Participants were 223 self-identified oboists recruited from oboe and double-reed Facebook groups (primarily English-speaking) and from emails to oboe professors at North American universities. The average age of the sample was 41.1 (Range = 18 - 81; SD = 16.50) (see Figure 1). Three respondents did not specify age. A large majority of respondents identified their gender as female (70.0%, n = 156). Other responses for gender were male (28.1%, n = 65), non-binary (0.4%, n = 1), and other (0.4%, n = 1, who specified in the comments being a trans man). Participants were diverse in professional status: 39 students (17.5%), 61 amateurs (27.4%), and 123 professional (55.2%). Participants had an average of 27.69 years of experience on oboe (Range = 0 - 69; SD = 16.96). The average age at which they started oboe was 13.37 (range = 3 - 62; SD = 7.04). The weekly time spent playing the oboe (see Figure 2) was reported as 1-3 h (13.6%, n = 30), 4-6 h (13.1%, n = 29), 7-9 h (16.7%, n = 37), 10-12 h (15.8%, n = 35), 13-15 h (10.9%, n = 24), 16-18 h (7.2%, n = 16), and 19+ h (22.6%, n = 50). 2 respondents did not specify weekly playing time.

The 10-question survey was conducted online using the SurveyMonkey platform. The survey took an estimated 5-7 min to complete. Three existing questionnaires were consulted in the making of this survey (Kuorinka et al., 1987; Ackermann and Driscoll, 2010; Lamontagne and Belanger, 2012). Ultimately, most questions were adapted from Ackermann and Driscoll (2010) because of its specific tailoring to musicians and the applicability of its questions to our research questions. The first 5 questions collected demographic data (age, gender, oboe start age, career level, weekly playing time), and the latter 5 questions asked about respondents’ experiences with injury and preventive methods. As in Ackermann & Driscoll’s questionnaire, a playing-related musculoskeletal disorder (PRMD) was defined as “any pain, weakness, numbness, tingling, or other symptoms that interfere with your ability to play your instrument at the level to which you are accustomed,” which itself was based on the definition used by Zaza and Farewell (1997). Respondents were asked to identify any PRMDs they had experienced, from a list of 22 specified body parts plus “other (please specify),” and to rate the severity of the injury on a scale of 1 (mild) to 10 (extremely severe). The list of injury locations was adapted from two surveys (Kuorinka et al., 1987; Ackermann and Driscoll, 2010). Two main changes were made to the lists: since the majority of PRMDs reported to date among woodwind players occur in the upper extremities (Nemoto and Arino, 2007), the list of upper extremity injury locations would ideally be more nuanced than that in the general purpose Nordic Musculoskeletal Questionnaire (NMQ; Kuorinka et al., 1987). Therefore, instead of broader categories such as “wrist/hand,” the upper extremities were subdivided into fingers, thumb, hand, wrist, forearm, elbow, upper arm, and shoulder as in Ackermann and Driscoll (2010). Furthermore, since injuries in the lower body (feet/ankles, legs, knees) are unlikely to be caused by or to have a strong effect on playing a wind instrument, these injury locations were omitted and replaced with “other (please specify).” The 1 to 10 or 0 to 10 scale for rating pain and effectiveness has been used previously in validated questionnaires (Ackermann and Driscoll, 2010; Lamontagne and Belanger, 2012). This scale was adopted based on its ability to gather nuanced data on the severity of pain and perceived effectiveness of treatments. The 0 to 10 scale has the added benefit of a definitive 0 point (relevant only for questions that have a definitive 0 answer) and is simple for most people to conceptualize, as similar scales are used in everyday life (e.g., percentage points). For each injury, they were asked to describe the pain using checkboxes (multiple selections allowed) for descriptors “tingling,” “numbness,” “pain,” “weakness,” “cramps,” “involuntary movement,” and “other.” Respondents were asked whether their PRMD had been diagnosed by a healthcare professional, and if so, to specify what the diagnosis was. They were also asked to what extent they had recovered from their PRMD(s) between a choice of “No, not recovered at all,” “Somewhat recovered,” “Mostly recovered,” and “Yes, completely recovered.” A final question asked respondents to rate the effectiveness of various remedial and preventive solutions that they had experience with (multiple selections allowed), on a scale of 0 (no effect) to 10 (greatest effect of all). The list of treatments was adapted from Ackermann and Driscoll (2010). Additional candidate treatments were added to the list based on consultation with numerous oboists and musicians’ health researchers. In these final five questions, respondents were instructed to skip any rows that discussed a PRMD or preventive measure they did not have experience with.

For quality control, the survey was informally pilot tested by performing arts medicine researchers at the University of Toronto and several professional and student oboists in Canada. Reviewers were asked whether the survey’s questions were clear, whether any information should be added or omitted, and whether, in their opinion, the survey would serve to answer the stated research questions. Changes to the survey prompted by these reviewers included changes to wording to increase accessibility of the survey, and the addition of treatments to the final question.

A copy of the questionnaire can be found in the Supplementary Materials.

Descriptive statistics and associated plots were created in Python using the pandas package (Mckinney, 2010) and in Microsoft Excel, respectively. Statistical analyses (e.g., logistic regression analysis) were conducted in R and in Python using the pandas and statsmodels packages (Seabold and Perktold, 2010). Statistical analyses were performed in regression contexts to assess risk factors associated with PRMDs both categorically (0 = no history of PRMD, 1 = history of PRMD, with models run in the logistic regression framework) and dimensionally (severity of worst PRMD on scale of 1 – 10 analyzed as a continuous outcome utilizing only individuals reporting at least one PRMD, with models run in an ordinary least squares framework). We were interested in five risk factors motivated by literature review (gender, hours of playing, career level, age, years of playing experience). No correction for multiple tests across these five factors was performed. A significance level of p(alpha) < 0.05 was selected for reporting. Additionally, results that survive Bonferroni correction for conducting two analyses within each risk factor (the categorical prevalence model and the two dimensional severity model) are marked in-text with an asterisk (*: p < 0.025). Subjects with missing age data were excluded from models that included the age variable. For models that included the gender variable, two gender groups that contained only a single observation were excluded (“non-binary” and “Other”).

Of all respondents, 74.9% (167/223) reported having at least one PRMD in their lifetime. Seventy-nine (47.3%) of these 167 reported having had a diagnosis from a medical professional, while 84 (50.3%) had not received a diagnosis. The remaining 4 (2.4%) individuals were either among the 8 who selected “not applicable,” or they did not respond to this question.

A majority of respondents reported having had severe injuries, with 52.0% (n = 116) of all respondents reporting at least one injury rated between 7 and 10, and 61.9% (n = 138) reporting at least one injury rated 5 or higher (see Figure 3).

A majority of injuries were on the right side of the body. The most common PRMD location was the right thumb (61.9%), followed by the right wrist (48.4%), right hand (47.8%), and right forearm (47.5%). The most common injuries also tended, on average, to be the most severe: the average severity rating was highest in the right thumb (5.77), right wrist (5.59), right hand (5.23), and right forearm (4.96). See Figures 4, 5 for a complete list of prevalence and average severity by location.

Pain was the most common descriptor for all injury locations except for respiratory system (weakness) and right and left fingers (tingling). See Figure 6 for complete sensation profiles for all injury locations.

Not all respondents reporting an injury answered the question on recovery. For unknown reasons, a surprising number of participants chose not to disclose recovery status. For example, of 138 people reporting a right thumb injury, only 121 disclosed recovery status. Out of a total of 1,545 injuries reported, recovery data for 1,072 of these injuries was provided (for a response rate on the recovery question of 69.4%). This caveat being said, of those who did respond to this question only 280 of these 1,072 injuries (26.1%) were reported as being fully recovered (see Table 1).

We were interested in a number of variables previously linked to PRMD prevalence and severity (see Introduction). In the following sections, we describe the results of regression analyses quantifying the effects of gender, playing hours, career level, age, and years of playing experience on PRMD prevalence and severity.

The treatment methods rated as most effective varied from individual to individual. As not all respondents had tried every method presented in question #10, there was a great deal of variability in the response rates for each method (e.g., 156 respondents evaluated “rest,” while only 13 evaluated “Feldenkrais”). Rest, stretching, exercise, massage, ice, and pain killers had the highest response rates, while Feldenkrais, occupational therapy, acupuncture, chiropractic, medical, and Alexander Technique had the lowest response rates. Despite this variability in response rates, some trends can be discerned. The mean rankings, on a scale of 0 to 10, of physical therapy (6.69), rest (6.37), stretching (6.22), occupational therapy (6.05), and massage (5.94), were higher than most. It is worth noting, however, that for every method there was a very wide variability in perceived effectiveness, and it is clear that there is no single universally effective method (see Figure 9).

Several respondents suggested treatment methods that were not listed in the survey (Table 4). The most common of these was instrument modifications, such as pegs and neck straps that help take weight of the oboe off of the thumb. Body mapping (an awareness technique related to Alexander Technique) and technique or hand position corrections were also mentioned multiple times. Most of these alternative methods were rated very highly, but a few (such as change of activity and marijuana) rated below 5.

Of the 223 survey respondents, 167 (74.9%) reported having had a PRMD in their lifetime. Though there is a wide range of prevalence rates in the literature, this is often attributed to the wide variety of definitions used for PRMD (e.g., Berque et al., 2016). Our result falls in line with results of comparable studies that also used Zaza and Farewell’s definition of PRMD (1997). These studies similarly reported a lifetime PRMD frequency among musicians of between 63 and 78% (Baadjou et al., 2015; Berque et al., 2016; Ballenberger et al., 2018; Kok et al., 2018; Shanoff et al., 2019). Our result is also in the range of oboe-specific PRMD rates previously determined by smaller studies: 75% (with a sample size of 12 oboists), 72% (with a sample size of 60), and 85.7% (with a sample size of 28) (Thrasher and Chesky, 2001; Nemoto and Arino, 2007; Stanek et al., 2017).

By gathering data on injury severity, we were also able to achieve a more nuanced understanding of the degree to which oboists are incapacitated by their injuries. A majority of respondents had experienced severe injuries: 52.0% of all respondents reported at least one injury rated between 7 and 10 out of 10, and 61.9% reported at least one injury rated 5 or higher.

The most common injuries were found to be on the right side of the body, which confirms the findings of previous studies on oboe players (Thrasher and Chesky, 2001; Nemoto and Arino, 2007). The injuries that were the most common (right thumb, right wrist, right hand, and right forearm) also tended to be the most severe. The increased frequency and severity of injuries on the right side of the body can be attributed to the asymmetrical playing posture of oboists, where the right thumb is statically loaded by the weight of the instrument.

As not every respondent had experience with all of the treatment methods listed, there was a great deal of variability in the response rate for each method. This response rate variability makes it difficult to meaningfully interpret much of the data. For example, only 13 respondents indicated having had experience with Feldenkrais, while 156 were able to rank the effectiveness of rest. Even for those methods that had higher sample sizes, there was a great deal of variability in perceived effectiveness rankings. There is no strong evidence of any particular therapy being universally beneficial for everyone, as in almost every case there was a group of respondents ranking the effectiveness of the therapy as “0 (no effect).” It is highly likely that the effectiveness of particular methods at treating PRMD is largely dependent on the individual and the particular characteristics of the PRMD at hand. It is worth noting that physical therapy and occupational therapy were among the highest-ranking, and their relative success at alleviating PRMD may be attributed to the more targeted nature of these methods.

Based on the written-in responses in the “other: please specify” section, future studies on treatment method efficacy should consider investigating options such as instrument modifications (e.g., neck strap, instrument pegs), changes to playing technique, and body mapping.

In the literature, there are mixed results in studies exploring the effects of exercise on PRMD prevalence. Some studies have found no association between exercise and PRMD (Brandfonbrener, 1997; Kreutz et al., 2008; Baadjou et al., 2015). Other studies have found increased exercise is associated with less risk of PRMDs (Roach et al., 1994; Yeung et al., 1999; Ranelli et al., 2011). In one study, increased exercise was associated with higher rates of knee pain (Zetterberg et al., 1998). Three studies examined the effect of targeted exercise programs on the rate of PRMD in a group of musicians, and all reported a reduction of injuries at the end of the program (de Greef et al., 2003; Lopez and Martinez, 2013; Chan et al., 2014). All these exercise studies examined a diverse population, consisting of a mixture of string, woodwind, brass, and other musicians. The effect of posture has been explored with sample of saxophone players, and it was found that a rounded upper back and/or backward pelvic tilt correlated strongly with higher pain ratings, and a rounded back, backward pelvic tilt, and excessive curve in the low back all corresponded with higher rates of PRMD in the right wrist (Shanoff et al., 2019). Since the oboe, like the saxophone, has been found to be prone to PRMDs in the right wrist due to static loading of the right thumb (Nemoto and Arino, 2007), it is likely that postural issues will have a similar effect on PRMDs among oboe players.

An important strength of this study was its ability to reach the largest sample of oboists of any comparable study (n = 223). This was aided by the strong online presence of communities of oboists and double reed players. The survey was filled out on a volunteer basis, rather than randomized, and so it is possible that this lack of randomization affected the results, since respondents who have dealt with injuries in the past may be more willing to participate. This study was able to gather valuable information on the perceived effectiveness of various PRMD treatments. However, since the effectiveness of these treatments appears to vary so much based on the individual and the type of injury at hand, it would be advisable for future studies to delve deeper into this issue.