Gastrointestinal (GI) endoscopists are at high risk of work-related musculoskeletal injury due to daily procedures requiring repetitive push-and-pull movements, torquing, prolonged grip, and awkward postures, and the prevalence of musculoskeletal symptoms among GI endoscopists has been reported at 37% to 89%.^1-3 After overuse syndrome was reported for endoscopists in 1994,⁴ several studies have examined the musculoskeletal symptoms experienced by these professionals.^3,5-7 In a previous Korean study, 89% of endoscopists experienced musculoskeletal pain, and 47% of participants had severe pain in multiple anatomic areas. In addition, the location of the pain differed according to the participants’ experience.⁵ In another Korean study, approximately 90% of gastroenterologists also experienced musculoskeletal pain. Women had a higher total pain score than men in all age groups.⁶

A dedicated rehabilitation program for frequent musculoskeletal symptoms and overuse syndrome is needed to alleviate the musculoskeletal pain experienced by GI endoscopists. Although an endoscopy suite layout with a height-adjustable patient bed and adjustable monitor has been recommended, the usefulness of such intervention has not been evaluated. Thus, we developed an equipment/posture correction and stretching program in collaboration with GI endoscopists and rehabilitation medicine doctors. This study aimed to investigate the prevalence of musculoskeletal pain related to the daily workload of GI endoscopists, develop an easy and sustainable rehabilitation program, and evaluate the effectiveness of the program.

1. Study population and flow

GI endoscopists of various ages, sexes, and experiences in tertiary referral centers were requested to participate and voluntarily enrolled in this study. The study protocol consisted of an 8-week prospective survey. In the first 2 weeks, informed consent was obtained, and a baseline survey including baseline characteristics, baseline daily workload, and musculoskeletal symptoms was distributed. In the following 6 weeks, a novel rehabilitation program including equipment/posture correction and stretching was conducted. The daily performance of the program was self-recorded in a formatted Excel file, and the first satisfaction survey was distributed in the 6th week using a 5-point Likert scale. Finally, in the last 2 weeks, follow-up daily workload and musculoskeletal symptom surveys were distributed while the program was still in progress, and a second satisfaction survey was also distributed in the 8th week using a 5-point Likert scale (Fig. 1A). The study protocol was reviewed and approved by the Institutional Review Board of Ewha Womans University Seoul Hospital (IRB number: 2020-03-032).

Figure 1. Study design. (A) Development of the program and study overview. (B) Equipment/posture program. The equipment program was composed of monitor, scope, table, and main processor. The posture program was composed of eight components; neck, shoulder, elbow, wrist, back, leg, foot, and whole body. A detail description is provided in Supplementary Table 2.¹ (C) Enrollment of participants.
MSK pain, musculoskeletal pain.

2. Baseline, daily workload, and musculoskeletal questionnaires

The baseline questionnaire included baseline characteristics (age, sex, height, weight, job position with duration, comorbid disease, and current medication). The participants recorded the amount of time spent daily in-hospital activities including the number of patients seen for practice and procedures and the amount of time spent daily in out-hospital/home activities in conjunction with musculoskeletal pain and severity using a structured spreadsheet for the first 2 weeks. The in-hospital workload consisted of outpatient/inpatient care, regular/on-call procedures, research, education, lectures, meetings, and administrative work. The out-hospital workloads consisted of academic activity (conference, workshop, any kinds if meetings related with academic activity), private networking unrelated with occupation or housework, housework, childcare, and transport (time to go to work, time to go home, time to go anywhere). Non-workload time, such as time for meals, sleeping, and leisure, was not recorded to protect participants’ privacy. For musculoskeletal symptoms, pain from six anatomic areas (neck, shoulder, arm/elbow, hand/wrist/finger, back, and leg/foot) was recorded using a 5-point numerical scale (grade 0–4) (Supplementary Table 1).

3. Development of the equipment/posture correction and stretching program

To minimize procedure-related musculoskeletal diseases afflicting GI endoscopists, we developed a novel rehabilitation program in collaboration with gastroenterologists (S.Y.N., K.N., K.N.S., and C.H.T.) and a physiatrist (S.Y.). A multimodal physiotherapy intervention, including education on equipment/posture correction and stretching exercises, was applied.

1) Equipment/posture correction program

Based on previous reports,^1,8 we developed specialized recommendations on good equipment/posture position including proper location of the monitor, scope, table, and other factors related to the ergonomic strategies to avoid excessive muscle strain (Fig. 1B, Supplementary Table 2). We provided four recommendations for equipment and eight recommendations for good posture. The participants recorded whether they had followed the recommendations during their daily activities in the 3rd to 8th weeks.

2) Stretching program

Stretching is a form of physical exercise that flexes or extends specific muscles or tendons. Stretching increases intramuscular pressure, improves blood circulation, and removes fatigue-related substances (e.g., lactic acid). We developed a novel stretching program based on the stretching programs of the Occupational Safety and Health Administration, American Academy of Orthopedic Surgeons, and Korea Occupational Safety and Health Agency. The stretching program consisted of 12 stretching exercises for the neck, shoulder, elbow, wrist, finger, back, leg, and ankle/foot. The participants were instructed to hold the position for 10 seconds, repeat each exercise three times, and breathe in and out naturally during the exercise. Overstretching or bouncing during stretching is not recommended because of the risk of muscle strain. This stretching program was conducted from the 3rd week to end of the study.

4. Internal review and education of the stretching program

After development, we created a figure (Supplementary Fig. 1) and video for the stretching program (https://www.youtube.com/watch?v=6h7f0WMMcAo) to improve utilization. Nine gastroenterologists (S.Y.N., K.N., K.N.S., C.H.T., N.K., S.M.P., Y.S.P., S.J.P., and S.A.J.) and a physiatrist (S.Y.) performed the developed program and provided feedback on the time needed to complete it and how easy to perform. Then, we provided figures and videos to all participants at the end of 2nd week. A self-recording form for daily performance was provided to all participants (Supplementary Table 3).

5. Satisfaction survey for the equipment/posture correction and stretching program

After the start of the stretching program, a satisfaction survey was distributed at the end of the 6th week and 8th week. The survey investigated satisfaction scores for the overall program, equipment/posture correction program, and stretching program, and willingness to use this program continuously on a 5-point Likert scale (1=very unsatisfied, 2=unsatisfied, 3=average, 4=satisfied, and 5=very satisfied) (Supplementary Table 4).

6. Statistical analysis

Categorical variables were analyzed using the chi-square test or Fisher exact test. Continuous variables were analyzed using the t-tests. Correlations between workload and musculoskeletal symptoms were analyzed using the Spearman correlation analysis.

1) Definition of pain score at baseline and follow-up

The site-specific musculoskeletal pain score for each participant was the sum of the all-day scores over 2 weeks. For example, the neck pain score of an individual at baseline compromised the summation of the neck pain scores from day 1 to day 14. The neck pain score of an individual during follow-up equaled the summation of the neck pain scores from day 43 to day 56.

2) Definition of musculoskeletal pain score change

The following definitions were adopted.

• Overall pain score at baseline=summation of the multisite pain score at baseline.

• Overall pain score at follow-up=summation of multisite pain scores during follow-up.

• FB-Tsx (overall musculoskeletal pain score change)= (overall pain score during follow-up)–(overall pain score at baseline).

• Improvement (decrease of overall pain score): FB-Tsx <–5.

• Aggravation (increase of overall pain score): FB-Tsx >5.

• No change: –5≤FB-Tsx≤5.

The correlation coefficient matrix between the musculoskeletal pain score and workload or performance status was visualized using the R program (R Foundation for Statistical computing, Vienna, Austria). All statistical analyses were performed using STATA software version 15 (StataCorp, College Station, TX, USA) or R (R4.0.5). All statistical tests were two-sided, and p-values <0.05 were considered statistically significant.

1. Baseline characteristics of endoscopists

A total of 128 participants from 23 medical institutions (hospital volume: >500 beds) completed the basic questionnaires and a baseline workload/musculoskeletal symptom sheet (Fig. 1C). A total of 118 participants (69 men and 49 women) completed the stretching/posture program performance status, follow-up workload/musculoskeletal symptom score, and satisfaction score. Age distribution, hospital size, chronic disease, and use of medication did not differ between men and women. However, the duration of endoscopy after board certification was longer for women than for men (107.9 months vs 61.7 months, p=0.011) (Table 1). Baseline characteristics by endoscopic experience (career) were also provided in Table 1.

Table 1

Baseline Characteristics.

Characteristic	Sex				Career
	Men (n=69)	Women (n=49)	p-value*		Professor (n=65)	Fellow (n=53)	p-value*
Age group
30–39 yr	41 (59.4)	28 (57.1)	0.078		22 (33.8)	47 (88.7)	<0.001
40–49 yr	23 (33.3)	11 (22.5)			28 (43.1)	6 (11.3)
≥50 yr	5 (6.3)	10 (20.4)			15 (23.1)	0
Sex
Men	NA	NA			33 (50.8)	36 (67.9)	0.060
Women	NA	NA			32 (49.2)	17 (32.1)
Position
Professor	33 (48.8)	32 (67.3)	0.060		NA	NA
Fellow	36 (52.2)	17 (32.7)			NA	NA
Body mass index, kg/m2	24.9±3.0	20.7±2.0	<0.001		22.8±3.1	23.7±3.7	0.159
Hospital size
>500 beds	61 (88.4)	43 (87.8)	0.914		59 (50.8)	45 (38.1)	0.327
200–500 beds	8 (11.6)	6 (12.2)			6 (9.2)	8 (15.1)
Endoscopy duration, mo	61.7±72.3	107.9±112.2	0.011		140.3±97.4	7.9±12.6	<0.001
Chronic disease	14 (20.3)	7 (14.3)	0.401		17 (26.2)	4 (7.5)	0.009
Use of any medication	15 (24.7)	9 (18.4)	0.654		19 (29.2)	5 (9.4)	0.008

Data are presented as number (%) or mean±SD..

NA, not available..

*p-values were derived from the chi-square test or Fisher exact test..

2. Prevalence and type of musculoskeletal pain

Among the 118 participants, 94% (n=111) complained of musculoskeletal symptoms at baseline. Pain was most prevalent in the shoulder (70%). The prevalence of moderate to severe pain (overall pain score ≥20) was 11% to 31%: shoulder (31%), neck (25%), wrist/hand/finger (25%), lower back (19%), leg/foot (13%), and arm/elbow (11%) at baseline (Table 2). After the implementation of the program, the prevalence of moderate to severe pain decreased, although the difference was not statistically significant: shoulder pain (31% to 24%) and wrist/hand pain (25% to 17%).

Table 2

Prevalence of Musculoskeletal Pain (n=118).

Pain site	Prevalence of any pain scores			Prevalence of pain scores ≥20
	Baseline	Follow-up		Baseline	Follow-up
Neck	76 (64.4)	76 (64.4)		29 (24.6)	27 (22.9)
Shoulder	83 (70.0)	84 (71.2)		36 (30.5)	28 (23.7)
Arm, elbow	37 (31.4)	38 (32.2)		13 (11.0)	11 (9.0)
Wrist, hand, finger	68 (57.6)	62 (52.5)		30 (25.4)	20 (16.9)
Lower back	73 (61.9)	68 (57.6)		22 (18.6)	21 (17.8)
Leg, foot	55 (46.6)	54 (45.8)		15 (12.7)	12 (10.2)

Data are presented as number (%). Pain scores: mean summation of the pain score at each site during a 2-week period..

Statistically non-significant between baseline and follow-up..

Women spent more time on private networking and housekeeping than men (each p<0.05), and the number of colonoscopies and therapeutic endoscopies performed by women was higher than that perfumed by men (each p<0.05). Neck and shoulder pain score in women was higher than that in men (each p<0.05) (Supplementary Table 5, Supplementary Fig. 2). Total in-hospital work time was higher in professors than in fellows (108.4 hours per 2 weeks vs 88.2 hours per 2 weeks, p=0.01) (Supplementary Table 5). Outpatient number, admission patient number, and therapeutic endoscopic number were higher in professors than in fellows. Shoulder pain score was higher in professors than in fellows (16.9 vs 10.9, p=0.02). Shoulder pain was positively correlated with the number of colonoscopies (R=0.215, p=0.02) and therapeutic endoscopies (R=0.214, p=0.02). Higher neck and shoulder pain scores in women appeared to be related to increased endoscopic duration after board certification and the number of colonoscopies and therapeutic endoscopies.

3. Correlation between workload and musculoskeletal pain at baseline

Various activities were correlated with multisite musculoskeletal pain at baseline (Fig. 2A). Shoulder pain was significantly associated with the number of colonoscopy (R=0.242) and therapeutic endoscopy (R=0.183) cases. Arm/elbow pain was positively correlated with the number of colonoscopy cases (R=0.181). Wrist/hand pain showed a positive correlation with the total endoscopy time (R=0.270), social activity (R=0.281), and the number of upper endoscopy cases (R=0.288), whereas outpatient time (R=–0.241) and number of outpatients seen (R=–0.228) were negatively correlated with this type pain. There was a significant positive correlation between back pain and education time (R=0.205), in-hospital conference time (R=0.193), and number of hospitalized patients seen (R=0.229). Leg/foot pain was positively related to the time spent engaged in other in-hospital work activities (R=0.273), out-of-hospital academic meeting time (R=0.236), and number of colonoscopy cases (R=0.237).

Figure 2. Correlation between workload and musculoskeletal pain. (A) The correlation between the baseline workload and multisite musculoskeletal pain is significant. (B) The correlation between the follow-up workload and musculoskeletal pain is less prominent. Numeric number means correlation coefficient (rho). Blue arrows mean inverse association and red arrows refer to positive association.
B, baseline; F, follow-up; OPD, outpatient department; Adm, admission; EGD, upper endoscopy; CFS, colonoscopy; EndoTx, therapeutic endoscopy; No, procedure number. *p<0.05.

4. Correlation between workload and musculoskeletal pain after the rehabilitation program

After the posture/equipment correction and stretching programs, only several workloads were correlated with follow-up musculoskeletal pain (Fig. 2B). Follow-up neck pain was positively correlated with follow-up endoscopic time (R=0.185). Follow-up shoulder pain was positively associated with follow-up colonoscopy number (R=0.199). Follow-up arm/elbow pain was positively correlated with follow-up admission work time (R=0.239). Follow-up hand/wrist pain was positively associated with follow-up social time (R=0.291). Follow-up back pain was positively correlated with the number of patients admitted (R=0.203).

5. Correlation between follow-up musculoskeletal pain and program performance

Most types of musculoskeletal pain were negatively correlated with posture/equipment program performance status (Fig. 3). This shows that musculoskeletal symptoms significantly decreased when program performance was high. Follow-up neck pain was negatively correlated with good performance in the shoulder (R=–0.195) and elbow (R=–0.204) postures. Follow-up arm/elbow pain was negatively correlated with good elbow performance (R=–0.307) and wrist posture (R=0.205). The equipment/postural correction programs, including monitoring position (R=–0.183), shoulder (R=–0.201), elbow (R=–0.224), wrist (R=–0.213), leg (R=–0.184), and foot (R=–0.192) postures showed a negative correlation with leg and foot pain.

Figure 3. Correlation between follow-up musculoskeletal pain and posture/equipment performance. The number refers to the correlation coefficient (rho). P_[ ]_P is the position performance score and F_[ ] is the follow-up pain score. *p<0.05.

6. Factors associated with improvement of musculoskeletal pain

Demographic factors such as age group, sex, body mass index, underlying disease, use of medication, and position (fellow or professor) were not associated with aggravation or improvement of musculoskeletal pain (data not shown). A higher performance score in the scope position (86.8% vs 71.3%, p=0.054) and proper table height (94.1% vs 79.1%, p=0.054) were noted in the musculoskeletal pain improvement group than in the aggravation group. However, stretching performance did not differ between the improvement and aggravation groups. The scores for stretching satisfaction (3.9 vs 3.5, p=0.014) and willingness to continue the program (4.0 vs 3.6, p=0.035) were significantly higher in the improvement group than in the aggravation group (Table 3).

Table 3

Factors Associated with Overall Symptom Improvement.

Factor	FB-Tsx*
	Improve (n=54)	Aggravation (n=26)	p-value†
Equipment position, %
P_monitor_P	69.4±44.5	59.3±44.3	0.346
P_scope_P	86.8±31.4	71.3±36.4	0.054
P_table_P	94.1±23.0	79.1±34.9	0.054
P_other_P	87.8±30.8	74.5±39.6	0.105
Personal position, %
P_head_P	79.6±38.7	66.9±39.5	0.179
P_shoulder_P	75.6±38.5	81.8±28.4	0.461
P_elbow_P	74.1±42.6	75.3±33.6	0.898
P_wrist_P	87.9±30.1	82.2±27.7	0.423
P_back_P	70.7±40.8	76.6±32.6	0.518
P_leg_P	71.8±41.3	71.1±36.7	0.940
P_foot_P	65.2±44.9	62.2±41.8	0.780
P_entire_P	66.6±42.6	69.8±39.9	0.746
Stretching program
Stretching, day	22.0±9.6	22.9±10.9	0.684
Stretching, n	37.2±26.6	36.9±26.5	0.967
Satisfaction of overall program at 6 wk
Overall	3.8±0.6	3.6±0.6	0.174
Equipment/posture	3.6±0.7	3.3±0.6	0.151
Stretching	3.9±0.6	3.5±0.7	0.014
Willing to continue forward	3.8±0.6	3.6±0.7	0.461
Satisfaction of overall program at 8 wk
Overall	4.0±0.6	3.7±0.6	0.144
Equipment/posture	3.8±0.7	3.6±0.7	0.329
Stretching	4.0±0.7	3.7±0.7	0.137
Willing to continue the program forward	4.0±0.8	3.6±0.8	0.035
No. of follow-up hospital work
F-OPDNo	133.1±114.0	96.8±100.9	0.170
F-AdmNo	58.0±50.5	78.3±85.3	0.269
F-EGDNo	53.1±37.2	71.9±44.8	0.051
F-CFSNo	21.4±15.3	25.9±16.1	0.236
F-EndoTxNo	9.0±10.5	4.3±5.6	0.011
F-ERCPNo	1.7±10.0	1.4±5.1	0.833
No. of workload change
FB-OPDNo	5.3±18.0	9.5±53.6	0.700
FB-AdmNo	–12.1±42.7	–3.6±39.6	0.039
FB-EGDNo	–2.7±34.7	12.4±35.2	0.074
FB-CFSNo	0.02±7.2	6.3±11.7	0.017
FB-EndoTxNo	0.1±4.0	–0.5±9.1	0.754
FB-ERCPNo	0.4±2.3	–0.7±3.3	0.145

Data are presented as mean±SD. P_[ ]_P is the position performance score and F-[ ] is the follow-up number of procedures or patients..

FB, (follow-up)–(base); Tsx, total symptom scores; No, procedure number; OPD, outpatient department; Adm, admission; EGD, upper endoscopy; CFS, colonoscopy; EndoTx, therapeutic endoscopy; ERCP, endoscopic retrograde cholangiopancreatography..

*FB-Tsx (overall musculoskeletal pain score change)=(overall pain score at follow-up)–(overall pain score at baseline). Improvement (decrease in the overall pain score): FB-Tsx <–5. Aggravation (increase in the overall pain score): FB-Tsx >5; ^†p-values were derived from the t-test..

Follow-up work time for any type of work was not associated with aggravation or improvement in musculoskeletal pain (data not shown). A lower follow-up upper endoscopy number (53.1% in the improvement vs 71.9% in the aggravation groups, p=0.051) was associated with improvement in musculoskeletal pain (Table 3).

A decrease in the number of patients admitted (follow-up–baseline = –12.1 in the improvement group and –3.6 in the aggravation group, p=0.039) was associated with improvement in musculoskeletal pain. An increased number of colonoscopy procedures (follow-up–baseline=6.27 in the aggravation group and 0.02 in the improvement group, p=0.017) was associated with aggravation of musculoskeletal pain (Table 3).

7. Satisfaction score for the equipment/posture correction and stretching program

In the first survey (end of the 6th week), 94% of participants selected “average or satisfied” regarding equipment/posture training, 97% selected “average or satisfied” regarding the stretching program, and 93% selected “average or satisfied” regarding the willingness to use the program continuously. In the second survey (end of the study), 97% of participants selected “average or satisfied” regarding equipment/posture training, 99% selected “average or satisfied” regarding stretching program, and 93% selected “average or satisfied” regarding the willingness to use the program continuously (Supplementary Fig. 3).

In this prospective cohort study, most participants complained of musculoskeletal symptoms at baseline. Various in-hospital activities were correlated with multisite musculoskeletal pain at baseline. After the equipment/posture correction and stretching program, only a few workloads were correlated with follow-up musculoskeletal pain, suggesting that our program had a favorable effect. Follow-up musculoskeletal pain was negatively correlated with equipment/posture program performance. Factors associated with improvement of musculoskeletal pain included fewer follow-up upper endoscopies, higher performance of scope and table posture, and a decrease in the number of admitted patients. A factor associated with the aggravation of follow-up musculoskeletal pain was increase in the number of colonoscopy cases. More than 90% of participants stated being average or satisfied with the rehabilitation program and their willingness to continue using the program.

In this study, 94% of the participants complained of musculoskeletal symptoms at baseline. Pain was most prevalent in the shoulder (70%). Previous studies have reported the prevalence of musculoskeletal symptoms in GI endoscopists as 37% to 89%.^1-3 The wide prevalence range may be related to the endoscopy workload, survey method, and definition of musculoskeletal pain, such as pain site and severity, in each study. The prevalence in our study was similar to those reported by previous Korean studies. In one study, 89.1% of endoscopists experienced musculoskeletal pain, and 47.3% of the participants had severe pain in multiple anatomic areas.⁵ Another study reported that the prevalence of musculoskeletal pain at any site was 89.6%.⁶ In our study, the prevalence of moderate to severe pain (overall pain score ≥20) was 11% to 31% in the order of shoulder, neck, wrist/hand/finger, lower back, leg/foot, and arm/elbow at baseline.

Various activities were correlated with multisite musculoskeletal pain at baseline. Shoulder pain was significantly associated with the number of colonoscopy and therapeutic endoscopy cases. Arm/elbow pain was positively correlated with the number of colonoscopy cases. Wrist/hand pain was positively correlated with total endoscopy time, social activity, and number of upper endoscopy cases, whereas it was negatively correlated with outpatient time and number of outpatients. Upper extremity pain (wrist, hand, arm, elbow, and shoulder) was positively associated with the endoscopic workload. However, back pain was positively correlated with non-endoscopic workload, including education time, in-hospital conference time, and number of hospitalized patients seen. Leg/foot pain was positively related to various workloads, including time engaged in other in-hospital activities, out-of-hospital academic meeting times, and the number of colonoscopy cases.

The motions during endoscopy require wrist flexion/extension or radial/ulnar deviation.^9,10 Tenosynovitis of the left thumb is attributed to the repeated application of force by the left thumb during endoscopy.¹¹ If the monitor is placed on the side of the endoscopist or is too high, endoscopists could adopt uneasy postures, involving extension or rotation of the spine.¹² The optimal monitor location is directly in front of the endoscopist to avoid rotation and flexion of the cervical spine, and the monitor should be nearly parallel to the eye level.¹³ High monitor placement above the eye level leads to excessive cervical extension and neck strain.¹⁴ The optimal viewing angle for the monitor was 15° to 25° below the horizon of the eyes.^1,15 The optimal viewing distance of monitors is between 52 and 182 cm.¹ This broad range depends on the monitor size and endoscopist preference.¹⁶ Considering the eye height of endoscopists, distance of the monitor, and viewing angle, the center of the monitor can be placed between 93 and 162 cm above the floor.¹ The optimal bed height is the upright position of the endoscopist without flexing the trunk (85 to 120 cm), and the optimal elbow position is 90° during the procedure, with a range of 0 to 10 cm below the elbow.¹

Interestingly, after the posture/equipment correction and stretching program, only a few activities were correlated with follow-up musculoskeletal pain. After the rehabilitation program, only neck pain and shoulder pain were positively associated with the endoscopic workload (follow-up endoscopic time and number of colonoscopy cases at follow-up), whereas arm/elbow pain, hand/wrist pain, and back pain were related to the non-endoscopic workload, such as time spent on admission work and social activities.

Furthermore, most follow-up musculoskeletal pains were inversely correlated with the posture/equipment program performance. Follow-up neck pain was negatively correlated with good shoulder and elbow postures. Several shoulder muscles originate from the neck. Furthermore, shoulder and elbow movements require neck-originating muscles. Therefore, good shoulder and elbow postures appeared to be negatively correlated with neck pain. Follow-up arm/elbow pain was negatively correlated with good elbow and wrist postures. This result explains the site-specific effect of the equipment/posture program. Follow-up leg/foot pain was negatively correlated with monitor position, and shoulder, elbow, wrist, leg, and foot postures. The negative correlation of leg/foot pain with good performance in leg and foot postures can be easily explained. The human body is well-organized from the head to the foot in an upright position. Forcing, pushing, pulling, and torque of the shoulder, wrist, and elbow can transmit pressure or force to the back, legs, and feet during endoscopy. Therefore, good performance of the monitor position, and shoulder, elbow, and wrist postures also appear to be negatively correlated with leg/foot pain.

We further analyzed the aggravation or improvement factors of musculoskeletal pain. Factors associated with improvement in overall musculoskeletal pain were lower number of follow-up upper endoscopy cases (53% vs 92%), higher performance score of the scope position (86.8% vs 71.3%), proper table height (94.1% vs 79.1%), and decrease in the number of patients admitted (decrease of 12 patients per 2 weeks vs increase of 3.6 patients per 2 weeks). A factor associated with aggravation of follow-up musculoskeletal pain was increase in the colonoscopy number (increase of 6.27 cases per 2 weeks in the aggravation group and no change in the improvement group). However, stretching performance did not differ between the improvement and aggravation groups. The scores for stretching satisfaction and willingness to continue the program were higher in the improvement group than in the aggravation group. These results suggest that both reduction in the number of endoscopy cases and admitted patients and good equipment/posture performance are needed to improve the musculoskeletal pain experienced by GI endoscopists.

This study had several strengths. First, we developed a posture/equipment correction and stretching program based on previous studies. This program was easily applicable, and most participants reported being satisfied with the program. Second, we revealed that various hospital workloads at baseline were significantly associated with multisite musculoskeletal pain, whereas only a few workloads were correlated with musculoskeletal pain after the rehabilitation program. Third, we found a negative correlation between good posture/equipment performance and multisite musculoskeletal pain. Fourth, we also showed improvement or aggravation factors for musculoskeletal pain in endoscopists. Nevertheless, this study had several limitations. First, it was performed during the early days of the coronavirus disease pandemic in the spring-summer of 2020. Thus, workloads during the study period may not reflect usual medical practice. Second, the habits of each endoscopist, which may affect musculoskeletal pain (e.g., preference for sitting or standing during the procedure), were not considered in the analysis. However, we aimed to develop an easily applicable rehabilitation program, not a personalized one, in specific conditions. Third, there are many factors that affect the improvement and aggravation of the musculoskeletal pain. Especially in the case of therapeutic endoscopy, there can be a significant difference in muscle pain after the procedure depending on the site, location, and difficulty of the procedure. It would have been difficult to include all these factors in this study. Future study needs to investigate the factors affecting musculoskeletal pain in therapeutic endoscopists. Last, we did not measure the sports activity such as golf, swimming, yoga, tennis, and others which may exacerbate or relieve the musculoskeletal symptoms. We also did not measure the sleep quality such as sleep length and depth may help or hinder musculoskeletal recovery.

In conclusion, various workloads were correlated with musculoskeletal pain at baseline. After the posture/equipment correction and stretching program, only a few workloads were correlated with follow-up musculoskeletal pain. Follow-up musculoskeletal pain was negatively correlated with equipment/posture program performance. Higher performance of equipment/posture was associated with improvement of musculoskeletal pain, whereas an increase in the number of colonoscopy performed was associated with aggravation of follow-up musculoskeletal pain. Most participants reported satisfaction with the stretching program and willingness to continue using the program. Therefore, our rehabilitation program is easily applicable, satisfactory, and helpful in improving and preventing musculoskeletal pain in GI endoscopists.

This research was supported by the Support Program for Women in Science, Engineering and Technology through the Center for Women in Science, Engineering and Technology (WISET) funded by the Ministry of Science and ICT (No. WISET202003GI01) and the Korean Women Medical Association. We thank the participants and the Diversity Committee of the Korean Society of Gastroenterology.

No potential conflict of interest relevant to this article was reported.

Study concept and design: S.Y.N., K.N., S.Y., K.N.S., S.A.J. Data acquisition: S.Y.N., K.N., C.H.T., K.N.S., S.A.J., N.K., S.M.P., Y.S.P., S.J.P. Data analysis and interpretation: S.Y.N., K.N., C.H.T., J.J. Drafting of the manuscript: S.Y.N., K.N., S.Y. Critical revision of the manuscript for important intellectual content: N.K., S.M.P., Y.S.P., S.J.P. Statistical analysis: S.Y.N., J.J. Obtained funding: K.N.S., S.A.J. Administrative, technical, or material support; study supervision: K.N.S., S.A.J. Approval of final manuscript: all authors.

Supplementary materials can be accessed at https://doi.org/10.5009/gnl220103.