Background

Endoscopists frequently develop musculoskeletal pain and are at risk of a number of occupational injuries, such as De Quervain’s tenosynovitis and carpal tunnel syndrome. Prior studies have reported the prevalence of musculoskeletal pain or injuries ranges from 29 to 89% among endoscopists in general [1, 2]. The types and frequency of pain symptoms include low back pain (6–27%), thumb pain (5–19%), shoulder pain (9–32%), elbow pain (8–15%), hand pain (9–17%), neck pain (9–28%), and hand numbness (12%) [1,2,3,4,5,6,7].

However, scant information is available regarding the development of musculoskeletal pain and injuries in advanced endoscopists performing endoscopic retrograde cholangiopancreatography (ERCP). Factors such as leaded protective gear, differences in patient and endoscopist positioning compared to standard endoscopic procedures, and longer procedural times likely place endoscopists who perform ERCPs at an increased risk of injuries in comparison with other endoscopists. A single study focused solely on endoscopists performing ERCP has been published previously–nearly two decades ago. In 2002, O’Sullivan and colleagues obtained examined survey responses from 114 endoscopists in Canada. Sixty-seven percent of respondents reported at least one musculoskeletal complaint, and 58% reported two or more complaints; 74% attributed their symptoms to endoscopy and/or ERCP, and 79% reported that their condition was aggravated by performing ERCP. The most frequently reported pain symptoms were back pain (57%), neck pain (46%), and hand pain (33%) [5].

Given the marked changes in the performance of ERCP in the last 20 years and the increasingly complex procedures, it is important to assess the current prevalence of musculoskeletal pain and injuries associated with ERCP and attempt to identify risk factors that may help direct future device and protective equipment development.

Methods

Physicians were identified using a pre-existing email list of American Society of Gastrointestinal Endoscopy members throughout the USA. An electronic survey containing 23 questions was sent to a pre-existing list of 3276 gastroenterologists who were members of ASGE. These survey questions were not validated. Only providers that performed ERCPs were asked to respond. Responses to the survey were kept anonymous.

The survey was designed to collect information about the respondents and their endoscopic practice, including variables considered potential factors in the development of musculoskeletal pain and injury, as well as their musculoskeletal pain symptoms and injuries. In addition, the survey questions of O’Sullivan and colleagues [5] were reviewed in the development of our survey questions. This draft survey was then reviewed and completed by three advanced endoscopists and one general gastroenterologist and revised based on their input.

The survey included a list of potential musculoskeletal pain symptoms or injuries. Participants could select only one musculoskeletal pain symptom and one musculoskeletal injury, thus enabling us to evaluate the predominant symptom and/or injury for each participant. They were also asked their sex, length of time performing ERCPs, glove size, average number of ERCPs performed yearly, average number of non-ERCP procedures performed yearly, need for treatment of symptoms and/or injuries, the presence of fluoroscopy tables with adjustable heights, the presence of anti-fatigue matting, whether they wore a one- or two-piece lead gown/apron, frequency of removal of lead gown/apron, time period performing ERCPs prior to symptom onset, predominant positioning of patients for ERCPs, the presence of adjustable monitor heights, frequency of assisting in lifting and moving of patients, and lead apron thickness. Respondents also were asked if they had received any education/training on ergonomics in endoscopy, and if so, whether those principles had been incorporated into their practice.

This was a cross-sectional, questionnaire-based descriptive study. The predefined analysis was to be a comparison of the proportion of respondents with pain in those with and without individual baseline characteristics and the proportion of respondents with injuries, with and without individual baseline characteristics. However, because almost all respondents reported pain, we were unable to meaningfully compare the proportion with pain related to baseline characteristics. We therefore performed a post hoc analysis using the outcome of musculoskeletal pain attributed to performing ERCPs in place of the overall pain outcome. In addition, due to the high rate of missing data regarding musculoskeletal injuries, we did not perform a comparison of the proportion with injuries related to baseline characteristics. Univariate analysis was performed using Chi-square test.

Results

A total of 203 (6.2%) gastroenterologists participated in the survey. A summary of demographic and procedure-related characteristics can be found in Table 1. Among the respondents, 184 (91%) reported a musculoskeletal pain symptom with 84 (46%) attributing this pain to performing ERCPs. Ninety-seven (48%) respondents reported a musculoskeletal injury, with 31 (32%) attributing these injuries to performing ERCPs. The most prevalent pain symptoms were neck pain (n = 49, 24%) and lower back pain (n = 34, 17%) (Fig. 1). The most prevalent musculoskeletal injuries were De Quervain’s tenosynovitis (n = 32, 16%) and cervical radiculopathy (n = 25, 12%) (Fig. 2).

Table 1 Demographic and procedure-related characteristics
Full size table
Fig. 1
figure 1

Pain symptoms

Full size image
Fig. 2
figure 2

Injuries

Full size image

Among the respondents, only 73 (36%) used anti-fatigue matting, 83 (40%) wore two-piece lead gowns, and 147 (72%) were not aware of the thickness of their lead gown. Only 50 (25%) participants had received any education/training on ergonomics in endoscopy, and 145 (71%) stated they are interested in learning more about preventative strategies regarding ERCP-related injuries.

Table 2 presents the comparisons of pain attributed to performing ERCPs related to procedure-related characteristics. Respondents who performed fewer ERCPs tended to be less likely to have pain attributed to ERCP, especially when they performed ≤ 50 ERCPs: Only about one-quarter performing this number of ERCPs attributed pain to ERCPs, while just over half of respondents performing > 100 ERCPs per year attributed to ERCPs (Table 2).

Table 2 Pain symptoms attributed to performing ERCPs
Full size table

Discussion

This survey of US endoscopists performing ERCPs found that approximately 90% suffer from a musculoskeletal pain symptom, with almost half attributing their pain to performing ERCP. In addition, almost half of endoscopists performing ERCP have a musculoskeletal condition such as De Quervain’s tenosynovitis or cervical radiculopathy. As might be expected, those respondents performing a greater number of ERCPs were more likely to report musculoskeletal pain they attributed to ERCPs. This relationship appeared to plateau at around 50% once an endoscopist exceeded 100 ERCPs annually, without further increase with increasing numbers of ERCPs performed.

The frequent reports of musculoskeletal pain symptoms and conditions support the need for improvements in education regarding ergonomic factors in the performance of ERCP. In addition, identification of modifiable risk factors is of importance to assist in the development of preventative strategies. We were unable to identify such factors, other than number of ERCPs performed. Further studies, including prospective assessments, are warranted to characterize the causes of musculoskeletal problems of endoscopists performing ERCPs.

We found that nearly three-quarters of endoscopists are interested in learning more about preventative strategies regarding ERCP-related injuries. As we acquire more knowledge of risk factors for musculoskeletal pain symptoms and injuries, it will help direct preventative measures. These preventative measures can then be used to formulate educational material for both endoscopists and trainees. It is important that this knowledge is introduced early in training as positional and technical habits develop quickly. Thus, our hope is that fellowship programs focus on critiquing of ergonomics as well as trainee competency.

Future studies should also investigate ergonomic measurements in endoscopists who perform ERCP. A study in which right-thumb pinch force and bilateral forearm muscle activity while performing colonoscopies were measured found that the highest mean right-thumb peak pinch forces exceeded the injury threshold, and the activity of the left abductor pollicis longus, left extensor carpi radialis, and right extensor carpi radialis exceeded the American Conference of Industrial Hygienists (ACGIH) hand activity level (HAL) action limit [8]. It would be useful to perform similar measurements during ERCPs as this could also direct future device designs to limit forces sustained by advanced endoscopists.

One limitation of the study was the low survey response rate (6.2%). However, a low response rate was expected as this survey was distributed broadly to ASGE members, but only providers who performed ERCP were asked to respond. For future studies, having a database of only providers who perform ERCP would be invaluable as it would allow us to focus survey distribution to only eligible participants. Although this study is the largest ergonomics study to date focused on providers who perform ERCP, the sample size was still relatively small. Given the high prevalence of musculoskeletal pain symptoms, future studies will need to involve more participants in order to attain enough asymptomatic endoscopists for comparative analysis. Given this survey was focused on musculoskeletal pain symptoms and injuries, it is possible bias could have skewed the results as endoscopists with injuries and/or pain symptoms may have been more likely to participate in the survey. Another limitation of this study was that participants were only able to select one pain symptom and one musculoskeletal injury. Thus, despite the strikingly high prevalence of these two outcomes, the true prevalence is likely much higher. In addition, half of the respondents skipped the question regarding musculoskeletal injuries; because we did not have an option to enter other injuries, we cannot be sure if these individuals did or did not have other musculoskeletal injuries, thus limiting our assessment of total musculoskeletal conditions.