous shoulder surgery, and existing systemic diseases such as diabetes mellitus and cardiovascular disease) were recorded. We also inquired whether the patients had received physical therapy for shoulder pathologies. Job characteristics and work duration were documented as well.

Shoulder pain severity was graded using the visual analog scale (VAS, 0-10 cm) at rest, at night, and during motion. The SPADI,^{10, 11} a patient-reported questionnaire with 13 items, was utilized. It comprises one domain for pain (five items) and another domain for the extent of difficulty during daily activities (eight items). The initial overall score ranges from 0 to 130. This score is subsequently transformed into a percentage score between 0 (no disability) and 100 (most disability). According to a systematic review,¹² the SPADI demonstrated good reliability, with intra-class correlation coefficients ranging from 0.84 to 0.94, and a minimal clinically important difference between 8 and 13. A study involving 129 patients with shoulder pain showed that SPADI has good validity, with a moderate correlation (r>0.60) with visual analogue scale of pain during active and passive movement.¹³

The sleep status of the participants was also evaluated using the PSQI,¹⁴ a self-reported questionnaire with 19 items that generate seven component scores. The overall score ranged from 0 to 21. A higher PSQI score indicates poorer sleep quality in the last month. In a study involving 80 patients with primary insomnia,¹⁵ the PSQI was found to have high test-retest reliability (r=0.87). The validity of PSQI was demonstrated by the high correlations between PSQI and sleep log data (r=0.81 for sleep duration and r=0.71 for sleep onset latency). The minimal clinically important difference of PSQI was identified as 5.5 through a study that enrolled 66 patients with arthroscopic rotator cuff repair followed up for six months.¹⁶

Assessment of work-related factors

The characteristic of their jobs and the length of time that they had worked for were documented. The work-related ergonomic risk factors were assessed, including dealing with heavy objects, repeated use and requiring forceful motion of the affected upper extremity (defined as manipulating a load greater than 4 kg).¹⁷ Job discontinuation was defined as sick leave associated with painful shoulder for more than two consecutive months.¹⁸ Return to work was regarded as being able to resume a stable job after resolution of shoulder symptoms.

Ultrasound examination

Ultrasound scanning was performed by a certified physiatrist (with more than 10 years of experience in musculoskeletal ultrasound) according to the EURO-MUSCULUS/USPRM Basic Scanning Protocols for Shoulders.^{19, 20} A 13-18 MHz high-frequency linear transducer (Aplio 500, Canon Medical Systems Europe B.V., the Netherlands) was used to obtain the images. Patients were seated during the examination. The transducer was initially placed at the level of the proximal bicipital groove in the horizontal plane to check the long head of the biceps tendon (short-axis view). The transducer was later pivoted in the sagittal plane between the greater and lesser tubercles to visualize the tendon along its long axis. To scan the subscapularis tendon, the transducer was returned to the horizontal plane at the level of the coracoid process. The tendon, as well as its underlying anterior glenohumeral capsule, was identified as extending to the lesser tubercle. The acromioclavicular joint was further inspected between the lateral clavicle and medial acromion by relocating the transducer more cranially.

To examine the supraspinatus tendon, the patient was asked to place the palm of the target side over the ipsilateral buttock, with the elbow pointing backward. The transducer was first positioned along the short axis of the tendon and moved anteriorly from the rotator interval to the middle facet of the humeral greater tubercle posteriorly. Next, the transducer was rotated to the tendon’s long axis to scrutinize its lateral bony attachment, as well as its overlying subdeltoid bursa. Finally, the participant changed the palm position by placing it over the contralateral shoulder. The transducer was placed along the inferior border of the scapular spine to observe the long axis of the infraspinatus tendon and the posterior glenohumeral joint. If the participant had bilateral symptoms, sonographic findings of the more painful site were used for the analysis.

Sonographic diagnoses

Tears of the long head of the bicipital tendon were defined by its splitting or non-visualization inside the bicipital groove. Bicipital peritendinous effusion was considered pathological if its thickness was more than 1 mm (Figure 1A).²¹ Synovitis of the acromioclavicular joint was diagnosed by maximal distension of the acromioclavicular joint capsule more than 3 mm above the line linking the lateral clavicle and acromion (Figure 1B). Subdeltoid bursitis was confirmed by a bursal thickness exceeding 2 mm (Figure 1C).²² Calcification was recognized by intra-tendinous hyperechoic plaques (with acoustic shadows) (Figure 1D). The diagnostic criteria used for tendinopathy included the presence of tendon enlargement, hypoechogenicity, and an increase in the distance between fibrils.²¹ Rotator cuff tendon tears were confirmed by a visible intratendinous gap or disappearance of the fibers. Owing to the relatively large size of the supraspinatus tendon, we sub-categorized the tears into partial-thickness (Figure 1E) and full-thickness (Figure 1F).

Figure 1.

—Exemplary shoulder pathologies on ultrasonographic evaluation: A) Bicipital peri-tendinous effusion (asterisks); B) acromioclavicular synovitis (black arrows); C) subdeltoid bursitis (double headed arrows); D) supraspinatus tendon calcification (white arrow); E) supraspinatus partial thickness tear (stars); F) supraspinatus full-thickness tear (black arrowheads). DEL: deltoid; BIC: long head of the biceps brachii tendon; HH: humeral head; ACR: acromion; CLA: clavicle; SS: supraspinatus.

Statistical analysis

Categorical variables are presented as numbers and percentages and compared using the Chi-square test (or Fisher’s Exact Test for sparse data). Continuous variables are shown as mean and standard deviation and compared using one-way analysis of variance (or Mann-Whitney U Test in case of non-normally distributed data). To check the normality of continuous variable distributions, the Shapiro-Wilk Test was used. Treating job discontinuation as a dichotomous dependent variable (i.e., present or absent), a logistic regression analysis was conducted. The following independent variables were included in a full model for analysis: sex, age, shoulder surgery, and various shoulder pathologies, as well as whether the patient’s job required forceful motion of the affected upper extremity. The association was quantified by using the risk ratio, which was derived from the following formula: risk ratio = odds ratio / [(1 ‒ P_ref)_{+ (}P_ref × odds ratio)].²¹ P_ref indicated the prevalence of the outcome in the reference group. Our primary outcome was the association of ultrasound-identified pathologies with job discontinuation, analyzed by logistic regression, as each shoulder pathology was likely to be auto-correlated. Given that the primary pathologies may differ significantly between the operated and non-operated participants (e.g., massive rotator cuff tear for the former group and tendinopathy for the latter group), we conducted a separate analysis by the status of job discontinuation. In the group of job discontinuation, we also investigated the association between return to work and clinical/sonographic features. Statistical analyses were conducted using SPSS v. 21.0 (SPSS Statistics for Windows, IBM Corp., Armonk, NY, USA). P<0.05 was considered statistically significant.

Ethics approval and consent to participate

This study was approved by the institutional review board of National Taiwan University Hospital (IRB No. 201809020RIND), and all participants were asked to provide written informed consent before enrollment.

Results

Participant characteristics

A total of 59 subjects were enrolled after excluding three participants who had been unemployed for more than three years. Participants who discontinued their jobs (n=27) accounted for 45.7% of the target population. Compared to those who continued working, patients who discontinued their job had a higher proportion of receiving shoulder surgery (P<0.001; Pearson Chi-square value 17.356) and experienced more occupational/forceful upper limb movements (P=0.042; Pearson Chi-square value 4.144) (Table I). Furthermore, 17 participants were able to return to work in the job discontinuation group (Figure 2A).

Table I. —Univariate comparisons between groups with and without job discontinuation.

Characteristics	With job discontinuation (N.=27)	Without job discontinuation (N.=32)	P value	Statistics
Clinical characteristics
Age (year)	51.78±11.52 (47.22 to 56.33)	54.00±10.57 (50.19 to 57.81)	0.443	0.597 c
Female	18 (66.67%)	15 (46.88%)	0.127	2.327 a
Height (cm)	161.50±8.75 (158.04 to 164.97)	165.89±9.06 (162.57 to 169.22)	0.067	3.497 c
Weight (kg)	64.68±13.34 (59.41 to 69.69)	64.10±11.30 (59.95 to 68.24)	0.857	0.033 c
Body Mass Index (kg/m2)	24.83±4.93 (22.88 to 26.78)	23.16±2.75 (22.15 to 24.17)	0.112	2.603 c
Diabetes	1 (3.70%)	3 (9.38%)	0.617	0.745 a
Cardiovascular diseases	3 (11.11%)	9 (28.13%)	0.106	2.616 a
Smoking	2 (7.41%)	3 (9.38%)	1.000	0.073 a
Shoulder surgery	15 (55.56%)	2 (6.35%)	<0.001*	17.356 a
Rehabilitation	19 (70.37%)	22 (68.75%)	0.546	0.365 a
Job tenure (year)	28.69±13.26 (23.09 to 34.29)	31.32±10.86 (27.11 to 35.53)	0.435	0.620 c
Dealing with heavy objects during work	23 (85.19%)	21 (65.63%)	0.086	2.955 a
Repeated use of the affected upper extremity during work	25 (92.59%)	29 (90.63%)	1.000	0.073 a
Requiring forceful motion of the affected upper extremity during work	25 (92.59%)	23 (71.88%)	0.042*	4.144 a
Pain duration (month)	4.13±2.05 (3.32 to 4.94)	5.11±1.52 (4.56 to 5.67)	0.069	314.500 b
Pain and disability
VAS at rest	3.20±2.20 (2.31 to 4.09)	2.40±1.70 (1.75 to 3.04)	0.133	2.330 c
VAS at night	4.79±2.30 (3.86 to 5.71)	4.86±2.56 (3.89 to 5.84)	0.907	0.014 c
VAS at overhead	5.02±2.88 (3.88 to 6.16)	5.55±3.06 (4.39 to 6.72)	0.373	337.500 b
Pain domain of SPADI	52.07±26.96 (41.41 to 62.74)	49.84±25.53 (40.64 to 59.05)	0.746	0.106 c
Function domain of SPADI	30.09±26.60 (19.57 to 40.61)	26.45±20.75 (18.97 to 33.93)	0.802	415.500 b
Total score of SPADI	38.55±24.81 (28.73 to 48.36)	35.41±21.58 (27.63 to 43.19)	0.605	0.270 c
Pittsburgh Sleep Quality Index	8.52±3.78 (7.02 to 10.01)	6.94±2.84 (5.91 to 7.96)	0.072	3.362 c
Ultrasound findings
Right site	16 (59.26%)	17 (53.13%)	0.636	0.224 a
Bicipital peritendinous effusion	7 (25.93%)	8 (25.00%)	0.935	0.007 a
Biceps tendon tear	0 (0.00%)	2 (6.25%)	0.495	1.747 a
Subscapularis tendinopathy	4 (14.81%)	4 (12.50%)	1.000	0.067 a
Subscapularis tendon calcification	5 (18.52%)	8 (25.00%)	0.550	0.358 a
Subscapularis tendon tear	2 (7.41%)	0 (0.00%)	0.205	2.454 a
Acromioclavicular joint synovitis	0 (0.00%)	4 (12.50%)	0.118	3.620 a
Subdeltoid bursitis	3 (11.11%)	8 (25.00%)	0.172	1.862 a
Supraspinatus tendinopathy	7 (25.93%)	12 (37.50%)	0.343	0.899 a
Supraspinatus partial thickness tear	5 (18.52%)	5 (15.63%)	1.000	0.087 a
Supraspinatus full thickness tear	7 (25.93%)	3 (9.38%)	0.162	2.850 a
Supraspinatus tendon calcification	5 (18.52%)	4 (12.50%)	0.719	0.410 a
Infraspinatus tendinopathy	1 (3.70%)	1 (3.13%)	1.000	0.015 a
Infraspinatus tendon tear	4 (14.81%)	0 (0.00%)	0.039*	5.086 a
Infraspinatus calcification	1 (3.70%)	2 (6.35%)	1.000	0.197 a

Continuous variables are given as mean±SD and 95% confidence interval. Categorical variables are given as number and percentage. VAS: Visual Analogue Scale; SPADI: Shoulder Pain and Disability Index; SPADI (total) = [SPADI (pain) × 0.5+SPADI (function) × 0.8] × (10/13). *P<0.05; ^a Pearson Chi-Square value given by the Chi-square or Fisher exact test; ^b Mann-Whitney U value given by Mann-Whitney U Test; ^c F value given by one-way analysis of variance.

Figure 2.

—Numeric data as regards job discontinuation and return to work in the overall population (A) and subgroups with/without related shoulder surgery (B). The percentage of each subgroup was approximated to the nearest whole number.

Sonographic findings

Univariate analysis revealed that the job discontinuation group had an increased percentage of infraspinatus tendon tears (P=0.039; Pearson Chi-square value 5.086). Logistic regression was adjusted for age, sex, shoulder surgery, and occupational forceful upper-limb motions. The sonographic variables, with case number ≤1 in either subgroup, were excluded from the analytic model as they were likely to affect the robustness of the effect estimation. Eventually, job discontinuation was independently associated with supraspinatus tendon full-thickness tears (risk ratio, 8.80; 95% CI, 1.77-10.56; P=0.018) (Table II).

Table II. —Multivariate analysis for the association between job discontinuation and sonographic findings (after adjusting for sex, age, shoulder surgery and work type).

Variables	Risk ratio (95% CI)	P value
Sex (male as reference)	1.61 (0.67 to 2.03)	0.198
Age (year)	0.91 (0.83 to 1.01)	0.070
Shoulder surgery	12.01 (3.68 to 15.29)	0.001*
Requiring forceful motion of the affected upper extremity during work	1.35 (0.97 to 1.38)	0.058
Bicipital peritendinous effusion	0.66 (0.12 to 2.19)	0.579
Subscapularis tendinopathy	2.00 (0.16 to 6.68)	0.541
Subscapularis tendon calcification	1.54 (0.29 to 3.31)	0.544
Subdeltoid bursitis	0.88 (0.09 to 3.09)	0.896
Supraspinatus tendinopathy	0.94 (0.19 to 2.13)	0.930
Supraspinatus tendon partial thickness tear	1.41 (0.17 to 4.73)	0.719
Supraspinatus tendon full thickness tear	8.80 (1.77 to 10.56)	0.018*
Supraspinatus tendon calcification	4.21 (0.65 to 7.46)	0.113

*P<0.05.

Comparison within the operated and non-operated groups

Of the participants, 17 individuals had undergone shoulder surgery on their affected shoulders and were denoted as the operated group, while the remaining 42 had not undergone surgery and were denoted as the non-operated group.

In the operated group, only two participants continued with their work (Figure 2B). Participants without job discontinuation also had a higher prevalence of cardiovascular disease and acromioclavicular synovitis. While the types of surgery mainly included rotator cuff repair (N.=16), only one patient underwent capsular release for shoulder adhesive capsulitis. The time interval between the surgery date and the completion of the questionnaire ranged from 92 to 758 days (mean 248.2; 95% CI, 144.9 to 351.5). All the patients who underwent rotator cuff repair had experienced a full-thickness tear of the rotator cuff tendons, which had caused either intractable pain or impaired their daily activities. They had previously undergone rehabilitation for at least three months, but it had failed to alleviate their symptoms. Eight patients in the operated group (46.6%) had recurrent rotator cuff tears (Figure 3 and Table III).

Figure 3.

—Exemplary sonographic findings of post-operative supraspinatus tendons: A) no retear; B) partial thickness retear (asterisk); C) full thickness retear (black arrowheads). Black arrows: suture materials; HH: humeral head.

Table III. —Subgroup analysis for groups with and without shoulder surgery.

Characteristics	Operated (N.=17)				Non-operated (N.=42)
	Job discontinuation (+) (N.=15)	Job discontinuation (-) (N.=2)	P value	Statistics	Job discontinuation (+) (N.=12)	Job discontinuation (-) (N.=30)	P value	Statistics
Age (year)	52.20±10.04 (46.64 to 57.76)	66.50±2.12 (47.44 to 85.56)	0.015*	0.500 b	51.25±13.59 (42.61 to 59.89)	53.17±10.38 (49.29 to 57.04)	0.773	169.000 b
Female	9 (60.00%)	0 (0.00%)	0.206	2.550 a	9 (75.00%)	15 (50.00%)	0.139	2.188 a
Height (cm)	163.14±8.77 (158.28 to 168.00)	163.00±5.66 (112.18 to 213.82)	0.824	13.500 b	159.46±8.65 (153.96 to 164.96)	166.09±9.28 (162.56 to 169.62)	0.052	106.000 b
Weight (kg)	63.46±13.38 (56.05 to 70.87)	69.50±4.95 (25.03 to 113.97)	0.441	9.000 b	66.21±13.72 (57.49 to 74.93)	63.72±11.56 (59.33 to 68.12)	0.745	162.000 b
Body Mass Index (kg/m2)	23.75±4.05 (21.51 to 26.00)	26.14±0.05 (25.70 to 26.58)	0.529	10.000 b	26.17±5.75 (22.52 to 29.82)	22.96±2.73 (21.92 to 24.00)	0.106	117.000 b
Diabetes	0 (0.00%)	0 (0.00%)	1.000	NA	1 (8.33%)	3 (10.00%)	1.000	0.028 a
Cardiovascular diseases	2 (13.33%)	2 (100.00%)	0.044*	7.367 a	1 (8.33%)	7 (23.33%)	0.402	1.251 a
Smoking	1 (6.67%)	0 (0.00%)	1.000	0.142 a	1 (8.33%)	3 (10.00%)	1.000	0.028 a
Rehabilitation	12 (80.00%)	2 (100.00%)	1.000	0.327 a	7 (58.33%)	20 (66.67%)	1.000	0.033 a
Job tenure (years)	29.68±13.07 (22.13 to 37.23)	39.50±9.19 (-43.09 to 122.09)	0.417	8.000 b	27.30±14.10 (17.22 to 37.38)	30.69±10.88 (26.30 to 35.09)	0.590	114.500 b
Dealing with heavy objects during work	12 (80.00%)	2 (100.00%)	1.000	0.486 a	11 (91.67%)	19 (63.33%)	0.128	3.372 a
Repeated use of the affected upper extremity during work	14 (93.33%)	2 (100.00%)	1.000	0.142 a	11 (91.67%)	27 (90.00%)	1.000	0.028 a
Requiring forceful motion of the affected upper extremity during work	14 (93.33%)	1 (50.00%)	0.228	3.192 a	11 (91.67%)	22 (73.33%)	0.247	1.711 a
Pain duration (month)	4.77±1.92 (3.71 to 5.83)	6.00±0.00 (6.00 to 6.00)	0.529	10.000 b	3.33±2.00 (2.06 to 4.61)	5.05±1.55 (4.46 to 5.64)	0.018*	92.000 b
VAS at rest	2.59±2.12 (1.36 to 3.81)	1.00±1.41 (-11.71 to 13.71)	0.333	7.000 b	3.92±2.15 (2.55 to 5.28)	2.50±1.69 (1.83 to 3.17)	0.042*	95.000 b
VAS at night	4.39±2.24 (3.10 to 5.68)	4.50±2.12 (-14.56 to 23.56)	0.933	13.000 b	5.24±2.39 (3.72 to 6.76)	4.89±2.62 (3.85 to 5.93)	0.845	155.500 b
VAS at overhead	4.50±3.11 (2.78 to 6.22)	2.50±3.54 (-29.27 to 34.27)	0.441	9.000 b	5.67±2.53 (4.06 to 7.28)	5.78±2.98 (4.60 to 6.95)	0.620	145.500 b
Pain domain of SPADI	43.47±26.77 (28.64 to 58.29)	29.00±15.56 (-110.77 to 168.77)	0.529	10.500 b	62.83±24.04 (47.56 to 78.11)	51.23±25.63 (41.66 to 60.80)	0.189	132.000 b
Function domain of SPADI	24.33±26.73 (9.53 to 39.14)	18.75±26.52 (-129.49 to 256.99)	0.721	12.500 b	37.29±25.71 (20.96 to 53.63)	26.96±20.77 (19.20 to 34.72)	0.276	140.500 b
Total score of SPADI	31.69±25.00 (17.85 to 45.54)	22.69±22.30 (-177.67 to 223.06)	0.721	12.000 b	47.12±22.70 (32.69 to 61.54)	36.26±21.65 (28.17 to 44.34)	0.198	133.000 b
Pittsburgh Sleep Quality Index	7.67±3.68 (5.63 to 9.70)	6.00±2.83 (-19.41 to 31.41)	0.618	11.500 b	9.58±3.78 (7.18 to 11.98)	7.00±2.88 (5.93 to 8.07)	0.034*	104.500 b
Right site	10 (66.67%)	2 (100.00%)	1.000	0.944 a	6 (50.00%)	15 (50.00%)	1.000	0.000 a
Bicipital peritendinous effusion	4 (26.67%)	0 (0.00%)	1.000	0.697 a	3 (25.00%)	8 (26.67%)	1.000	0.012 a
Biceps tendon tear	0 (0.00%)	1 (50.00%)	0.118	7.969 a	0 (0.00%)	1 (3.33%)	1.000	0.410 a
Subscapularis tendinopathy	3 (20.00%)	1 (50.00%)	0.426	0.883 a	1 (8.33%)	3 (10.00%)	1.000	0.028 a
Subscapularis tendon calcification	2 (13.33%)	1 (50.00%)	0.331	1.633 a	3 (25.00%)	7 (23.33%)	1.000	0.013 a
Subscapularis tendon tear	2 (13.33%)	0 (0.00%)	1.000	0.302 a	0 (0.00%)	0 (0.00%)	1.000	NA
Acromioclavicular joint synovitis	0 (0.00%)	2 (100.00%)	0.007*	17.000 a	0 (0.00%)	2 (6.67%)	1.000	0.840 a
Subdeltoid bursitis	2 (13.33%)	0 (0.00%)	1.000	0.302 a	1 (8.33%)	8 (26.67%)	0.247	1.711 a
Supraspinatus tendinopathy	4 (26.67%)	0 (0.00%)	1.000	0.697 a	3 (25.00%)	12 (40.00%)	0.485	0.840 a
Supraspinatus tendon partial thickness tear	3 (20.00%)	0 (0.00%)	1.000	0.486 a	2 (16.67%)	5 (16.67%)	1.000	0.000 a
Supraspinatus tendon full thickness tear	3 (20.00%)	1 (50.00%)	0.426	0.883 a	4 (33.33%)	2 (6.67%)	0.046*	4.978 a
Supraspinatus tendon calcification	2 (13.33%)	0 (0.00%)	1.000	0.302 a	3 (25.00%)	4 (13.33%)	0.387	0.840 a
Infraspinatus tendinopathy	0 (0.00%)	0 (0.00%)	1.000	NA	1 (8.33%)	1 (3.33%)	0.495	0.473 a
Infraspinatus tendon tear	2 (13.33%)	0 (0.00%)	1.000	0.302 a	2 (16.67%)	0 (0.00%)	0.077	5.250 a
Infraspinatus calcification	0 (0.00%)	0 (0.00%)	1.000	NA	1 (8.33%)	2 (6.67%)	1.000	0.036 a
Rotator cuff retear after surgical repair	7 (46.67%)	1 (50.00%)	1.000	0.008 a	NA	NA	NA	NA

Values are given as mean±SD and 95% confidence interval. SPADI: Shoulder Pain and Disability Index. SPADI (total) = [SPADI (pain) × 0.5+SPADI (function) × 0.8] × (10/13). NA: not applicable. *P<0.05; ^a Pearson Chi-square value given by the Chi-square or Fisher’s Exact Test; ^b Mann-Whitney U value given by Mann-Whitney U Test.

In contrast, 28.5% (N.=12) of the non-operated group discontinued their work. Additionally, patients who discontinued their jobs had a shorter duration of shoulder pain, higher VAS scores at rest, higher PSQI scores, and a higher prevalence of supraspinatus tendon full-thickness tear than those without job discontinuation (Table III).

Factors associated with return to work

In the job discontinuation group (N.=27), 62.9% (N.=17) were able to return to work. Participants who experienced persistent job discontinuation showed borderline higher VAS scores during overhead activities and SPADI functional subscale scores compared to those who returned to work (both P=0.059; Mann-Whitney U=value 47.000). Sonographic findings were similar between the two groups (Table IV).

Table IV. —Univariate comparisons for participants with and without return to work in the group of job discontinuation.

Characteristics	Without return to work (N.=10)	With return to work (N.=17)	P value	Statistics
Clinical characteristics
Age (year)	50.50±15.01 (39.77 to 61.23)	52.53±9.34 (47.73 to 57.33)	0.863	81.000 b
Female	7 (70.00%)	11 (64.71%)	1.000	0.079 a
Height (cm)	161.40±8.80 (155.11 to 167.69)	161.56±9.00 (156.94 to 166.19)	1.000	85.000 b
Weight (kg)	60.15±9.84 (53.11 to 67.19)	67.35±14.64 (59.82 to 74.88)	0.223	60.500 b
Body Mass Index (kg/m2)	23.05±2.79 (21.05 to 25.05)	25.87±5.65 (22.96 to 28.78)	0.309	64.000 b
Diabetes	0 (0.00%)	1 (5.88%)	1.000	0.611 a
Cardiovascular diseases	0 (0.00%)	3 (17.65%)	0.274	1.985 a
Smoking	0 (0.00%)	2 (11.76%)	0.516	1.271 a
Surgery	5 (50.00%)	10 (58.82%)	0.706	0.199 a
Rehabilitation	6 (60.00%)	13 (76.47%)	0.175	2.339 a
Job tenure (years)	27.56±16.46 (14.91 to 40.20)	29.37±11.52 (22.99 to 35.75)	0.953	66.000 b
Dealing with heavy objects during work	10 (100.00%)	13 (76.47%)	0.264	2.762 a
Repeated use of the affected upper extremity during work	9 (90.00%)	16 (94.12%)	1.000	0.156 a
Requiring forceful motion of the affected upper extremity during work	9 (90.00%)	16 (94.12%)	1.000	0.156 a
Pain duration (month)	3.85±1.92 (2.48 to 5.22)	4.29±2.17 (3.18 to 5.41)	0.639	75.000 b
Pain and disability
VAS at rest	3.24±2.59 (1.25 to 5.23)	3.18±2.05 (2.12 to 4.23)	1.000	76.000 b
VAS at night	4.54±2.97 (2.26 to 6.83)	4.91±1.95 (3.91 to 5.92)	0.634	67.000 b
VAS at overhead	6.50±2.37 (4.81 to 8.19)	4.15±2.85 (2.68 to 5.61)	0.059	47.000 b
Pain domain of SPADI	58.00±26.95 (38.72 to 77.28)	48.59±27.15 (34.63 to 62.55)	0.386	67.000 b
Function domain of SPADI	42.00±28.87 (21.35 to 62.65)	23.09±23.24 (11.14 to 35.04)	0.059	47.000 b
Total score of SPADI	48.15±26.01 (29.55 to 66.76)	32.90±22.98 (21.08 to 44.71)	0.127	54.500 b
Pittsburgh Sleep Quality Index	9.70±3.20 (7.41 to 11.99)	7.82±4.00 (5.77 to 9.88)	0.155	56.000 b
Ultrasound findings
Right site	6 (60.00%)	10 (58.82%)	1.000	0.952 a
Bicipital peritendinous effusion	2 (20.00%)	5 (29.41%)	0.678	0.590 a
Biceps tendon tear	0 (0.00%)	0 (0.00%)	1.000	NA
Subscapularis tendinopathy	1 (10.00%)	3 (17.65%)	1.000	0.589 a
Subscapularis tendon calcification	3 (30.00%)	2 (11.76%)	0.326	0.239 a
Subscapularis tendon tear	1 (10.00%)	1 (5.88%)	1.000	0.693 a
Acromioclavicular joint synovitis	0 (0.00%)	0 (0.00%)	1.000	NA
Subdeltoid bursitis	1 (10.00%)	2 (11.76%)	1.000	0.888 a
Supraspinatus tendinopathy	3 (30.00%)	4 (23.53%)	1.000	0.711 a
Supraspinatus tendon partial thickness tear	1 (10.00%)	4 (23.53%)	0.621	0.382 a
Supraspinatus tendon full thickness tear	2 (20.00%)	5 (29.41%)	0.678	0.590 a
Supraspinatus tendon calcification	2 (20.00%)	3 (17.65%)	1.000	0.879 a
Infraspinatus tendinopathy	1 (10.00%)	0 (0.00%)	0.370	0.184 a
Infraspinatus tendon tear	1 (10.00%)	3 (17.65%)	1.000	0.589 a
Infraspinatus calcification	1 (10.00%)	0 (0.00%)	0.370	0.184 a

Values are given as mean±SD and 95% CI. Categorical variables are given as number and percentage. VAS, Visual Analogue Scale; SPADI: Shoulder Pain and Disability Index: SPADI (total) = [SPADI (pain) × 0.5+SPADI (function) × 0.8] × (10/13); NA: not applicable. *P<0.05; ^a Pearson Chi-square value given by the Chi-square or Fisher’s Exact Test; ^b Mann-Whitney U value given by Mann-Whitney U Test.

Discussion

Our study yielded several important findings. First, there was a positive association between job discontinuation and related shoulder surgery, as well as the type of work (jobs requiring forceful upper-limb movements). Second, job discontinuation was positively related to full-thickness tears of the supraspinatus tendon. Third, in participants who discontinued their jobs, return to work was related to clinical (e.g., VAS during overhead activities) rather than sonographic findings.

The main factor associated with job discontinuation was shoulder surgery. The majority of the participants in the operated group had primary rotator cuff repairs in our study. Additionally, our findings suggest that all patients had a gap of more than 90 days between their surgery and the date of questionnaire completion. This implies that those who were unable to return to work might not have experienced temporary job discontinuation. It is noteworthy that common indications for rotator cuff repair include age less than 60 years, full thickness tears, poor response to conservative treatments, vocational movements involving the shoulder, and capability to comply with post-operative care.²³ In this sense, it is also reasonable to speculate that the operated group would have a worse pre-operational baseline than those who were not operated on, thus, hindering strenuous use of the affected upper extremities. Similar to our results, if the forceful motion of the upper extremities is required during work, discontinuing the job for a certain period seems to be inevitable. After rotator cuff repair surgery, it typically takes at least six weeks for the repaired tendon to heal and three months for a strong bone attachment to form.^{24, 25} However, even after this healing period, the majority of patients in our group were unable to continue working due to the need for an extended rehabilitation period. This rehabilitation is necessary to improve shoulder strength and range of motion, and can be a time-consuming process that requires considerable effort and compliance from the patient.

Regarding sonographic findings, the supraspinatus tendon full-thickness tears had an independent positive association with job discontinuation. The supraspinatus muscle/tendon complex functions as the most important shoulder abductor, which is routinely involved in overhead activities.²⁶ Among the four rotator cuff tendons, this tendon has the highest risk to be torn, with a prevalence of 38.1-61.9% in patients with painful shoulders.²⁷ Compared with pathologies, such as tendinitis and partial thickness tears, the continuity and mechanical strength of the supraspinatus tendon are more severely compromised in full-thickness tears, thus, hampering the patients from performing labor-intensive work.²⁸

In the operated group, 46.6% of the patients were found to have recurrent rotator cuff tears, as recognized by ultrasound imaging. Notably, an ultrasound study²⁹ found that the prevalence of recurrent tears was higher in patients with massive tears than in those without (40.7% vs. 19.0%, respectively). Recurrent tears were also likely to develop in the aging population, accounting for more than 50% of patients older than 60 years. The clinical impact of recurrent tears has also been summarized in a meta-analysis.³⁰ The authors found that patients with complete rotator cuff retears had significantly lower Constant scores and University of California, Los Angeles scores compared with those with an intact or partially torn rotator cuff. The ultrasound’s ability to delineate the rotator cuff following surgery has been reported in another study.³¹ Using high-resolution ultrasound, significant changes in tendon echotexture, fibrillar pattern, and bony surface irregularity were identified from postoperative five weeks to six months. Therefore, in patients undergoing rotator cuff repair, an ultrasound examination should be performed when post-operative recovery is suboptimal (e.g., for early detection of recurrent tears).

Another important finding of our study was that a substantial portion of the participants who discontinued their jobs could not return to work when they were enrolled in the study. According to a recent meta-analysis,³² the percentage of patients who underwent rotator cuff repair without the ability to resume work at their previous work level was approximately 37.7%, similar to the finding in our study. The meta-analysis also indicated that the average interval between the surgery and the moment of work recommencing was 8.15 months. The aforementioned observations may also be attributed to the cross-sectional design of our study. If the patients had been followed up for at least eight months, the rate of return to work would likely have been better.

Participants who could not return to work were likely to experience more pain during overhead activities and worse shoulder disability than those who could. Similarly, a systematic review³³ revealed that an increased DASH Score (Disabilities of Arm, Shoulder, and Hand) was associated with a reduced probability of returning to work. This finding is reasonable because higher disability scores of the affected upper extremity indicate poorer recovery potential, worse injury, and more time required for rehabilitation. Surprisingly, in our study, none of the sonographic findings were associated with return to work, reflecting the complexity of the underlying mechanisms. Numerous factors should be considered, such as the nature of the job, psychological traits, workers’ compensation, and the extent of the injury.³³ Therefore, imaging findings should not be treated as the only reference for judging the patient’s work capability. Instead, ultrasonography can serve as an adjuvant tool for validating the integrity of the involved shoulder before returning to work.

Limitations of the study

This study has several limitations. First, the sample size was relatively small. Since no prior study had used ultrasound imaging to evaluate the correlation between shoulder pathologies and job status, it was challenging to obtain prior information regarding the prevalence of shoulder pathologies in the target groups for estimating the required sample size. A large-scale study is needed to validate our pilot observation. Second, this study used a cross-sectional design. Without longitudinal follow-ups, dynamic changes in sonographic findings in relation to job status cannot be thoroughly explored. Third, shoulder treatment strategies varied across patients and were difficult to quantify or categorize. Therefore, it was difficult to analyze the impact of different treatment options on work status. Fourth, we compared job status (returned to work versus did not return to work) in the group with job discontinuation. Although our sample size was limited, we acknowledge that our study was designed as a pilot investigation. Therefore, our objective was to identify potential trends or factors that could be investigated by future researchers. Importantly, return to work after job discontinuation is a significant issue in occupational medicine, and any factors identified as significantly different in our study would be worthy of further examination.

Conclusions

Job discontinuation is mainly associated with shoulder surgery and the need for occupational/forceful upper-limb movements. Ultrasound examination is helpful in recognizing recurrent rotator cuff tendon tears following surgery. This significantly impacts the functional reserve for handling tasks that require frequent upper limb use. Importantly, sonographic findings should not be used as the only standard for evaluating the patient’s work capability.