Work-Related Musculoskeletal Disorders in Physical Therapists: Prevalence, Severity, Risks, and Responses.

Physical therapy practice can lead to work-related musculoskeletal disorders (WMSDs) in physical therapists. We know little, however, about the range of problems, their severity, or the implications for affected therapists. Existing studies have focused on back pain,[1-4] but that underestimates the range of problems that may develop. Only one study[5] we found recognized and investigated other areas in which WMSDs might develop as a consequence of physical therapy practice. Beyond that, there are many questions.

This study was designed to investigate unanswered questions about physical therapists and WMSDs and to provide a basis for further work expected to lead to developing preventive strategies. The aim of this study was to investigate the following:

* Distribution, prevalence, and severity of WMSDs.

* Associations among specialty areas, tasks, risk factors, and WMSDs.

* Strategies used by physical therapists to minimize the effects and risks of developing WMSDs.

* Responses of physical therapists who developed WMSDs.

Distribution, Prevalence, and Severity of WMSDs

A focus on back pain among physical therapists is consistent with expectations of prevalence of symptoms among health care workers.[6-10] Of the studies of back pain in physical therapists, one study[2] showed a 29% prevalence of work-related low back pain (LBP). Most disturbingly, younger therapists had the highest prevalence. The initial onset was most commonly within the first 4 years of experience, and most initial episodes occurred in acute care or rehabilitation settings. In other studies of physical therapists, the researchers found an annual prevalence of back pain of 38%[1] and a prevalence of work-related LBP of 49.2%.[3] A problem with these studies is that they used different definitions of back pain, limiting the opportunities for direct comparisons.

The most comprehensive study of WMSDs in physical therapists investigated their prevalence in 9 different body areas.[5] The highest annual prevalence of WMSDs was in the low back (45%), followed by the wrists and hands (29.6%), upper back (28.7%), neck (24.7%), and shoulders, elbows, hips and thighs, knees, and ankles and feet (each less than 20%). More female therapists than male therapists had spinal symptoms and wrist and hand symptoms. Although comprehensive, this study did not consider thumbs separately. Two unpublished studies[11,12] and practitioner anecdotes suggest that therapists using particular mobilization and manipulation techniques are prone to developing WMSDs in their thumbs. This finding suggests the importance of investigating the prevalence of thumb symptoms, given the popularity of relevant techniques in current physical therapy practice.

Severity of symptoms has not been addressed in the literature. Inferences can be made from therapist responses to WMSDs. The percentage of physical therapists who sought treatment from a physician was reported in 3 studies.[1,2,5] The percentage of therapists who reported limitations due to LBP was reported in 1 study,[3] and the percentage of therapists who took time off from work was reported in another study.[5] Although these findings may indicate the severity of at least some of the WMSDs that physical therapists experience, more information is needed.

Specialty Areas, Tasks, Risk Factors, and Development of WMSDs

The specialty area of practice or job setting is thought to be a risk factor for WMSDs in physical therapists.[2,3] The underlying assumption apparently is that a particular specialty area has inherent risks because practitioners use a limited number of techniques. Although this may be partly true, mode of practice and clientele may vary considerably within a specialty area, altering the risk factors for injury. Bark et al[5] related specialty areas to WMSDs and found that therapists working in hospitals had a higher prevalence of LBP and ankle and foot symptoms than those working in other settings. They also investigated the relationship between tasks and symptoms and found that manual therapy was related to wrist and hand symptoms and elbow symptoms and that neurologic rehabilitation was related to LBP and upper back and knee pain. Clearly, more information is needed, if only to provide the basis for possible preventive strategies. This means considering both specialty areas and the specific tasks relating to an area when determining risks involved.

Risk Factors

In a study by Bark et al,[5] therapists selected 17 job-related risk factors for the development of WMSDs and ranked them as problematic on a scale of 0 to 10 (0 represented "no problem," and a score of 8 or higher represented a "major problem"). By implication, individual risk factors contributed to the development of symptoms. An alternative explanation for Bark and colleagues' findings, however, could be that the respondents did not consider the factors as problematic until after the onset of symptoms. No attempt was made to relate any risk factors to particular WMSDs.

The set of 17 job-related risk factors[5] fall into 4 broad areas: (1) activities (6 risk factors pertaining to specific activities), (2) postural factors (4 risk factors relating to the working posture or position of the physical therapist), (3) workload issues (4 risk factors relating to the frequency or repetitiveness of treatment and time management issues such as scheduling and rest breaks), and (4) personal factors (3 risk factors pertaining to work in relation to the physical work capacity, state of health, and knowledge of the physical therapist). Given that risk factors may be related to specialty areas or tasks, we believe there is a need to investigate these risk factors in relation to WMSDs.

Strategies Used by Physical Therapists to Minimize Effects and Risks of Developing WMSDs

Strategies used by physical therapists to avoid the development of WMSDs include the use of aids and equipment and the use of what we call "self-protective behaviors." Aids and equipment include height-adjustable beds, lifting belts, slide boards, splints, and stools on casters ("wheelie stools"), which are available to health care personnel to reduce the physical demands of their work.[13-17] Physical therapists may also use self-protective behaviors. We identified 10 behaviors from the literature and from our discussions with physical therapists as self-protective behaviors and categorized them as either outsourcing, preventive, or reactive strategies.

Outsourcing strategies shift all or part of the therapist's workload to another person. These strategies include reducing the load by obtaining help when transferring patients, a safer strategy for the therapist than lifting alone.[18] Similarly, a therapist can use physical therapist assistants.

Preventive strategies are meant to alter the technique or the environment to avoid placing stress on the therapist's body. Postures used during work can be related to the presence of LBP.[19] Strategies to modify a therapist's position or to adjust bed height to prevent injury are in this category. Another possible preventive strategy is to use pauses and changes in posture to reduce the risk of injury, as well as warming up before performing a technique.[20]

Reactive strategies are those developed by a therapist in response to injury (or perceived risk of injury). These strategies include actions that help avoid aggravating factors. For example, using a different part of the body to administer a manual technique can be a way of protecting the upper limb from overuse,[21,22] as is substituting electrotherapy for some manual techniques.

Responses of Physical Therapists Who Develop WMSDs

Responses to WMSDs may include adopting one or more self-protective behaviors or using aids and equipment. In addition, physical therapists may seek treatment, modify activities of daily living (ADL) and leisure (lifestyle), or make changes to their specialty area, either within the profession or by leaving altogether, as a consequence of WMSDs.

Physical therapists' responses to WMSDs include taking time off from work,[1-3,5] modifying leisure and ADL to relieve symptoms,[3] seeking treatment from a health care practitioner,[1-3,5] lodging a workers' compensation claim,[2-4] and continuing to work with discomfort.[5] This latter option may include modifying treatment techniques or choosing alternative methods of treatment to reduce the strain on the affected body part.

The different types of responses can be divided into 2 categories: those responses aimed at reducing effects of stress on the body as much as possible and those responses designed to avoid future exposure. The effects of strain may be reduced by receiving treatment, modifying lifestyle, or removal from stressors. Sick leave and workers' compensation are 2 ways of reducing the effects of strain. Some investigators[2,3] have reported that 5.2% to 10.3% of therapists took sick leave for LBP and that 3.4% to 3.5% of therapists lodged workers' compensation claims for LBP (percentages calculated from data provided by the investigators). Bark et al[5] reported that 2.8% of therapists missed work because of LBP, but it was unclear whether their respondents used sick leave, some other kind of leave, or workers' compensation. Bark et al considered 9 body areas, and they also described treatment. These findings suggest that responses to be considered include treatment, modification of ADL and leisure (lifestyle), and interference with work. The findings also suggest the need to include responses of physical therapists to WMSDs in all body areas.

The category of responses designed to avoid future exposure is the category we call "leaving and moving." Molumphy et al[2] reported that 18% of therapists with LBP changed their work setting and 12% of therapists with LBP reduced their hours of patient contact time, but that none left the profession as a consequence of WMSDs. Instead, therapists tended to move from acute care and rehabilitation settings to settings where patients needed less acute care. Bark et al[5] reported that 25% of therapists had to change their work activities as a consequence of WMSDs, most frequently by changing techniques, work postures, or body mechanics.

The aims of our study extended beyond documenting the prevalence, distribution, and severity of WMSDs to include investigating the relationships among risk factors, specialty areas, and WMSDs; the strategies used to minimize the effects and risks of developing WMSDs; and the responses of therapists to WMSDs.

Method

This study was part of a larger study that investigated the musculoskeletal, reproductive, and general health of physical therapists. We report on only the musculoskeletal part of the study in this article.

A sample of 824 therapists was chosen by taking every fourth therapist, after a randomly selected starting point, from a list of all physical therapists (N=3,296) who were registered in the state of Victoria and resident in Australia. This list comprises the compulsory state register of all physical therapists.

Distribution, Prevalence, and Severity of WMSDs

Distribution. A questionnaire was mailed to 824 physical therapists, with a letter of explanation and a postage-paid return envelope. Respondents were advised that return of the completed questionnaire constituted informed consent. All questionnaires were given numbers corresponding to names on a master list, to allow follow-up of nonrespondents. To ensure anonymity of respondents, a numbered section was returned separately, opened independently, and respondents' names removed from the master list. A reminder letter was sent to nonrespondents after 2 weeks requesting return of the completed questionnaires, and a second reminder letter was sent to nonrespondents after a further 3 weeks. A sample of 1 in 10 nonrespondents was telephoned to determine their characteristics.

Prevalence. Musculoskeletal symptoms were investigated using a self-administered, purpose-designed questionnaire. Questions were based on the standardized Nordic Questionnaire, which is used to record work-related musculoskeletal symptoms in working populations.[23] Respondents were asked to indicate the number of hours per week they had spent during the previous 12 months performing various tasks as part of their therapy practices.

Respondents were asked whether they had ever experienced work-related pain or discomfort. To determine the 12-month prevalence of symptoms (defined as "job-related ache, pain, etc"), they were asked whether they had experienced work-related symptoms in the past 12 months in 10 different anatomical areas. These were the same anatomical areas as those reported by Bark et al,[5] with the addition of the thumbs. In addition to questions regarding the prevalence of work-related musculoskeletal symptoms, the therapists were asked to indicate whether symptoms in each anatomical area had interfered with work, ADL, or leisure; whether symptoms had lasted more than 3 days; and whether they had sought treatment in the preceding 12 months.

Severity. Severity of symptoms was addressed by asking subjects with symptoms to record whether the symptoms had prevented them from working, had prevented them from performing normal ADL or leisure activities, required treatment from a health care professional, or lasted more than 3 days. A point was given if work-related discomfort was present, and an additional point was given for each relevant applicable descriptor selected. The minimum score was 1, indicating the presence of symptoms, and the maximum score was 5, indicating that all 4 descriptors applied. The total number of points given was recorded as the severity score, and a score of 3 or higher was viewed as moderately severe WMSD.

Specialty Areas, Tasks, Risk Factors, and Development of WMSDs

The 17 job-related risk factors identified by Bark et aP were included in the survey. To extend our understanding of the impact of risk on physical therapists and to allow some insight into causality, we asked the therapists to indicate the degree to which they believed each factor contributed to their work-related discomfort or injury. Respondents indicated whether the risk factor was a minor, moderate, or major contributor to their discomfort or injury.

Strategies Used by Physical Therapists in an Effort to Minimize Effects and Risks of Developing WMSDs

The use of self-protective strategies to reduce the risk or perceived risk of injury was investigated. Respondents indicated the aids and equipment they used while practicing physical therapy. They indicated which of the 10 self-protective behaviors (previously described) were relevant to their work and, if relevant, the frequency with which they practiced the behavior, using a 5-point Likert scale ranging from "almost always" to "almost never."

Responses of Physical Therapists Who Developed WMSDs

Physical therapists who reported ever having a work-related episode of pain or discomfort were asked which of the following they had ever done: lodged a workers' compensation claim, taken sick leave, continued working with discomfort, or taken other action. Additional questions asked whether they had changed or modified their treatments or their area of practice as a result of work-related discomfort. They were asked whether they had changed specialty area or left the physical therapy profession as a consequence of their WMSDs. Those therapists who were no longer practicing were asked to indicate why they had left the profession.

Data Analysis

Data were analyzed using SPSS 7.0 for Windows.(*) We calculated the 12-month prevalence of symptoms in each of the 10 anatomical areas and the percentage of all therapists who reported that symptoms had prevented them from working or had prevented normal ADL or leisure activities, that they had sought treatment, and that the symptoms lasted more than 3 days. We also calculated the percentage of therapists who had ever experienced work-related pain or discomfort and who took sick leave, lodged a workers' compensation claim, or continued working with discomfort. Finally, the percentage of therapists who had changed specialty area or left the profession because of WMSDs was calculated.

Risk and WMSDs. Respondents who indicated that particular job risks were major contributing factors in the development of their musculoskeletal symptoms or injury were compared with those who did not. Chi-square analyses were used to investigate the relationships between WMSDs and job-related risk factors, tasks, age, sex, and specialty area. Mantel-Haenszel odds ratios (ORs) and upper and lower 95% confidence intervals (CIs) were calculated to estimate the relative risks.

Responses to WMSDs. The relationship between severity of symptoms and changing specialty area and the use of self-protective behaviors was investigated using chi-square analysis, and Mantel-Haenszel ORs and upper and lower 95% CIs were used to estimate the relative risks.

Results

Questionnaires were sent to 824 therapists. Thirty-five of the therapists reported that they did not currently reside in Australia and, therefore, were not eligible to participate in the study. Of the 789 questionnaires sent to therapists who were eligible to participate, 541 questionnaires were returned. Four questionnaires were eliminated because of incomplete data, and 1 questionnaire was returned too late, giving a response rate of 67.9% (n=536). The sex balance of respondents (118 male [22%] and 418 female [78%]) was comparable to that of the population from which the sample was drawn (25% male and 75% female).[24] The differences between respondents and nonrespondents were not significant ([[chi square].sub.(1)]=0.56, P=.85), and their characteristics are summarized in Table 1.

Table 1. Comparison of Respondents and Nonrespondents

                     Population     Sample
                     (N=3,296)     (n=789)

Male (%)               25.5         25.1
Female (%)             74.5         74.9
Mean age (y)            ...(a)       ...
Musculoskeletal
  symptoms (%)          ...          ...

                    Respondents   Nonrespondents
                      (n=536)         (n-20)

Male (%)               22.0           30.0
Female (%)             78.0           70.0
Mean age (y)           38.0           40.0
Musculoskeletal
  symptoms (%)         91.0           86.4

(a) Ellipsis indicates unknown.

Distribution, Prevalence, and Severity of WMSDs

Four hundred eighty-eight respondents (91%) reported experiencing work-related musculoskeletal pain or discomfort at some time in their working life. For 225 (48%) of these respondents, the most serious work-related problem concerned their low back. Neck symptoms (57 [12.2%]) and upper back symptoms (57 [12.2%]) were the next most common symptoms, followed by those in the thumb (52 [11.0%]).

Table 2 shows the 12-month prevalence of WMSDs, symptoms lasting more than 3 days, and severity scores of 3 or greater. More than 80% of all therapists (n=444 [82.8%]) had musculoskeletal symptoms in at least one part of their body during the 12-month period. The area that most frequently received scores of 3 or higher (moderately severe) on the severity scale was the low back, with 187 respondents (34.9%). One hundred forty-five respondents (27.1%) reported moderately severe work-related neck symptoms. All other body areas had a prevalence of moderately severe symptoms of less than 20%.

Table 2. Percentage of Therapists (n=536) Reporting Musculoskeletal Symptoms, Therapists Reporting Symptoms Lasting More Than 3 Days, and Therapists With a Severity Score of 3 or Greater

                           12-Month        Symptoms Lasting
                         Prevalence (%)      > 3 Days (%)

Low back                      62.5               38.4
Neck                          47.6               29.7
Upper back                    41.0               22.9
Thumbs                        33.6               14.9
Shoulders                     22.9               12.5
Wrists/hands                  21.8               13.2
Elbows                        13.2                8.2
Knees                         11.2                4.3
Hips                           7.3                4.7
Ankes                          7.1                2.8
Percentage of total
 respondents who
 were affected                82.8               63.8

                           Therapists With Severity
                            Score of 3 or More (%)

Low back                           34.9
Neck                               27.1
Upper back                         19.0
Thumbs                              8.7
Shoulders                          10.6
Wrists/hands                        9.9
Elbows                              6.7
Knees                               3.5
Hips                                3.5
Ankes                               3.0
Percentage of total
 respondents who
 were affected                     58.2

Twelve-month prevalence of upper back, low back, and thumb symptoms was inversely related to age (Fig. 1). Analysis showed that younger therapists reported more neck symptoms ([[chi square].sub.(4)]=10.98, P=.027), upper back symptoms ([[chi square].sub.(4)]=15.27, P=-.004), low back symptoms ([[chi square].sub.(4)] = 19.02, P=.001), and thumb symptoms ([[chi square] (4)=20.64, P [is less than] .001) than did older therapists. The relationship between other upper-limb symptoms and age did not reach statistical significance. Knee symptoms were related to increased age ([[chi square].sub.(4)]=12.41, P=.015); other symptoms were not. Figure 2 shows that the first episode of WMSD occurred for the majority of therapists in the first 5 years of practice.

[Figures 1-2 ILLUSTRATION OMITTED]

The prevalence of symptoms was not different between male and female therapists in most areas (Fig. 3). Male therapists had increased odds of reporting neck symptoms (OR=1.9, 95% CI=1.3-2.9), wrist symptoms (OR=2.0, 95% CI=1.3-3.2), and thumb symptoms (OR=2.2, 95% CI=1.5-3.4) in the last year compared with their female colleagues. Male therapists reported using more mobilization and manipulation techniques than did female therapists ([[chi square].sub.(1)] = 12.49, P [is less than] .01).

[Figure 3 ILLUSTRATION OMITTED]

Severity of symptoms was related to the hours per week spent performing some tasks. Therapists who spent more time performing manipulation or mobilization had more severe low back symptoms ([[chi square].sub.(1)]=12.92, P=.005) and thumb symptoms ([[chi square].sub.(1)]=30.38, P [is less than] .001) than did therapists who spent fewer hours performing manipulation or mobilization. Other tasks were not related to severity of symptoms.

Specially Areas, Tasks, Risk Factors, and Development of WMSDs

Table 3 shows the ORs linking WMSDs with particular specialty areas of physical therapy practice. Therapists who had ever worked in private practice, sports physical therapy, or pediatrics had increased odds of reporting WMSDs in the last 12 months. Therapists currently working in private practice reported more neck symptoms ([[chi square].sub.(1)]=8.49, P=.004), upper back symptoms ([[chi square].sub.(1)]=10.28, P=.001), elbow symptoms ([[chi square].sub.(1)]=8.53, P=.003), wrist and hand symptoms ([[chi square].sub.(1)]=16.48, P=.001), and thumb symptoms ([[chi square].sub.(1)]=27.64, P=.001) than did therapists working in other areas. The association between WMSDs and other specialty areas did not reach statistical significance. No relationship was found between low back symptoms and any particular specialty area.

Table 3. Musculoskeletal Symptoms Related to Therapists "Ever" Working in Different Specialty Areas of Physical Therapy

                             Symptoms      Odds Ratio
                             in Last       (95% Confidence
Specialty Area               12 months     Interval)

Sports physical therapy      Thumb          2.9 (1.8-4.8)
                             Neck           2.4 (1.4-3.8)

Private practice             Wrist          4.2 (2.6-6.6)
                             Elbow          2.9 (1.6-4.9)
                             Thumb          2.8 (1.9-3.9)
                             Neck           2.5 (1.7-3.4)
                             Upper back     1.7 (1.2-2.5)

Pediatrics                   Knees          2.3 (1.2-4.2)

Across all specialty areas, only 2 tasks were revealed by chi-square analysis to be related to WMSDs. Table 4 shows that although mobilization and manipulation techniques and other hands-on treatments were associated with increased risk of WMSDs, the ORs obtained for mobilization and manipulation were generally higher than for other hands-on techniques. Four of the 6 increased ORs for mobilization and manipulation were greater than 2.5, whereas none of those for other hands-on techniques exceeded 2.5. Three of the 6 ORs for other hands-on treatments were less than 2.0. Electrotherapy, cardiothoracic (acute and cardiac rehabilitation), neurological (acute and long-term rehabilitation), hydrotherapy, general and outpatient rehabilitation, and education and training and administration tasks were not associated with the presence of WMSDs.

Table 4. Tasks Performed at All During the Last 12 Months and Related to the Presence of Musculoskeletal Symptoms During That Time

                     Symptoms       Odds Ratio
Task Performed       in Last        (95% Confidence
in Last 12 Months    12 Months      Interval)

Mobilization/        Thumbs         7.7 (4.6-12.7)
  manipulation       Wrists         4.0 (2.3-6.8)
                     Elbows         2.8 (1.5-5.3)
                     Neck           2.7 (1.8-3.8)
                     Upper back     2.3 (1.6-3.3)
                     Low back       1.7 (1.2-2.4)

Other "hands-on"     Wrists         2.4 (1.5-3.7)
  treatment          Elbows         2.3 (1.3-4.1)
                     Shoulders      2.1 (1.3-3.3)
                     Upper back     1.9 (1.3-2.7)
                     Low back       1.7 (1.1-2.3)
                     Neck           1.6 (1.1-2.2)

Thumb symptoms in particular were related to the number of hours per week that therapists used mobilization and manipulation techniques. Figure 4 illustrates that the prevalence of thumb symptoms increased as the number of hours of using these techniques increased. The relationship was linear initially, with a prevalence of thumb pain near 60% among therapists who reported using these techniques for more than 20 hours per week.

[Figure 4 ILLUSTRATION OMITTED]

Not all job-related risks were relevant to all therapists. For example, lifting patients was not relevant to therapists working exclusively in administration. Of the 377 therapists to whom the risk was relevant, 203 therapists (53.8%) believed that performing manual orthopedic techniques had contributed to their injury in a major way. Of the 421 therapists to whom performing the same task repeatedly was relevant, 220 therapists (52.3%) responded that this risk factor was a major contributor to their WMSDs. Three hundred fourteen therapists reported that they lifted or transferred patients who were heavy and dependent on therapists for transfer, and 137 therapists (43.6%) believed this was a major contributor to their injury. Of the 403 therapists who treated a large number of patients a day, 167 therapists (41.4%) reported this was a major factor contributing to their WMSDs. Of the 412 therapists who reported working in the same position for long periods, 171 therapists (41.5%) indicated that this risk factor contributed to their work-related symptoms in a major way. Only 15 (3.1%) of all therapists who had experienced WMSDs responded that inadequate training in injury prevention was a major contributing factor in the development of their work-related symptoms.

Table 5 shows that performing manual orthopedic techniques was associated with increased risk of neck symptoms (OR=l.9, 95% CI=1.2-2.8), shoulder symptoms (OR= 1.9, 95% CI= 1.2-3.0), elbow symptoms (OR=3.5, 95% CI=1.9-6.7), wrist and hand symptoms (OR=5.1, 95% CI=3.0-8.6), and thumb symptoms (OR=5.5, 95% CI=3.5-8.6). Lifting or transferring heavy patients was related to increased risk of low back symptoms (OR=2.4, 95% CI=1.4-4.1).

Table 5. Job-Related Risk Factors That Therapists Identified as Major Contributors to Their Work-Related Musculoskeletal Disorders and Their Relationship to Particular Musculoskeletal Symptoms

                   Job-Related Risk Factor      Affected Areas

Activities         Performing manual            Neck, shoulders
                   orthopedic techniques        Elbows, wrist/hands,
                                                thumbs

                   Lifting or transferring      Low back
                   dependent patients

                   Assisting patients during
                   gait activities

                   Carrying, lifting, or
                   moving heavy materials or
                   equipment

                   Working with confused or
                   agitated patients

                   Unanticipated sudden
                   movements or falls by
                   patient

Position/posture   Working in awkward or        Low back
                   cramped positions

                   Working in the same          Upper back
                   position for long periods    Low back, neck

                   Bending or misting in an     Low back
                   awkward way

                   Reaching or working away
                   from your body

Workload issues    Performing the same task     Neck, shoulders
                   over and over                Elbows
                                                Wrists/hands, thumbs

                   Treating a large number of   Thumbs, elbows,
                   patients in one day          shoulders
                                                Neck, wrists/hands

                   Working scheduling           Elbows
                   (overtime, irregular         Shoulders
                   shifts, length of workday)

                   Not enough rest breaks       Neck, shoulders
                   during the day               Upper back, elbows,
                                                wrists/hands

Personal work      Working at or near your      Wrists/hands
factors            physical limits              Elbows

                   Continuing to work when      Neck
                   injured or hurt              Wrists/hands,
                                                shoulders

                   Inadequate training in
                   injury prevention

Postural risk factors were associated with increased risk of spinal symptoms. Working in awkward positions was associated with increased risk of low back symptoms (OR=2.1, 95% CI=1.2-3.9). Working in the same position for long periods was associated with increased risk of upper back symptoms (OR=1.7, 95% CI=1.1-2.5), low back symptoms (OR=2.0, 95% CI=1.3-3.1) and neck symptoms (OR= 1.8, 95% CI= 1.2-2.7). Bending or twisting the back was associated with increased risk of low back symptoms (OR=2.0, 95% CI= 1.04-3.8).

All four workload risk factors were associated with increased risk of WMSDs in up to 5 different body areas. Performing the same task repeatedly was associated with increased risk of neck symptoms (OR=1.6, 95% CI=1.1-2.3), shoulder symptoms (OR=1.7, 95% CI=1.1-2.7), elbow symptoms (OR=2.4, 95% CI=1.4-4.2), wrist and hand symptoms (OR=2.6, 95% CI=1.6-4.1), and thumb symptoms (OR=2.9, 95% CI=2.0-4.4). Therapists who treated a large number of patients in one day had increased risk of thumb symptoms (OR= 1.9, 95% CI=1.3-2.9), elbow symptoms (OR=2.1, 95% CI=1.2-3.7), shoulder symptoms (OR=1.8, 95% CI=1.2-2.9), neck symptoms (OR=2.5, 95% CI=1.6-3.8), and wrist and hand symptoms (OR=3.2, 95% CI=2.0-5.1). Work scheduling issues were associated with increased risk of elbow symptoms (OR=2.2, 95% CI=1.01-4.9) and shoulder symptoms (OR=2.6, 95% CI=1.3-5.2). Not enough rest breaks during the day was associated with an increased risk of neck symptoms (OR=1.8, 95% CI=1.1-2.9), shoulder symptoms (OR=1.8, 95% CI=1.1-3.0), upper back symptoms (OR=2.1, 95% CI=1.3-3.4), elbow symptoms (OR=2.6, 95% CI=1.4-4.7), and wrist and hand symptoms (OR=2.2, 95% CI=1.4-3.8).

Working at or near the therapists' physical limits was associated with increased risk of wrist and hand symptoms (OR=2.2, 95% CI=1.3-3.8). Continuing to work when injured was associated with increased risk of elbow symptoms (OR=1.9, 95% CI=1.03-3.4), neck symptoms (OR=2.1, 95% CI=1.2-3.5), wrist and hand symptoms (OR=2.5, 95% CI=1.5-4.1), and shoulder symptoms (OR=2.5, 95% CI=1.5-4.2).

Job-related risk factors that were not related to WMSDs included one postural work factor (reaching or working away from the body), one personal work factor (inadequate training in injury prevention), and 4 activities (assisting patients during gait activities; carrying, lifting, or moving heavy materials and equipment; working with confused or agitated patients; and unanticipated sudden movements or falls by patients).

Strategies Used by Physical Therapists to Minimize Effects and Risks of Developing WMSDs

Respondents used various aids to reduce the strain on their body while working. The majority of therapists (n=415 [77.4%]) used an adjustable bed or plinth, 243 therapists (45.3%) used a wheelie stool, 86 therapists (16%) used lifting belts, 55 therapists (10.3%) used slide boards, 36 therapists (6.7%) used splints, and 123 therapists (22.9%) used other (unspecified) assistance. Only 50 therapists (9.3%) indicated that they used no aids to reduce the physical strain on their bodies. Three hundred sixty-nine therapists (73.4%) changed or modified treatment at some time as a result of WMSDs, but the exact nature of the treatment modification was unspecified.

Table 6 summarizes self-protective strategies therapists used to reduce the strain on their bodies while working. The majority of respondents used 3 of the 4 preventive strategies. Almost all therapists (n=503 [98.2%]) reported modifying the patient or therapist position at least sometimes, 455 therapists (95.4%) reported that they adjusted the plinth or bed height at least sometimes, and 393 therapists (78%) reported that they paused to stretch and change posture at least sometimes. Although 468 respondents indicated that warming up and stretching before performing manual techniques (the remaining preventive strategy) was relevant, only 96 therapists (20.5%) did this at least sometimes.

Table 6. Percentage of Therapists Who Reported Using Various Self-Protective Behaviors to Reduce Work-Related Strain on Their Bodies

Strategy             Self-Protective Behavior

Outsourcing          I get someone else to help me handle a heavy
                       patient (n=405)
                     I use physical therapist assistants to perform
                       physically stressful tasks (n=298)
Preventive--modify   I pause regularly so I can stretch and change
  technique or         posture (n=504)
  environment        I adjust plinth/bed height before treating a
                       patient (n=477)
                     I modify patient position/my position (n=512)
                     I warm up and stretch before performing
                       manual techniques (n=468)
Reactive--response   I stop a treatment if it causes or aggravates
  to discomfort        my discomfort (n=477)
  or injury          I use electrotherapy instead of manual
                       techniques to avoid stressing an injury
                       (n=392)
                     I select techniques that will not aggravate or
                       provoke my discomfort (n=473)
                     I use a different part of my body to administer
                       a manual technique (n=460)

                                                   Percentage of
                                                      Physical
                                                     Therapist
                                                       Using
                                                      Strategy

Self-Protective Behavior                           Almost Always

I get someone else to help me handle a heavy           64.7
  patient (n=405)
I use physical therapist assistants to perform         10.1
  physically stressful tasks (n=298)
I pause regularly so I can stretch and change          36.7
  posture (n=504)
I adjust plinth/bed height before treating a           88.7
  patient (n=477)
I modify patient position/my position (n=512)          86.1
I warm up and stretch before performing                 6.7
  manual techniques (n=468)
I stop a treatment if it causes or aggravates my       36.7
  discomfort (n=477)
I use electrotherapy instead of manual                  5.9
  techniques to avoid stressing an injury
  (n=392)
I select techniques that will not aggravate or         40.2
  provoke my discomfort (n=473)
I use a different part of my body to administer        42.2
  a manual technique (n=460)

                                                   Percentage
                                                   of Physical
                                                   Therapist
                                                     Using
                                                   Strategy

Self-Protective Behavior                           Sometimes

I get someone else to help me handle a heavy         22.2
  patient (n=405)
I use physical therapist assistants to perform       22.8
  physically stressful tasks (n=298)
I pause regularly so I can stretch and change        41.3
  posture (n=504)
I adjust plinth/bed height before treating a          6.7
  patient (n=477)
I modify patient position/my position (n=512)        12.1
I warm up and stretch before performing              13.8
  manual techniques (n=468)
I stop a treatment if it causes or aggravates my     35.2
  discomfort (n=477)
I use electrotherapy instead of manual               18.6
  techniques to avoid stressing an injury
  (n=392)
I select techniques that will not aggravate or       37.2
  provoke my discomfort (n=473)
I use a different part of my body to administer      38.7
  a manual technique (n=460)

                                                   Percentage of
                                                     Physical
                                                    Therapist
                                                      Using
                                                     Strategy

Self-Protective Behavior                           Almost Never

I get someone else to help me handle a heavy           13.1
  patient (n=405)
I use physical therapist assistants to perform         67.1
  physically stressful tasks (n=298)
I pause regularly so I can stretch and change          22.0
  posture (n=504)
I adjust plinth/bed height before treating a           4.6
  patient (n=477)
I modify patient position/my position (n=512)          1.8
I warm up and stretch before performing                79.5
  manual techniques (n=468)
I stop a treatment if it causes or aggravates my       28.1
  discomfort (n=477)
I use electrotherapy instead of manual                 75.5
  techniques to avoid stressing an injury
  (n=392)
I select techniques that will not aggravate or         22.6
  provoke my discomfort (n=473)
I use a different part of my body to administer        19.1
  a manual technique (n=460)

Table 7 documents self-protective behaviors that were associated with particular WMSDs. Outsourcing and reactive behaviors were related to WMSDs, but preventive behaviors were not. Getting help with a heavy patient was related to decreased symptoms in neck ([[chi square].sub.(1)]=5.61, P=.018), shoulders ([[chi square].sub.(1)]=7.15, P=.007), upper back ([[chi square].sub.(1)]=8.44, P=.004), wrists and hands ([[chi square].sub.(1)]=9.24, P=.002), and thumbs ([[chi square].sub.(1)]=19.85, P=.001). Using physical therapist assistants to perform physically stressful tasks was related to decreased symptoms in the wrists and hands ([[chi square].sub.(1)]=3.93, P=.048) and thumbs ([[chi square].sub.(1)]=5.38, P=.02).

Table 7. Relationship Between Self-Protective Behavior Physical Therapists Reported Using and Work-Related Musculoskeletal Disorders

                                                     Related to
Strategy               Self-Protective Behavior      Symptoms:

Outsourcing            I get someone else to help    Neck
                         me handle a heavy patient   Shoulders,
                         (related to decreased       upper back
                         symptoms)                   Wrists/hands
                                                     Thumbs

                       I use physical therapist      Wrists/hands
                         assistants to perform       Thumbs
                         physically stressful
                         tasks (related to
                         decreased symptoms)

Preventive--modify     I pause regularly so I can
technique or             stretch and change
environment              posture

                       I adjust plinth/bed height
                         before treating a patient

                       I modify patient position/
                         my position

                       I warm up and stretch
                         before performing manual
                         techniques

Reactive--response     I stop a treatment if it
to discomfort or        causes or aggravates my
injury                  discomfort

                       I use electrotherapy          Low back, knees
                         instead of manual           Ankles/feet
                         techniques to avoid         Shoulders
                         stressing an injury         Wrists/hands
                         (related to increased       Elbows
                         symptoms)

                       I select techniques that      Shoulders
                         will not aggravate or       Elbows
                         provoke my discomfort       Knees
                         (related to increased
                         symptoms)

                       I used a different part of    Wrists/hands
                         my body to administer a     Elbows
                         manual technique (related
                         to increased symptoms)

Using electrotherapy to avoid stressing an injury was associated with increased symptoms in the previous year in the low back ([chi square] (1)=4.13, P=.042), knees ([[chi square].sub.(1)]=4.23, P=.04), shoulders ([[chi square].sub.(1)]=9.15, P=.002), ankles and feet ([[chi square].sub.(1)]=8.56, P=.003), wrists and hands ([[chi square].sub.(1)]=7.20, P=.007), and elbows ([[chi square].sub.(1)]=15.27, P=.001). Selecting techniques that would not aggravate or provoke discomfort was associated with increased shoulder symptoms ([[chi square].sub.(1)]=5.01, P=.024), elbow symptoms ([[chi square].sub.(1)]=5.61, P=.018), and knee symptoms ([chi square] (1)=14.24, P=.001). Using a different part of the body to administer a manual technique was associated with increased wrist and hand symptoms ([[chi square].sub.(1)]=5.0, P=.025) and elbow symptoms ([[chi square].sub.(1)]-5.25, P=.022).

Therapists who used outsourcing had fewer symptoms in the areas indicated than did therapists who almost never used these options. Therapists who used reactive strategies almost always or sometimes had a higher prevalence of WMSDs in the areas indicated (Tab. 7).

Electrotherapy was the protective behavior most commonly related to the presence of moderately severe symptoms. Therapists who used electrotherapy at least sometimes to reduce the strain on their bodies were more likely to have symptoms scoring 3 or more in the neck ([[chi square].sub.(1)]=7.58, P=.006), shoulders ([[chi square].sub.(1)]=4.1l, P=.043), upper back ([[chi square].sub.(1)]=11.22, P=.001), low back ([[chi square][.sub.(1)]=6.15, P=.013), wrists and hands ([[chi square].sub.(1)]=10.16, P=.001), and thumbs ([[chi square].sub.(1)]=4.71, P=.030). Using a different part of the body to administer a manual technique ([[chi square].sub.(1)]=4.72, P=.030) and selection of techniques that would not aggravate an injury ([[chi square].sub.(1)]=4.10, P=.043) were related to moderately severe wrist symptoms.

Responses of Physical Therapists Who Developed WMSDs

Only 36 (7.4%) of the 488 (91.0%) therapists who had experienced WMSDs had lodged a workers' compensation claim, 66 therapists (13.6%) had taken sick leave, and 411 therapists (84.2%) continued working with discomfort. Table 8 shows the percentages of therapists who were prevented from working, who were prevented from performing their usual ADL and leisure activities, and who sought treatment. Of all therapists, the majority (n=461 [86.0%]) were not prevented from working by WMSDs in the last year. More than half of all therapists (n=327 [61.0%]) sought treatment, and 226 therapists (42.2%) were prevented from performing their normal ADL and leisure activities. Of the subgroup of 335 physical therapists (62.5%) who reported low back symptoms in the last year, only 42 (12.5%) were prevented from working as a consequence, but 139 (41.5%) were prevented from doing their normal leisure activities and ADL.

Table 8.

Consequences of Musculoskeletal Symptoms Over the Last Year in 10 Body Areas (n=536)

                                           Prevented
                                        From Performing    Sought
                      Prevented From    ADL(a)/Leisure    Treatment
Body Area              Working (%)      Activities (%)       (%)

Low back                   7.8               25.9           36.6
Neck                       3.5               13.6           34.0
Thumbs                     2.6                5.4            8.6
Upper back                 1.7                9.3           26.7
Wrists/hands               1.7                6.5            8.4
Shoulders                  1.1                6.9           12.7
Knees                      1.1                3.9            3.4
Elbow/forearm              0.7                5.0            6.3
Ankles/feet                0.7                1.9            2.8
Hips/thighs                0.6                2.1            4.1
Percentage of total
  respondents who         14.0               42.2           61.0
  were affected

(a) ADL=activities of daily living.

Ninety-five respondents (17.7%) changed their specialty area of practice or left the profession altogether as a result of WMSDs. Only 17 therapists (3.2%) left the profession altogether for work-related health reasons, and all except 3 therapists had tried changing their specialty area of practice within the profession before finally leaving. Therapists who had changed their specialty area of practice or left the profession were more likely to have moderately severe LBP ([[chi square].sub.(1)]=10.82, P=.001) in the preceding year than those who had not.

Figure 5 shows the specialty areas therapists left because of WMSDs. Thirty-nine therapists (42% of those who changed their specialty area of practice) left neurology and rehabilitation to work in another area. Nineteen therapists (21%) left manipulative therapy or private practice, and 14 therapists (14.8%) left orthopedics. Other areas that physical therapists left because of WMSDs were general hospital work, pediatrics, and nursing homes and gerontology.

[Figure 5 ILLUSTRATION OMITTED]

Figure 6 shows the 15 specialty areas these therapists entered. Only those areas entered by more than 3% of those therapists who changed their specialty area of practice are shown. Twelve therapists (12.9%) went into ergonomics and occupational rehabilitation, 11 therapists (11.8%) entered women's health, and 10 therapists (10.6%) entered administration. The "other" areas entered by the 17.7% who changed their specialty area of practice included cardiorespiratory, sports, academia, not working, and areas other than physical therapy.

[Figure 6 ILLUSTRATION OMITTED]

Discussion

The main finding of this study was that 1 in 6 physical therapists changed their specialty area or left the physical therapy profession because of WMSDs. This finding represents unknown personal and financial costs to the therapists, the profession, and the community. This figure is likely to understate the problem, as those therapists whose state registration had lapsed were not included in this study.

A second major finding was the relationship between thumb symptoms and the use of mobilization and manipulation techniques. This finding is consistent with anecdotal and unpublished reports by physical therapists and our knowledge of WMSDs experienced by physical therapists.

The previously undocumented relationships between risk factors and WMSDs in physical therapists suggest that principles of injury prevention techniques utilized in industry generally may be applicable to the physical therapy profession. These principles indicate practical ways of altering clinical practice to reduce the risk of injury.

Distribution, Prevalence, and Severity of WMSDs

The majority (91%) of physical therapists reported that they had experienced WMSDs at some time. The WMSDs were related to age, sex, specialty area, and specific tasks. The respondents identified risk factors contributing to their symptoms. Analysis showed that some of these risk factors were related to the presence of WMSDs in specific body areas. Workload issues were identified as being related to the presence of WMSDs, particularly upper-body symptoms.

The prevalence of low back symptoms among physical therapists in this study was generally higher than that reported by other authors.[1-3,5] Two of the previous studies[1,2] were published in 1985 and 1989, and it is plausible that the observed differences may be due to changes in practice over that time. Alternatively, these findings may reflect regional differences in how therapists practice.

The increased prevalence of symptoms among younger therapists has been attributed to various factors. Our data are consistent with the reluctance of younger therapists to seek assistance with physically demanding tasks and with their inexperience,[2,3] as more than 50% had their first episode as a student or in their first 5 years of practice.

The explanation proposed by Bork et al,[5] that the higher prevalence of WMSDs among younger therapists was due to survivor bias, is supported, in part, by the data from our study. We found the practice areas most frequently left by respondents were neurology and rehabilitation, neither of which demonstrated increased prevalence of low back symptoms in the previous year. This finding suggests survivor bias, where therapists who have LBP remove themselves from the specialty area. Another suggestion, that older physical therapists are likely to move into less physically demanding work (eg, administration),[2,5] was not supported directly by our data. We found that only 10.6% of therapists who changed their specialty area of practice within the profession because of WMSDs went into administration.

The finding that male physical therapists had more neck, wrist and hand, and thumb symptoms than did female therapists contrasts with the finding by Bork et al.[5] This increased prevalence of symptoms in male therapists may relate to their greater usage of mobilization and manipulation techniques.

The association between the use of mobilization and manipulation techniques and thumb symptoms suggests implications for the way in which therapists practice. High ORs and a dose-response relationship support the notion of cause and effect and imply that there should be some limits placed by therapists on the number of hours for which they use these techniques.

Work-related musculoskeletal disorders affected the therapists to varying degrees. Some therapists simply recorded the presence of symptoms, whereas other therapists variously reported that symptoms required treatment or interfered with leisure activities, ADL, and work. Low back symptoms were most intrusive, interfering with ADL, leisure activities, and work more often than those in other body areas. That almost 60% of all therapists had moderately severe symptoms and more than 40% compromised their ADL or leisure activities indicates that the issue of musculoskeletal injury within the physical therapy profession is widespread and not without cost.

Specialty Areas, Tasks, Risk Factors, and Development of WMSDs

The only specialty areas of practice related to WMSDs were sports physical therapy, private practice, and pediatrics. The increased prevalence of musculoskeletal symptoms among therapists employed in sports physical therapy and private practice may relate to the type of tasks they perform, rather than the area itself. The higher prevalence of knee symptoms among pediatric physical therapists is consistent with the findings of a previous study[5] and presumably due to the large amount of time spent by these therapists in kneeling and crouching.

Physical therapists may be exposed simultaneously to a number of different risk factors. It is likely that risk factors may interact, making identification of the cause of injury difficult. That upper-limb, neck, and upper back symptoms were related to mobilization and manipulation techniques and other hands-on treatment suggests that something about the performance of these techniques contributes to symptoms in these areas.

Lifting dependent patients was related to the development of low back symptoms. This is a commonly accepted belief, supported by research,[25-28] particularly in nursing. However, given that therapists self-identified the contributing risk factors, the association may have been due to bias and what therapists believed to be true, rather than the actual contribution this factor made to their injuries. This finding should thus be viewed cautiously until it is independently verified.

Workload issues, relating to the way physical therapists practice, were related to symptoms in the neck, upper back, and upper limbs. The prevalence of these symptoms was also higher in therapists who had worked in private practice. It seems that the way physical therapists work is related to their musculoskeletal health, particularly in the area of private practice, where these issues are directly related to the income of the practice.

Performing the same task repeatedly was related to the presence of symptoms in many areas and calls into question the wisdom of practicing in such a way. Concepts such as job rotation and variety in work are commonly applied in industry to avoid overloading any particular anatomical area, either by sustained posture or repetitive actions. Repeated muscle contractions and static loading are known to be risk factors in the development of cumulative trauma disorders,[29-33] Kroemer stated that provision of alternating work "which allows breaks in otherwise repetitive or maintained activities" is essential in the prevention of such disorders.[30(p280)] Thus, physical therapists should ensure that they vary their techniques in order to place varying stresses on different anatomical areas. Within specialty areas, therapists need to have at their disposal a variety of treatment tools. This is not only so that the ideal treatment may be given, but also so that they can vary the way in which they use their body, thereby reducing the risk to any one body part.

The range of conditions and type of clients treated, the financial arrangements of the therapist, and the setup of the practice may also influence the development of symptoms. These possibly confounding factors were not considered in this study, and they suggest a focus for further research.

Strategies Used by Physical Therapists to Minimize Effects and Risks of Developing WMSDs

The majority of therapists used some aids to reduce the strain on their bodies. The most commonly used aid was the height-adjustable bed, which reduces the postural strains on the spine. However, postural strains are only one of the risk factors to which physical therapists are exposed. Other factors such as workload issues, personal factors, and specific tasks also play a role and should be considered when planning ways to reduce the occurrence of WMSDs.

The inverse relationship between outsourcing options and upper-limb symptoms may be because the specialty, areas where assistance with patients is used (neurology and rehabilitation or another "heavy" area) are not those where upper-limb symptoms were most prevalent (private practice and sports physical therapy). It appears that the tasks performed within the different specialty areas contribute to the presence or absence of WMSDs, rather than the use of outsourcing options.

Another reasonable inference from the data is that it is the use of the outsourcing options that has resulted in the decrease in symptoms in these anatomical areas. It seems intuitively probable that the specialty area where therapists get help with heavy patients is unlikely to be one where therapists use manipulation and mobilization techniques extensively. The data, however, do not support either explanation for the relationship between outsourcing and symptoms.

No WMSDs were related to the use of preventive strategies, supporting the idea that these strategies are effective in preventing work-related musculoskeletal injury. Most therapists used at least 3 of the 4 nominated strategies.

Reactive strategies were those used by therapists in response to the presence or perceived risk of WMSDs. Our findings suggest that injured physical therapists may sometimes select treatment techniques and modalities for reasons other than the needs of the patient, namely self-preservation. The greater use of these strategies among therapists with moderately severe upper-limb and spinal symptoms suggests that reactive strategies aid symptom management, enabling therapists to continue working. Although the majority of therapists almost never used the option of treating with electrotherapy instead of manual techniques in order to avoid stressing an injury, up to 24% reported that they sometimes used this as a strategy to protect themselves. This finding may help explain previously puzzling findings reported by Robertson and Spurritt[34] of high clinical use of electrotherapy with little basis in research findings. Robertson and Spurritt's finding is consistent with ours, that using electrotherapy was the protective behavior most commonly related to the presence of moderately severe symptoms.

Selecting techniques that will not aggravate or provoke the therapists' discomfort or using an alternate body part to administer a technique implies that the therapist has an ample range of options to use in treating patients. The range of available options may expand with experience (and possibly in response to injury), which may partly explain the higher prevalence of symptoms among younger therapists.

Responses of Physical Therapists Who Developed WMSDs

The 1 7.7% of therapists who changed their specialty area of practice within the profession or who left the profession is a substantial group. This percentage means that 1 in 6 therapists can expect to change their specialty area of practice or leave the physical therapy profession because of WMSDs. A partial explanation for the difference between our finding and the previous finding of 5.2% of therapists changing their specialty area of practice[2] is that we included therapists who changed their specialty area of practice because of WMSDs in all body areas, not only those with LBP. Other factors may also contribute to the discrepancy between the findings, such as changes to the health care system over time and differing emphases and scope of practice in different parts of the physical therapy community. Other researchers have not reported movement within (or away from) the profession.

Incongruent time frames preclude causal inferences that therapists who changed their specialty area of practice within the profession did so because of a severity score of 3 or higher for the low back, but this finding suggests that back symptoms may be a factor in changing specialty area. This conclusion is consistent with the evidence that low back symptoms were the symptoms that most frequently interfered with ADL, leisure activities, and work and may be more disabling than symptoms in any other parts of the body.

Limitations

In our study, we used a cross-sectional design; thus, causal inferences cannot be drawn from the results. A second limitation of this study is the reliance on self-reported data. With all self-reported data, there is a possibility that individuals with symptoms tend to overestimate their exposure.[35] In our study, some therapists with injuries may have overestimated the number of hours per week they spent performing techniques they subsequently perceived as contributing to the injuries.

Physical therapists are trained to understand injury and its causes, which lends some credibility to their self-reported symptoms. At the same time, they may be more self-aware than other populations because of their training and thus tend to over-report symptoms. There is little evidence to support either of these views, and further research is necessary to clarify the accuracy of the self-reporting of symptoms by physical therapists. However, the Ors lend weight to the notion of cause and effect, particularly in the relationship between the performance of mobilization and manipulation techniques and the development of thumb symptoms and hand and wrist symptoms.

Risk factors were self-determined, with therapists indicating the degree to which a risk factor had contributed to their injury. This is a potential source of bias in the study. Given the training of physical therapists in biomechanics and the principles of injury, however, they may be expected to give a reasoned account of the risk. At the same time, commonly held beliefs about the risk of patient handling may influence their perceptions of risk. Validation of this means of determining risk in this population suggests a need for further investigation. The severity measure was a constructed variable, and the scale has not been validated. The data should be taken at face value.

Conclusions

One in 6 physical therapists changed their specialty area or left the physical therapy profession because of work-related musculoskeletal problems. The greatest proportion left neurology and rehabilitation, and those therapists who changed their specialty area entered a variety of specialty areas. Little is known of this group of therapists, and further research is under way to better understand the issues and costs involved in changing specialty area or leaving the profession.

The dose-response relationship between the number of hours spent performing mobilization and manipulation techniques and the prevalence of thumb symptoms has not previously been documented and suggests that causality is probable. Further study is needed to establish a more precise relationship and to determine what proportion of work time can be safely spent using these types of techniques. Objective criteria for measuring exposure are needed to enable the specific risk factor to be identified.

Postural risk factors, the performance of manual orthopedic techniques, and workload issues were related to symptoms in the low back, neck, upper back, and the wrists, hands, and thumbs. Personal factors also were instrumental in upper-body injuries. As has been demonstrated by other researchers, a knowledge of ergonomics, injury, and treatment does not offer the physical therapist immunity from injury. Further research is needed to identify those aspects of the job and associated work practices contributing to injury, with a view to formulating preventive strategies.

The increased prevalence of symptoms among younger physical therapists in particular underlines the need for them to have at their disposal a range of strategies to reduce risks posed by their work and avoid injury. Most importantly, there is a need for further research to identify aspects of

physical therapy practice that place therapists at greatest risk and to develop methods of reducing that risk.

(*) SPSS Inc, 444 N Michigan Ave, Chicago, IL 60611

References

[1] Scholey M, Hair M. Back pain in physiotherapists involved in back care education. Ergonomics. 1989;32:179-190.

[2] Molumphy M, Unger B, Jensen GM, Lopopolo RB. Incidence of work-related low back pain in physical therapists. Phys Ther. 1985;65: 482-486.

[3] Mierzejewski M, Kumar S. Prevalence of low back pain among physical therapists in Edmonton, Canada. Disabil Rehabil. 1997;19: 309-317.

[4] van Doorn JWC. Low back disability among self-employed dentists, veterinarians, physicians and physical therapists in The Netherlands: a retrospective study over a 13-year period (N=1,119) and an early intervention program with 1-year follow-up (N=134). Acta Orthop Scand Suppl. 1995;263:1-64.

[5] Bork BE, Cook TM, Rosecrance JC, et al. Work-related musculoskeletal disorders among physical therapists. Phys Ther. 1996;76:827-835.

[6] Knibbe JJ, Friele RD. Prevalence of back pain and characteristics of the physical workload of community nurses. Ergonomics. 1996;39: 186-198.

[7] Moffett JA, Hughes GI, Griffiths P. A longitudinal study of low back pain in student nurses. Int J Nurs Stud. 1993;30:197-212.

[8] Pheasant S, Stubbs D. Back pain in nurses: epidemiology and risk assessment. Applied Ergonomics. 1992;23:226-232.

[9] Smedley J, Egger P, Cooper C, Coggon D. Manual handling activities and risk of low back pain in nurses. Occup Environ Med. 1995;52: 160-163.

[10] Hignett S. Work-related back pain in nurses. J Adv Nurs. 1996;23: 1238-1246.

[11] Balon MB. Thumb and Wrist Symptoms of Manipulative Therapists' [thesis]. Melbourne, Victoria, Australia: Lincoln Institute; 1984:40.

[12] Jensen JB. Stress and Joint Symptoms Survey Related to Work Environment of Physiotherapists' and Manipulative Therapists [thesis]. Melbourne, Victoria, Australia: La Trobe University; 1983:23.

[13] Blizzard P. Save our thumbs. Physiotherapy. 1991;77:573-574.

[14] Garg A, Owen B, Beller D, Banaag J. A biomechanical and ergonomic evaluation of patient transferring tasks: bed to wheelchair and wheelchair to bed. Ergonomics. 1991 ;34:289-312.

[15] Garg A, Owen B, Beller D, Banaag J. A biomechanical anti ergonomic evaluation of patient transferring tasks: wheelchair to shower chair and shower chair to wheelchair. Ergonomics. 1991;34: 407-419.

[16] Garg A, Owen BD. Reducing back stress to nursing personnel: an ergonomic intervention in a nursing home. Ergonomics. 1992;35: 1353-1375.

[17] Garg A, Owen B. Prevention of back injuries in healthcare workers. International Journal of Industrial Ergonomics. 1994; 14:315-331.

[18] Robertson LD, Changsut R, Ramos LS, Jones DW. Influence of job and personal risk factors on safety limits for kinesiotherapists performing a stressful clinical lifting task. Clinical Kinesiology. Spring 1993:7-16.

[19] Holmstrom EB, Lindell J, Moritz U. Low back and neck/shoulder pain in construction workers: occupational workload and psychosocial risk factors, part 1: relationship to low back pain. Spine. 1992;17: 663-671.

[20] Vuori I. Exercise and physical health: musculoskeletal health and functional capabilities. Res Q Exerc Sport. 1995;66:276-285.

[21] Sutton GS, Bartel MR. Soft-tissue mobilization techniques for the hand therapist. J Hand Ther. 1994;7:185-192.

[22] Greene L. Designing your life and work to prevent injury. In: Greene R, ed. Save Your Hands/ Injury Prevention for Massage Therapists. Seattle, Wash: Infinity Press; 1995:127-152.

[23] Kuorinka I, Jonsson B, Kilbom A, et al. Standardised Nordic questionnaires for the analysis of musculoskeletal symptoms. Applied Ergonomics. 1987;18:233-237.

[24] Australian Institute of Health and Welfare. Physiotherapy Labour Force 1993. Canberra, Australia: Health Labour Force Bulletin; 1995.

[25] Jorgensen S, Hein HO, Gyntelberg F. Heavy lifting at work and risk of genital prolapse and herniated lumbar disc in assistant nurses. Occup Med (Lond). 1994;44:47-49.

[26] Dehlin O, Jaderberg E. Perceived exertion during patient lifts: an evaluation of the importance of various factors for the subjective strain during lifting and carrying of patients--a study at a geriatric hospital. Scand J Rehabil Med. 1982;14:11-20.

[27] Gagnon M, Sicard C, Sirois J-P. Evaluation of forces on the lumbo-sacral joint and assessment of work and energy transfers in nursing aides lifting patients. Ergonomics. 1986;29:407-421.

[28] Gagnon M, Smyth G. Biomechanical exploration on dynamic modes of lifting. Ergonomics. 1992;35:329-345.

[29] Kilbom A. Repetitive work of the upper extremity, part II: the scientific basis (knowledge base) for the guide. International Journal of Industrial Ergonomics. 1994; 14:59- 86.

[30] Kroemer KHE. Cumulative trauma disorders: their recognition and ergonomics measures to avoid them. Applied Ergonomics. 1989;20: 274-280.

[31] Kenny D, Powell NJ, Reynolds-Lynch K. Trends in industrial rehabilitation: ergonomics and cumulative trauma disorders. Work. 1995;5:133-142.

[32] Ranney D, Wells R, Moore A. Upper limb musculoskeletal disorders in highly repetitive industries: precise anatomical physical findings. Ergonomics. 1995;38:1408-1423.

[33] Roquelaure Y, Mechali S, Dano C, et al. Occupational and personal risk factors for carpal tunnel syndrome in industrial workers. Scand J Work Environ Health. 1997;23:364-369.

[34] Robertson VJ, Spurritt D. Electrotherapy agents: implications of EPA availability and use in undergraduate clinical placements. Physiotherapy. 1998;84:335-344.

[35] Viikari-Juntura E, Rauas S, Martikainen R, et al. Validity of self-reported physical work load in epidemiologic studies on musculoskeletal disorders. Scand J Work Environ Health. 1996;22:251-259.

JE Cromie, PT, BAppSc(Phty), GradDipOccHealth, is a PhD student in the Schools of Physiotherapy and Occupational Therapy, LaTrobe University, Bundoora, Victoria, Australia 3083 ([e-mail]). Address all correspondence to Ms Cromie.

VJ Robertson, PT, PhD, BAppSc(Phty), BA(Hons), is Associate Professor, School of Physiotherapy, LaTrobe University.

MO Best, PT, MPH, BAppSc(Phty), GradDipErg, is Senior Lecturer, School of Occupational Therapy, LaTrobe University.

All authors contributed to concept and research design, writing, project management, fund procurement, and consultation (including review of manuscript before submission). Ms Cromie provided data collection and analysis, subjects, and clerical support. Dr Robertson and Ms Best provided facilities and equipment.

Ethics approval for this study was granted by the Faculty Human Ethics Committee, La Trobe University.

This project was financially supported by an Australian Physiotherapy Association Victorian Branch Research Grant.

The findings of this study were presented, in part, at the Fifth International Australian Physiotherapy Association Congress; May 1998; Hobart, Tasmania.

This article was submitted June 11, 1999, and was accepted November 9, 1999.