Journal of Occupational Rehabilitation

, Volume 20, Issue 2, pp 199–219

Occupational Safety and Health Interventions to Reduce Musculoskeletal Symptoms in the Health Care Sector

Authors

    • School of Public Health, Institute for Health PolicyThe University of Texas
  • Shelley Brewer
    • Chemplan, Inc.
  • Benjamin C. AmickIII
    • The Institute for Work & Health
    • School of Public Health, Division of Health Promotion and Behavioral SciencesThe University of Texas
  • Emma Irvin
    • The Institute for Work & Health
  • Quenby Mahood
    • The Institute for Work & Health
  • Lisa A. Pompeii
    • School of Public Health, Division of Epidemiology and Disease ControlThe University of Texas
  • Anna Wang
    • The Institute for Work & Health
  • Dwayne Van Eerd
    • The Institute for Work & Health
  • David Gimeno
    • School of Public Health at San Antonio Campus, Division of Environmental and Occupational HealthThe University of Texas
  • Bradley Evanoff
    • School of MedicineWashington University at St. Louis
Article

DOI: 10.1007/s10926-010-9231-y

Cite this article as:
Tullar, J.M., Brewer, S., Amick, B.C. et al. J Occup Rehabil (2010) 20: 199. doi:10.1007/s10926-010-9231-y

Abstract

Introduction Health care work is dangerous and multiple interventions have been tested to reduce the occupational hazards. Methods A systematic review of the literature used a best evidence synthesis approach to address the general question “Do occupational safety and health interventions in health care settings have an effect on musculoskeletal health status?” This was followed by an evaluation of the effectiveness of specific interventions. Results The initial search identified 8,465 articles, for the period 1980–2006, which were reduced to 16 studies based on content and quality. A moderate level of evidence was observed for the general question. Moderate evidence was observed for: (1) exercise interventions and (2) multi-component patient handling interventions. An updated search for the period 2006–2009 added three studies and a moderate level of evidence now indicates: (1) patient handling training alone and (2) cognitive behavior training alone have no effect on musculoskeletal health. Few high quality studies were found that examined the effects of interventions in health care settings on musculoskeletal health. Conclusions The findings here echo previous systematic reviews supporting exercise as providing positive health benefits and training alone as not being effective. Given the moderate level of evidence, exercise interventions and multi-component patient handling interventions (MCPHI) were recommended as practices to consider. A multi-component intervention includes a policy that defines an organizational commitment to reducing injuries associated with patient handling, purchase of appropriate lift or transfer equipment to reduce biomechanical hazards and a broad-based ergonomics training program that includes safe patient handling and/or equipment usage. The review demonstrates MCPHI can be evaluated if the term multi-component is clearly defined and consistently applied.

Keywords

Health careOccupational healthInterventionsSystematic reviewMusculoskeletal

Introduction

Health care workers are at a high risk of developing musculoskeletal (MSK) symptoms, injuries and disorders. Health care workers injury rates equal or exceed rates in other industries that are traditionally considered hazardous [1]. The total cost of such injuries is unknown, but in 2000, the U.S. Veteran’s Administration—a large hospital system—spent over $23 million (U.S.) for job-related injuries related to patient care [2]. The prevalence of low back pain in nursing personnel has been reported at rates between 30 and 60% [36]. In a survey of Canadian nurses 37% said that in the past 12 months they experienced pain serious enough to prevent them from carrying out normal daily activities [7]. Low back pain has been identified as a major reason why nurses leave their profession [4].

MSK disorders in health care workers have been attributed in large part to patient transfer and lifting activities. Biomechanical studies have shown that these activities place high levels of compressive force on low back structures, exceeding the lifting limits recommended by the U.S. National Institute for Occupational Safety and Health (NIOSH) [8]. Shoulder, knee and other disorders have also been associated with patient lifting and transferring tasks [9]. Various interventions have been implemented to reduce back and other MSK disorders among health care workers. The interventions include worker education programs, physical conditioning or exercise programs, disability management, organizational policies and the use of mechanical lifts or other patient transfer equipment [1012].

Mechanical patient handling and transfer devices have been a major focus of injury prevention efforts in the health care setting. Numerous facilities have instituted “zero-lift” policies banning manual lifting [13]. Some U.S. states have enacted legislation encouraging or requiring health care facilities to have lifting devices available [14]. The provincial government of Ontario, Canada has committed over $89 million (CDN) to purchase and install more than 19,000 patient lifting devices in the provincial health care institutions and to support the provision of staff training [15]. The financial expenditures and policy initiatives demonstrate employers’ belief that biomechanical exposures contribute to the high MSK injury rates in health care workers.

In addition to the hazards of patient handling activities, recent research suggests MSK injuries in health care settings may also result from non-patient handling activities or tasks such as patient-related assaults, [16, 17] slips, trips and falls [18] and maintenance work [19]. Although systematic reviews have been conducted on nurses and specific patient-handling risk factors [1012], to the best of our knowledge, no systematic review has been conducted on the broad spectrum of interventions to prevent MSK injuries or reduce their impact in all employees in health care settings. Stakeholders such as facility managers, occupational safety and health professionals, ergonomic consultants, disability managers, etc. are faced with making decisions based on known practices and a plethora of programs and products being marketed. The need exists for a broad-based review that examines the range of programs and practices commonly accepted as prevention solutions to build an evidence platform for informed decision-making.

The question evaluated in this review was: “do occupational safety and health interventions in health care settings have an effect on musculoskeletal health status?” The review included both primary and secondary prevention intervention studies. Data from relevant studies were extracted and synthesized. Based on our synthesis, we make recommendations about improving work-related MSK health outcomes in the health care sector.

Methods

Review Process Overview

Health care intervention studies were systematically reviewed using an adaptation of the systematic review process originally developed by the Cochrane Collaboration [20]. The review team was comprised of 10 researchers from the U.S., Canada and the U.K. Reviewers were identified based on their expertise in conducting epidemiologic or occupational safety and health (OS&H) intervention studies related to MSK disorders among health care workers or in conducting systematic reviews. Review team members had backgrounds in epidemiology, ergonomics, nursing, library science, occupational medicine and safety engineering. The review team used a consensus process in decision making.

The basic steps for conducting the review are listed below.
  1. 1.

    Review question development

     
  2. 2.
    Search development and execution
    1. a.

      Stakeholder engagement

       
    2. b.

      Literature search

       
    3. c.

      Expected article comparison

       
     
  3. 3.

    Relevant documents identification (Level 1)

     
  4. 4.

    Quality appraisal (Level 2)

     
  5. 5.

    Data extraction (Level 3)

     
  6. 6.

    Evidence synthesis

     

Review Question Development

The review team agreed on the primary research question: “Do occupational safety and health interventions in health care settings have an effect on musculoskeletal health status?” We reviewed primary and secondary prevention intervention studies conducted in workplaces.

Search Development and Execution

Definitions for the three key terms “health care setting or worker,” “OS&H intervention” and “musculoskeletal health” were developed prior to performing a literature search.

“Health care setting or worker” was defined broadly as inpatient care settings or workers in such settings. These included hospitals, assisted living facilities, long-term care facilities, rehabilitation hospitals, medical centers, emergency departments, tertiary care centers and nursing homes. We excluded home health care workers or stand-alone ambulatory care facilities where patients were not part of an inpatient facility or an emergency department because exposures were expected to be different in inpatient versus outpatient facilities. Also excluded were studies in laboratories, commercial pharmacies, optometry stores, psychologists’ offices, chiropractic clinics and stand-alone alternative medicine centers including acupuncture, homeopathic, naturopath and massage clinics.

“Interventions” were defined as any occupational safety and health intervention designed to protect MSK health. We used the traditional hazard control tiers of engineering controls, administrative controls and personal protective equipment use to identify types of occupational safety and health interventions [21]. We excluded interventions in which the only outcome was violence reduction or regulation effectiveness.

“Musculoskeletal health” included MSK symptoms, disorders or clinical diagnoses. We included workers’ compensation and regulatory injury record keeping systems, despite the validity and reliability vulnerabilities of these data sources. Hereafter, we refer to workers’ compensation claims and regulatory injury record keeping systems as “administrative outcomes.” These are important to stakeholders because of their role in management planning and decision making. We included studies that used scales with anchors of MSK discomfort as it could relate to descriptions of MSK symptoms, but excluded studies where muscle loading or comfort was the only outcome as well as studies about surgeries, cancers and gynecological/pregnancy-related MSK symptoms, disorders and clinical diagnoses.

The review team considered peer-reviewed scientific articles published or in press in four languages, English, French, Spanish and Swedish, based on the language proficiency of team members. Book chapters, dissertations and conference proceedings (a.k.a. gray literature) were excluded as it was expected that key findings would be reported in the peer-reviewed literature.

Stakeholder Engagement

Stakeholders including practitioners, health care managers, health care employees and policy makers within the healthcare sectors in Canada and the United States were involved from the beginning of the review. Involving participants from two jurisdictions allowed for a broad range of stakeholder perspectives to be captured. The stakeholders provided input about the: research question, key definitions, and search terms. The stakeholders also participated in a meeting at the end of the review to provide feedback regarding the usefulness of the results.

Literature Search

Medical Subject Headings (MeSH) and free text terms were identified and combined to search the following databases: MEDLINE, EMBASE, CINAHL, Academic Source Premier, PsyclNFO and Business Source Premier. Search terms were identified for three broad areas: OS&H intervention terms, health care setting/worker terms and MSK health outcome terms (Table 1). The search categories were chosen to be exclusive within each area. The search strategy combined the three areas using the AND Boolean operator and combined terms within each category using an OR operator.
Table 1

Search terms (search strategy: terms within a row are combined with OR and between rows with AND)

Intervention terms

Training programs, orientation programs, lifting, lift devices, mechanical lift devices, zero lift, “no lifts,” minimal lifts, maximum lifts, active lifts, passive lifts, hoists, patient transfer, patient assist, material transfer, material handling, manual lift devices, manual assistance, transfer aides, transfer assistance, transfer device, slide board, organizational and policy (administrative) changes, disability management, medical management, participatory ergonomics programs, staffing, shift-work, ergonomic, job redesign, work redesign, equipment redesign, job enlargement, task rotation, work hardening, work place safety, work safety, return to work programs, prevention exercises, strength training, flexibility program, body mechanics, lifting teams, back school, psychosocial work organization, patient handling, resident handling, ceiling lifts, overhead lifts, functional abilities evaluation, functional abilities screening, physical demand analysis, engineering controls, personal protective equipment (PPE), administrative control, antifatigue mats, back belts, non-skid flooring, shoe choice, non-slip soles, slippery floor signs, wet floor signs, umbrella covers, intervention research, intervention studies, interventions

NOT:

Blood-borne pathogens, infection prevention and control, EAP or substance abuse and drug treatment programs, radiation safety courses, hazard(ous) communication, smoking cessation programs, hazwoper, needle-stick, violence prevention (if not in conjunction with something else on the inclusion), cytotoxics, glutaraldehyde formalin, formaldehyde, spill training, lab safety, chemical hygiene, hazardous waste emergency planning, lockout/tagout, energy control, mercury, infection control, ethylene oxide, universal precautions, fire prevention and control, suicide prevention, conflict resolution

Health care setting terms

Assisted-care, assisted living, nursing home, hospital, acute-care, skilled nursing facility, old age facilities, old age homes, residential care facility, long-term care, long term care facility, medical centre, tertiary care centres, direct-care workers, patient techs, CNAs (certified nursing assistants), nurses aides, nursing assistant, personal support workers, patient sitter, transporter, porter, orderlies, attendants, LVNS (licensed vocational nurses), PCA (patient care assistants), nurses, nursing, RN, registered practical nurses, nurse practitioner, nurse clinicians, clinical health nurses, medical aides, nurse anesthetist, physician, surgeon, surgical techs, or techs, residents, clerk, intern, radiology tech, diagnostic imaging, ultrasound, OT/PT, rehab therapists, recreation workers/activities workers, hospital workers, health care workers, health care aides, restorative care aides, long term care aides, retirement aides, nursing home workers, dietary aides, laundry aides, laundry, food services, housekeeping, maintenance, part-time workers, contract workers, mental health in-patient, emergency department, emergency services, pharmacists, pharmacist aides, pharmacist techs, allied health personnel, paramedic

NOT:

Out-patient, ambulatory, pharmacy, stand-alone ambulatory, stand-alone medicine centres, chiropractic, walk-in clinic, homeopathic, naturopath, massage, psychologist, urgent care clinic, urgent care centre

Musculoskeletal health outcome terms

Arm injuries, cumulative trauma disorders, tendonitis, tendinopathy, tenosynovitis, rotator cuff, neck injuries, synovitis, muscle weakness, forearm injuries, wrist injuries, hand injuries, osteoarthritis, “sprains and strains”, soft tissue injuries, arthralgia, finger injuries, tendon injuries, bursitis, nerve compression syndromes, myofascial pain syndromes, neuralgia, causalgia, radiculopathy, polyradiculoneuritis, polyneuritis, muscular diseases, carpal tunnel syndrome, shoulder impingement syndrome, thoracic outlet syndrome, tennis elbow, epicondylitis, cervico-brachial neuralgia, ulnar nerve compression syndrome, musculoskeletal diseases, musculoskeletal disorders, repetitive trauma, musculoskeletal system, musculoskeletal injuries, musculoskeletal symptom, RSI, neck pain, back pain, back injuries, degenerative disc disorders, degenerative disc diseases, intervertebral disk displacement, herniated disc, bulging disc, lumbar strain, cervical strain, thoracic strain, upper extremity/AND pain, lower extremity/AND pain, knee injuries, hip injuries, leg injuries, disability

NOT:

Cancer, surgery, pregnancy, gynecological symptoms, gynecological diseases

Expected Article Comparison

Prior to conducting the search, the review team identified a list of articles it felt were key to this topic and we would therefore hope to capture in the search strategy. Our search strategy including input from stakeholders captured all 21 relevant published articles identified by the review team. The literature search was therefore considered to have adequate search sensitivity.

The review team also contacted 16 international content experts to solicit relevant articles, conference proceedings or articles in press. Six experts responded and four suggested articles. Five of these articles had been accepted for publication and therefore were moved forward to Level 1 review.

Relevant Documents Identification (Level 1)

Review Tool Training

At every review level, all review team members participated in a pilot of the review tool. Response differences were discussed and a group consensus was reached on the review tool to insure all reviewers were consistently applying the review guides.

The broad search strategy captured many studies not relevant to the research question. A Level 1 relevance review was designed to efficiently identify and exclude non-relevant articles. Article relevance was determined by applying six criteria to the article title and abstract (Table 2). Reviewers entered answers using Systematic Review Software (SRS) [22]. SRS, a web based software, allowed centralized article tracking and access.
Table 2

Level 1—relevancy screening questions

1. Did an intervention occur in a health care setting?

2. Was the reference from a peer-reviewed publication (in-press or accepted for publication)?

3. Was the language English, Spanish, Swedish or French?

4. Did the study only have post-intervention measurements, with no control group?

5. Was individual health data collected?

6. Was outcome measure MSK symptoms/disorders/injury?

A negative response led to the article being excluded from further review, except for Q4

An additional review step was added because a large number of articles passed the title and abstract review stage with unclear responses. The additional step, Level 1b, was completed by applying the same Level 1a criteria to the full article. One team member reviewed each article at Level 1a and two reviewed each article at Level 1b. Relevant articles were moved forward for Level 2 review once the two reviewers reached consensus.

A possibility for selection bias existed since single reviewers conducted the Level 1a review. A quality control (QC) check was done with an independent reviewer (QC reviewer). The QC reviewer assessed a randomly chosen set of 12 studies from each of the seven reviewers who participated in Level 1 review. Each set included six excluded articles and six that progressed to the next review level. QC reviewer responses were entered into SRS software so they could be directly compared to a team member’s responses. The QC reviewer disagreed with the original reviewer on 28 articles. In 25 of 28 cases (90%), the QC reviewer excluded the study while the original reviewer included it. The original reviewer was more likely to be inclusive than the QC reviewer. Over-inclusion was not considered a problem since upon inspection all 25 articles were excluded at the next review step. On three occasions the reviewer excluded the article and the QC reviewer included it. Of those articles, two were less than three pages long and the third article was a regulatory intervention; therefore, the original exclusions were appropriate. The Level 1a review process was considered reasonable and not vulnerable to selection biases.

Individual Health Data Criterion

While conducting the Level 1 review, it became clear that if we excluded studies that did not collect individual health data, as per our screening questions (Table 2), we would exclude relevant studies. Some studies use work unit or worksite aggregated injury rates as outcomes as opposed to following a cohort of individuals over time. Consequently studies that used aggregate injury information at a work unit or a work site were re-examined for possible inclusion. However, included studies must have calculated an “injury rate” and not simply reported injury counts. The group reviewed all studies that had been excluded because of lack of individual health data (n = 16) and eight went onto Level 2 review. The other 50% had injury counts, not rates, and were excluded at Level 1.

Quality Appraisal (Level 2)

The articles that passed the Level 1 review were evaluated for methodological quality during Level 2 review. The team identified 19 methodological criteria to assess quality (Table 3). Study quality is important for a number of reasons, among them the view that lower quality studies have been shown more likely to report positive intervention effects [23]. Each article was independently reviewed by two team members using a review guide. To reduce bias, the same two members did not review all of the same articles. Each reviewer was randomly rotated with at least two other team members. Team members did not review articles they had consulted on, authored or co-authored. Reviewer pairs were required to reach consensus on all 19 quality criteria. In cases where agreement could not be reached, a third reviewer was consulted to ensure consensus was obtained.
Table 3

Level 2—quality appraisal questions and weights

Question

Weight

1. Was the research question/objective clearly stated?

2

2. Was a primary hypothesis clearly stated?

1

3. Was the intervention implementation described?

3

4. Was the calendar duration of the intervention documented?

1

5. Was the length of follow-up 3 months or greater?

2

6. Were concurrent comparison (control) group(s) used?

4

7. Was the intervention allocation randomized?

4

8. Were sample inclusion/exclusion criteria described?

2

9. Was the sampling frame representative of the target population?

2

10. Was the participation rate reported and greater than 40% for employees?

3

11. Did the researchers describe the study participants at baseline by demographics, exposure or outcome?

2

12. Were baseline characteristics presented by group?

3

13. Were differences between those employees who remained in the study and those who dropped out analyzed?

3

14. Did withdrawals affect groups equally?

3

15. Were the effects of the intervention on some exposure parameters documented?

3

16. Was contamination between groups described or documented?

1

17. Were covariates/potential confounders for MSK disorders measured (i.e. gender, age, non-work activities)?

3

18. Was adjustment made for covariates/potential confounders?

2

19. Were statistical methods adequately described?

3

The review team assigned weights, a priori, for each criterion because each methodological criterion was not considered equal. The four-point weighting ranged from “somewhat important” (1 point) to “very important” (4 points). Summary quality scores were created for each article by dividing the weighted sum of the 19 quality criteria by 47 (highest possible weighted score) and multiplying by 100. The quality ranking score was used to group articles into four quality categories: high (80–100%), medium–high (60–79%), medium (40–59%) and limited (less than 40%). The categories were determined by team consensus with reference to the literature [20, 24, 25]; and past prevention reviews [26].

Data Extraction (Level 3)

The quality ranking represents the review team’s assessment of each study’s internal, external, construct and statistical conclusion validity [27]. Each validity type is important in determining how much weight to give to any one study’s reported effects. A lower overall validity reflected greater uncertainty as to whether the findings were the result of chance or design. Full data extraction and evidence synthesis were only completed on medium–high and high quality studies.

Data were extracted by reading and recording details from each paper. The extracted data were used to build summary tables to inform evidence synthesis and to develop overall conclusions.

Data extraction was performed independently by two reviewers. Reviewer pairs were again rotated to reduce bias. Team members did not review articles they had consulted on, authored or co-authored. Differences in data extracted between reviewers were identified and resolved. In cases where agreement could not be reached, a third reviewer was consulted to ensure consensus was obtained. The team developed standardized data extraction forms based on existing forms and data extraction procedures [28, 29]. Reviewer pairs extracted data on: year of study; type of health care setting; study design; sample characteristics; length of follow-up; intervention characteristics; MSK health outcomes and whether those outcomes were self-reported, administrative or clinically-based; statistical analyses; covariates/confounders and study effects (see Table 4 for the complete list of data extraction questions). The review team decided to record the effects reported for the longest follow-up period when considering study findings. Reviewers reconsidered the methodological quality rating scores recorded in the Level 2 review during data extraction. Any quality rating changes that the reviewer pairs identified were proposed to the full team for consensus. Final ratings are documented in the methodological quality appraisal table (Table 6).
Table 4

Data extraction items

1. State the research question/objective.

2. State the primary hypothesis.

3. State additional hypotheses not listed in question #2.

4. Write the last name of the first author and the year of publication.

5. List the jurisdiction where the study was completed.

6. Describe what type of health care organization(s) that the study was conducted in.

7. List the job titles/classification of the participants in the study.

8. List the inclusion criteria described in the study.

9. List the exclusion criteria described in the study.

10. What is the study design?

11. What type of prevention intervention did the study investigate?

12. Describe all interventions evaluated.

13. Was there confirmation the intervention occurred?

14. How long after the intervention implementation did confirmation occur?

15. What was the duration of the intervention in months/days/hours?

16. Indicate the time period between the baseline measurement and all subsequent follow-up measurements.

17. Describe overall (study) group.

18. Describe the intervention group(s).

19. Describe the referent group(s).

20. When were potential covariates/confounders measured?

21. Select from this list all covariates/confounders that were evaluated for inclusion in the final analysis.

22. Provide a list of covariates/confounding variables that were controlled for in the final test of the intervention’s effectiveness.

23. Describe the differences in covariates/confounders for those that participated in the study vs. those that were invited but did not participate, if possible by experimental group.

24. Describe the differences in covariates/confounders for those that participated in the study vs. those that were lost to follow-up, if possible by experimental group.

25. Does the study use “administrative” records to collect measurements of MSK health outcomes?

26. Does the study use self-reported questionnaire records as completed by the employee to collect measurements of MSK health outcomes?

27. Does the study use clinical exams or clinical records as completed by the clinician to collect measurements of MSK health outcomes?

28. Was the population studied “fixed” or “open”?

29. What sources were used to “count” employee injuries?

30. How were employee hours collected?

31. Indicate at what level employee hours were ascertained and/or estimated.

32. Were injury rates calculated?

33. If injury rates were calculated, list the equation(s).

34. Did the study discuss how researchers handled any of the following special issues related to administrative record keeping: temporary or contract employees; employees who floated between units; turnover rate; reinjury to the same employee?

35. Check all body regions where symptoms were ascertained by questionnaire.

36. Describe when follow-up MSK health outcomes (symptoms) were measured.

37. Were MSK symptoms measured at the same time of day or shift?

38. Check all body regions where specific clinical disorders were ascertained by physical examination or laboratory test.

39. Was masking of physical assessment done?

40. Was a standard protocol used for the clinical exams?

41. Please check the types of final analyses done for testing the observed effects of the intervention.

42. Describe for each outcome of interest (MSK) the observed intervention effects.

43. Remark on the findings or enter information that is unique about the study that may not be adequately captured in the other DE questions.

Evidence Synthesis

The high level of heterogeneity in the reviewed studies required the use of a synthesis approach adapted from Slavin (1995) and others [26, 28, 30] known as “best evidence synthesis.” This approach considers the study’s quality, the quantity of studies using the same prevention strategy and the consistency of the findings (Table 5). A strong level of evidence exists when there are three high-quality studies with convergent effects. Our evidence synthesis guidelines were adapted from other IWH prevention intervention reviews [26, 28, 31, 32]. While the review team first used the evidence synthesis to answer the global question (“Do occupational safety and health interventions in health care settings have an effect on musculoskeletal health status?”), levels of evidence were also reviewed for intervention categories represented in the literature. The review team concluded that it would not be possible to calculate effect sizes due to the heterogeneity of interventions and outcomes.
Table 5

Best evidence synthesis guidelines

Level of evidence

Minimum quality

Minimum quantity

Consistency

Strong

High >80%

≥3 studies

All high-quality studies converge on the same findings

Moderate

Medium–high (60–79%)

≥2 studies

Majority of medium–high quality studies converge on the same findings

Mixed

Medium–high (60–79%)

≥2 studies

Medium–high and better quality studies have inconsistent findings

Insufficient

None of the above criteria are met

A further challenge to evidence synthesis was created because few studies declared a primary outcome or a small set of outcomes. Rather, multiple MSK outcomes were used to assess intervention effectiveness. Consequently, decision rules were created to allow for concise and meaningful evidence synthesis. A study with any positive results and no negative results on a single intervention was classified as a positive effect study. A study with both positive effects and no effects was also classified as a positive effect study (e.g. there was a positive effect on one outcome such as back pain, but no effect on another outcome, such as neck pain). A study with only no effects was classified as a no effect study. A study with any negative effects was classified as a negative effect study.

Finally, the review team agreed to not include studies in evidence synthesis when the study did not report a statistical test of the intervention’s effect on the MSK outcome. Statistical tests included a description of a test statistic (e.g. a chi square value) with a p-value, or presentation of confidence intervals. A statistical test was needed to provide statistical confidence that the observed intervention was due to the intervention’s effect, and not to chance.

Results

Literature Search and Relevant Documents Identification

The literature search identified 8,465 articles after the results from the different databases were merged and duplicates were removed (Fig. 1). The Level 1a review resulted in exclusion of 8,350 articles. Using the exclusion criteria in Table 2, two team members reviewed the full article leading to the additional exclusion of 67 articles. Eight articles were grouped with other articles that described results from the same study. This left 40 studies to be reviewed for methodological quality at Level 2. The 40 studies were each reviewed by two reviewers using the quality assessment questions (Table 3).
https://static-content.springer.com/image/art%3A10.1007%2Fs10926-010-9231-y/MediaObjects/10926_2010_9231_Fig1_HTML.gif
Fig. 1

Flowchart of systematic review process

Methodological Quality Appraisal

The 40 studies that met the relevance criteria were assessed for methodological quality and assigned a quality ranking score. The studies were placed into four quality categories: high (80–100%), medium–high (60–79%), medium (40–59%), and limited (less than 40%).

Two studies were classified as meeting the criteria to be determined as high quality [33, 34]. The high quality studies were quite consistent in their quality scores meeting 15 of the 19 criteria (85%). However, neither study stated a hypothesis nor adjusted for covariates in testing for the intervention effect. Only one of the two studies included at least a 3-month follow-up time or described contamination between groups.

We classified 14 studies as having met the criteria of medium–high quality (range 60–74%). These studies generally scored well on the following criteria: using concurrent comparison (control) group(s); describing sample inclusion/exclusion criteria; presenting baseline characteristics; and measuring covariates/confounders. However, few of these studies met the following criteria: stating a hypothesis (4/14); describing differences between participants and those lost to follow-up (3/14); indicating contamination between groups (4/14); and adjusting for covariates/confounders (4/14). The medium–high quality studies also did not meet the criteria of randomizing the intervention (5/14) and reporting participation rates ≥40% (8/14).

We classified 20 studies as having met the medium quality criteria (quality score ranging from 40 to 57%). Most of these studies presented baseline characteristics (19/20). Few of these studies met the following criteria: randomizing the intervention (2/20); describing differences between participants and those lost to follow-up (3/20); indicating contamination between groups (1/20); and adjusting for covariates/confounders (2/20). The medium quality studies often did not include a control or concurrent comparison group (9/20) or present baseline characteristics by group (7/20). Only four studies were classified as having met the criteria to be of limited quality (range 17–26%) (Table 6).
Table 6

Methodological quality assessment (QA) (n = 40)

Criteriaa

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

Quality ranking

Author/weight

2

1

3

1

2

4

4

2

2

3

2

3

3

3

3

1

3

2

3

 

High quality (H)

Harma, 1988

1

0

1

1

0

1

1

1

0

1

1

1

1

1

1

1

1

0

1

85%

Maul, 2005

1

0

1

1

1

1

1

1

1

1

1

1

1

0

1

0

1

0

1

85%

Criteria Met

2

0

2

2

1

2

2

2

1

2

2

2

2

1

2

1

2

0

2

 

% Criteria Met

100

0

100

100

50

100

100

100

50

100

100

100

100

50

100

50

100

0

100

 

Mediumhigh quality (MH)

Bru, 1994

1

0

1

0

1

1

1

1

0

1

1

1

0

0

0

1

1

1

1

72%

Carrivick, 2002

1

0

1

0

1

0

0

1

0

1

1

1

0

0

1

1

1

1

1

62%

Collins, 2004

1

1

1

0

1

1

0

1

0

1

1

0

1

0

0

0

1

1

1

64%

Dehlin, 1978

1

0

1

1

0

1

0

1

0

0

1

1

1

0

1

0

1

0

1

62%

Dehlin, 1981

1

1

1

1

0

1

0

1

0

0

1

1

0

0

1

1

1

0

1

60%

Donchin, 1990

1

0

1

1

1

1

1

1

0

1

1

1

0

0

1

0

1

0

1

74%

Gundewall, 1993

1

0

1

1

0

1

1

0

0

0

1

1

1

0

1

0

1

0

0

60%

LeClerc, 1997

1

1

1

0

1

1

0

1

0

1

1

1

0

0

1

0

1

0

1

66%

Li, 2004

1

0

1

1

1

1

0

1

0

1

1

0

1

0

1

0

1

0

1

66%

Linton, 1989

1

0

1

1

1

1

1

1

0

0

1

1

0

0

1

0

1

0

0

62%

Oldervoll, 2001

1

0

1

1

1

1

0

1

0

1

1

1

0

0

1

0

1

0

1

66%

Smedley, 2003

1

0

1

0

1

1

0

1

0

1

1

1

0

0

1

0

1

1

1

68%

Videman, 1989

1

1

1

1

1

1

0

1

0

0

1

1

0

0

1

0

1

0

1

62%

Yassi, 2001

1

0

1

0

1

1

1

1

0

0

1

1

0

0

1

0

0

0

1

60%

Criteria Met

14

4

14

8

11

13

5

13

0

8

14

12

4

0

12

3

13

4

12

 

% Criteria Met

100

29

100

57

79

93

36

93

0

57

100

86

29

0

86

21

93

29

86

 

Medium quality studies (M)

Alexandre, 2001

1

1

1

1

0

0

1

1

0

0

1

1

0

0

0

0

1

0

1

51%

Best, 1997

1

1

1

0

1

1

1

0

0

0

1

1

0

0

1

0

1

0

0

57%

Carrivick, 2002

1

0

1

0

1

0

0

0

0

1

1

0

0

0

0

0

1

1

1

43%

Davis, 2004

1

1

1

0

1

1

0

1

0

1

1

1

0

0

0

0

0

0

1

53%

Evanoff, 1999

1

1

1

1

1

1

0

1

0

1

1

0

0

0

0

0

1

0

1

55%

Evanoff, 2003

1

1

1

0

1

1

0

1

0

0

1

0

0

0

1

0

0

0

1

47%

Fanello, 1999

1

0

1

0

1

0

0

1

0

1

1

1

0

0

1

0

1

0

1

55%

Fujishiro, 2005

1

1

1

0

1

0

0

1

0

0

1

0

0

0

1

0

1

0

1

45%

Garg, 1992

1

0

1

0

1

0

0

0

0

1

1

0

0

0

1

0

1

0

1

45%

Lagerstrom, 1997

1

1

1

1

1

0

0

0

0

0

1

0

0

0

1

0

1

0

1

43%

Landstad, 2001

1

0

1

1

0

1

0

1

0

1

1

1

0

0

0

1

1

0

1

57%

Nassau, 1999

1

1

1

1

1

1

0

1

0

0

0

0

0

0

1

0

0

0

1

45%

Nelson, 2005

1

0

1

1

1

0

0

1

0

1

1

0

0

0

1

0

1

1

1

55%

Nevala, 2004

1

0

1

0

0

1

0

0

0

0

1

0

0

0

1

0

1

0

1

43%

Peterson, 2004

1

1

1

1

0

1

0

0

0

1

1

0

0

0

1

0

0

0

0

40%

Ronald, 2002

1

0

1

1

1

0

0

1

0

1

1

0

1

0

0

0

1

0

1

51%

Skargren, 1996

1

0

1

1

0

0

0

1

0

1

1

1

1

0

1

0

1

0

0

53%

Sobaszek, 2001

1

0

1

1

1

0

0

1

0

1

1

0

0

0

0

0

1

0

1

45%

Tiesman, 2003

1

0

1

1

1

0

0

1

0

0

1

0

0

0

1

0

1

0

1

45%

Yassi, 1995

1

0

1

0

1

1

0

1

0

1

1

1

1

0

0

0

0

0

1

57%

Criteria Met

20

9

20

11

15

9

2

14

0

12

19

7

3

0

12

1

15

2

17

 

% Criteria Met

100

45

100

55

75

45

10

70

0

60

95

35

15

0

60

5

75

10

85

 

Limited quality studies

Charney, 1997

1

0

1

0

0

0

0

1

0

0

0

0

0

0

0

0

0

0

1

21%

Guthrie, 2004

1

0

1

0

0

0

0

0

0

0

0

0

0

0

1

0

0

0

0

17%

Lynch, 2000

1

0

0

0

0

1

0

0

0

0

0

0

0

0

1

0

0

0

1

26%

Ryden, 1988

1

0

1

0

1

0

0

1

0

0

0

0

0

0

0

0

0

0

1

26%

Criteria Met

4

0

3

0

1

1

0

2

0

0

0

0

0

0

2

0

0

0

3

 

%Criteria Met

100

0

75

0

25

25

0

50

0

0

0

0

0

0

50

0

0

0

75

 

aRefer to Table 3 for the QA criteria

Data Extraction and Evidence Synthesis

The review team extracted data from the 16 studies of high or medium–high quality. The review team grouped the various interventions described in the studies to create intervention categories. Table 7 shows the intervention categories and descriptions, study design and whether the intervention was designed for primary or secondary injury prevention (additional data from the studies reviewed can be found in a detailed report of this review available at http://www.iwh.on.ca/research/sr-program.php).
Table 7

Description of interventions

Author, year

Intervention description

Study design

Prevention type

QA

Multi-component patient handlinga

Collins, 2004

I1: zero lift policy; mechanical lifting equipment and repositioning aids; lift equipment training and medical management

Pre-post w/statistical control

Both

MH

Smedley, 2003

I1: revised manual-handling policy; equipment: sliding sheets, hi/lo baths, hoists and transfer belts; two-day health and safety training

C: no policy change; a non-formal ergonomic program—usual

NR field trial

Both

MH

Yassi, 2001

I1: “safe-lift” policy; lifting and transfer equipment; 3 h of education on back care, patient assessment and handling techniques

I2: “no strenuous lifting” policy; new mechanical patient lifts and transfer equipment on each ward; 3 h of education on back care, patient assessment and handling techniques

C: no policy change; one mechanical total body lift available on the ward and access to sliding devices from a central equipment depot on request only; no training provided

RFT

Both

MH

Exercise training

Dehlin, 1978

I1: muscle training (exercise)

C1: lectures on geriatric medicine and nursing care

C2: no training or lectures

C3: asymptomatic exposure parameter controls; no training or lectures

NR field trial

Secondary prevention

MH

Gundewall, 1993

I1: exercise to increase dynamic endurance, isometric strength and functional coordination

C: no exercise

RFT

Both

MH

Harma, 1988

I1: physical training (exercise)

C: no exercise

RFT

Both

H

Maul, 2005

I1: low back school (three sessions 1 h each) and physical exercises

C: low back school and no exercise

RFT

Secondary prevention

H

Oldervoll, 2001

I1: endurance (aerobic capacity promoting) training twice a week for 17 weeks

I2: strength promotion (SP) classes twice/week for 17 weeks

C: wait listed controls

NR field trial

Secondary prevention

MH

Patient handling training

Videman, 1989

I1: increased practical patient handling training (ergonomic training)

C: traditional patient handling training

NR field trial

Both

MH

Exercise training, patient handling training

Dehlin, 1981

I1: physical fitness training (exercise)

12: ergonomic education on lifting technique

C: no training or ergonomic education

NR field trial

Secondary prevention

MH

Back school, exercise training

Donchin, 1990

I1: calisthenics (exercise)

I2: back school

C: wait listed controls

RFT

Secondary prevention

MH

Cognitive behavioral training

Bru, 1994

I1: cognitive-behavioral training

I2: relaxation training

I3: combined I1 and I2

C: wait listed controls

RFT

Secondary prevention

MH

Intensive off-site MIPP

Linton, 1989

I1: low back program at an off-site clinic & hotel (exercise training, ergonomics training, behavioral training on pain management)

C: wait listed controls

RFT

Secondary prevention

MH

Equipment and equipment training

Li, 2004

I1: one portable full body sling lift, two portable stand-up sling lifts (‘‘E-Z Lift’’ and ‘‘E-Z Stand’’ by EZ Way Inc, Minneapolis, Minnesota) and friction reducing sheets (Maxislides) and training sessions in lift usage

C: no lifts purchased

NR field trial

Both

MH

Participatory ergonomics team

Carrivick, 2001

I1: participatory ergonomics team of cleaners identified, assessed and recommended controls of manual handling

C1: hospital orderlies not receiving the intervention

C2: cleaners from another hospital not receiving the intervention

C3: hospital and non-hospital cleaners in the state insurance system

NR field trial

Both

MH

Broad-based MIPP

Leclerc, 1997

I1: training with exercise and ergonomic changes following a site visit by an ergonomist

C: usual injury prevention policies

NR field trial

Both

MH

aMulti-component patient handling—an intervention which included three components: organizational policy change, lift or transfer equipment purchase and training on equipment usage or patient handling

MIPP MSK injury prevention program, I intervention group, C control group, RFT randomized field trial, NR, non-randomized, MSK musculoskeletal

The most common interventions were multi-component patient handling (3 of 16) and exercise training (5 of 16) interventions. The team agreed that “multi-component patient handling” interventions (MCPHI) included three components: an organizational policy that defines an organizational commitment to reducing injuries associated with patient handling, purchase of the appropriate lift or transfer equipment to reduce biomechanical hazards and a broad-based ergonomics training including safe patient handling and/or equipment usage. Most exercise interventions intended to reduce disabling outcomes associated with an MSK injury. A majority of the interventions involved some type of training (13 of the 16 studies). The remaining interventions (n = 8) were evaluated by single studies (see Table 7). Substantial heterogeneity was observed within the intervention categories for the specific equipment employed, training methods used and intervention protocols followed.

Nine studies were both primary and secondary prevention trials where participants were not excluded based on symptoms or disorders. Seven studies were secondary prevention only, where only subjects who reported symptoms or diagnosed disorders were included. Only seven of the sixteen studies had randomized intervention assignment.

Some of the study characteristics that were considered important when examining comparability and generalizability are shown in Table 8. The most common geographic setting was Sweden (n = 4). Two studies each were from the U.S., Norway and Finland. Each of the following countries contributed one study: U.K., France, Canada, Israel and Australia. The primary job titles studied were nurse, nursing aide, nursing assistant and licensed practical nurses.
Table 8

Study description for full data extraction studies (n = 16)

Author, year

Health care setting (country)

Job titles

Study design

Population description

Participation rate

Sample size

Multi-component patient handlinga

Collins, 2004

Nursing homes (U.S.)

Certified nursing assistants, registered and licensed practical nurses, physical therapists, restorative aides

Pre-post w/statistical control

Open

100%

NP

Smedley, 2003

Public hospitals (England)

Nurses

NR field trial

Open

56%

N = 1,239

I1 = NP

C = NP

Yassi, 2001

Acute and tertiary care hospital (Canada)

Nurses, unit assistants

R field trial

Admin: open

Self-report: fixed

NP

N = 346

I1 = 116

I2 = 127

C = 103

Exercise training

Dehlin, 1978

Geriatric/long-term care hospital (Sweden)

Nursing aides

NR field trial

Fixed

NP

N = 66

I1 = 18

C1 = 14

C2 = 14

C3 = 20

Gundewall, 1993

Wards of a geriatric hospital (Sweden)

Nurses, nurses’ aides

R field trial

Fixed

NP

N = 69

I1 = NP

C = NP

Harma, 1988

University hospital (Finland)

Nurses, nursing aides

R field trial

Fixed

79%

N = 119

I1 = 76

C = 43

Maul, 2005

University hospital (NP)

All hospital employees

R field trial

Fixed

51%

N = 148

I1 = 74

C = 74

Oldervoll, 2001

University hospital (Norway)

Registered nurses, auxiliary nurses, laboratory staff, administration staff, cleaning department staff

NR field trial

Open

54%

N = 65

I1 = 24

I2 = 22

C = 19

Patient handling training

Videman, 1989

Nursing school (Finland)

Student nurses

NR field trial

Self-report: unclear

NP

N = 308

I1 = 151

C = 157

Exercise training, patient handling training

Dehlin, 1981

Geriatric hospital (Sweden)

Nurses’ aides

NR field trial

Fixed

NP

N = 45

I1 = 15

I2 = 14

C = 16

Back school, exercise training

Donchin, 1990

University hospital (Israel)

Clinical, administrative, technical employees

R field trial

Fixed

70%

N = 142

I1 = 46

I2 = 46

C = 50

Cognitive behavioral training

Bru, 1994

Hospital (Norway)

Physicians, registered nurses, auxiliary nurses, laboratory staff, kitchen staff

R field trial

Fixed

71%

N = 119

I1 = NP

I2 = NP

I3 = NP

C = NP

Intensive off-site MIPP

Linton, 1989

Hospital (Sweden)

Licensed practical nurses, nursing aides

R field trial

Fixed

NP

N = 66

I1 = 36

C = 30

Equipment & equipment training

Li, 2004

Community hospital (U.S.)

Nurses, nursing assistants, patient care attendants

NR field trial

Admin: open

Self-report: fixed

Admin: NP

Self-report: 44%

N = NP

I1 = 61

C = NP

Participatory ergonomics team

Carrivick, 2001

Hospital (Australia)

Cleaners, orderlies

NR field trial

Open

100%

N = NP

I1 = 507

C1 = 279

C2 = NP

C3 = NP

Broad-based MIPP

Leclerc, 1997

Hospital (France)

Nurses and auxiliary nurses

NR field trial

Admin: open

Self-report: fixed

90%

NP

aMulti-component patient handling—an intervention which included three components: organizational policy change, lift or transfer equipment purchase and training on equipment usage or patient handling

MIPP MSK injury prevention program, NP not provided, I intervention group (subscripts indicate number of groups, e.g. I2), R field trial randomized field trial, C control group (subscripts indicate number of groups, e.g. C2), NR field trial non-randomized field trial, MSK musculoskeletal, N sample size, Fixed followed same participants over time, Open participants allowed to enter and/or leave during study

The sample sizes tended to be small and varied from 45 [35] to 1,239 [36]. Eight studies had a fixed population (employees were not allowed to enter the study after a specific date) and followed the same participants over time. Four studies had open populations (employees being tracked at the end of study were not the exact same employees present at the beginning of study) where participants were allowed to enter and/or leave during the study period. Participation rates were not provided in almost half of the studies (7/16). When participation rates were provided they varied between 44 and 100%.

A summary of the intervention effects are presented in Table 9. Two studies, Gundewall (1993) and Videman (1989), are not included in the evidence synthesis because they did not report statistical tests of intervention vs. control group effects [37, 38]. The review team did not find any negative or adverse effects. Therefore, we consistently report positive effects or no effects. The evidence is summarized by intervention category.
Table 9

Intervention effects on MSK health outcomes as reported in evidence synthesis studies (n = 14) [refer to key at end of table for abbreviations]

Author, yeara

Outcome

Effect (direction of effect°, comparison) on MSK health outcomes

°even if non-significant

Length of follow-up

QA

Multi-component patient handlingb

Collins, 2004

Admin: “Resident handling” injury rates from OSHA logs, employer records and WC; also “Resident Handling” LWD rate and RWD rate from OSHA logs

Positive effect for resident handling injury incidence from OSHA (decrease in I1 vs. C, where C is “all other injuries”) Positive effect for resident handling WC rate, injury rate from employer records, LWD rate and RWD rate (decrease in post- vs. pre-intervention)

36 months

MH

Smedley, 2003

Self-report: low back pain

No effect on back pain prevalence (increase in post- vs. pre-intervention)

NP

MH

Yassi, 2001

Admin: “Patient handling” injury rates from WC Self-report: low back pain and shoulder pain

All interventions: No effect on injury rates (decrease in I1, I2 vs. C at 12 months) “Safe Lifting”: Positive effect on low back pain and shoulder pain (decrease in I1 vs. C at 12 months) “No Lift”: No effect on low back pain and shoulder pain (decrease in I2 vs. C at 12 months) “No Lift”: Positive effect on low back pain and shoulder pain (decrease in I2 vs. C at 6 months) “Safe Lifting”: No effect on low back pain and shoulder pain (decrease in I1 vs. C at 6 months)

12 months

MH

Exercise training

Dehlin, 1978

Self-report: low back insufficiency—driven by symptoms

Positive effect for low back insufficiency (decrease in I1 vs. C1) No effect for low back insufficiency (decrease I1 vs. C2)

NP

MH

Harma, 1988

Self-report: MSK symptom index

Positive effect for MSK symptom score (decrease in I1 vs. C)

NP

H

Maul, 2005

Self-report: pain assessed by Nordic questionnaire of low back pain; low back pain intensity (numeric rating scale, pain drawing and medication use); McGill pain characteristics scale; perceived treatment effectiveness

Positive effect on perception of therapy reducing pain (decrease in I1 vs. C1 at 120 months)

No effect for current pain in comparison with pre-treatment pain (decrease in I1 vs. C at 120 months)

Positive effect on pain drawing (decrease in I1 vs. C at 12 months)

No effect for numeric rating scale of pain or McGill pain characteristics (decrease in I1 vs. C at 12 months)

120 months

H

Oldervoll, 2001

Self-report: pain index

Positive effect on pain index (I1 and I2: decrease in post- vs. pre-intervention at 7 months)

All interventions: Positive effect on pain index (I1, I2 vs. C at 0 months)

 Endurance Training: decrease in post- vs. pre-intervention

 Strength Promotion: decrease in post- vs. pre-intervention

7 months

MH

Exercise training, patient handling training

Dehlin, 1981

Self-report: low back insufficiency—driven by symptoms

All interventions: No effect on low back insufficiency (I1, I2 vs. C)

NP

MH

Back school, exercise training

Donchin, 1990

Self-report: duration, rate and type of low back pain episode

All interventions: Positive effect for low back pain duration (I1, I2, vs. C)

 Exercise Training: decrease in I1 vs. C Back School: no difference between I2 vs. C

12 months

MH

Cognitive behavioral training

Bru, 1994

Self-report: pain intensity and pain duration in neck, shoulders and low back

All interventions: Positive effect (I1, I2, I3, vs. C) on pain intensity; body area affected varied by intervention

 Cognitive training: decrease neck and shoulders in I1 vs. all others

 Relaxation training: decrease low back in I2 vs. all others

All interventions: No effect (I1, I2, I3, vs. C) on pain duration

4 months

MH

Intensive off-site MIPP

Linton, 1989

Self-report: low back pain intensity

Positive effect on low back pain intensity (decrease in I1 vs. C)

6 months

MH

Equipment & equipment training

Li, 2004

Admin: “lifting” MSK injury rate and LWD rate from OSHA 200 log and from WC

Self-report: MSK discomfort in: neck, shoulders/upper arm, upper back, lower back, forearm, wrist/hand, hips/buttocks, knees and feet/ankles

No effect for “lifting” MSK injury rates, LWD rates from OSHA & WC (decrease in I1 vs. C)

Positive effect for MSK discomfort in all areas (decrease in post- vs. pre-intervention)

Admin: 24 months

Self-report: 1 month

MH

Participatory ergonomics team

Carrivick, 2001

Admin: lost-time injury rate

Positive effect for lost time injury rate (decrease in post- vs. pre-intervention) Positive effect for manual handling lost-time injuries (decrease in post- vs. pre-intervention)

36 months

MH

Broad-based MIPP

Leclerc, 1997

Self-report: MSK pain intensity and duration in low back, upper back, neck and shoulders

Positive effect for combined spine and shoulder disorder, upper back disorder and lower back disorder scores (decrease I1 vs. C) No effect for neck disorder or shoulder disorder scores (increase I1 vs. C)

NP

MH

aGundewall 2003 and Videman 1989 not included in evidence synthesis

bMulti-component patient handling—an intervention which included three components: organizational policy change, lift or transfer equipment purchase and training on equipment usage or patient handling

OSHA occupational safety and health administration, WC workers’ compensation, VAS visual analog scale, QA quality appraisal, LWD lost work days, MSK musculoskeletal, MH medium–high, I intervention group, RWD restricted work day, MIPP MSK injury prevention program, H high, C control group

Multi-component patient handling intervention (MCPHI) Three studies of medium–high quality evaluated interventions that included all three components [13, 36, 39]. Two showed positive effects [13, 39]. Therefore, we concluded there was moderate evidence that a MCPHI had a positive effect on improving MSK health.

Exercise training Six studies evaluated exercise training. Two were high quality [33, 34] and four were medium–high quality [35, 4042]. Four described “physical fitness” or “calisthenics” while two described exercises that specifically improved strength and/or endurance. Two high quality and four medium–high quality studies showed positive effects on MSK outcomes. We concluded there was moderate evidence that exercise training interventions had a positive effect on improving MSK health.

For the majority of the remaining interventions there was only a single study. These included: back school; exercise & patient handling & stress management; intensive off-site MSK injury prevention programs; participatory ergonomics; equipment & patient handling training; equipment & equipment training; broad-based MSK injury prevention program; ergonomic devices & consultation; shower trolleys; integrated disability management; and pre-employment screen & RTW policy. With a single study, there was insufficient evidence to determine whether the interventions had an effect on MSK outcomes.

Due to the time lapsed from the original search to publication; an updated search was run to ensure the level of evidence had not changed since the original project. A new search limited to the publication years 2006–2009 was run in 2009 in all the databases except for CINAHL. This search identified 2,918 articles (excluding duplicates). Two reviewers reviewed the titles and abstracts and then, where abstracts were insufficient, a small subset of full articles were reviewed. Following the review of articles ten relevant studies remained [4352] to be quality assessed. Two studies were of medium–high quality [50, 52] and one was high quality [49] meeting the evidence synthesis quality cut-point. The three studies tested two interventions found in the original review: patient handling training [49, 52] and cognitive behavioral training [49, 50]. One of the studies also evaluated a new intervention type, patient handling training and exercise [52]. None of the studies found significant effects and consequently all three were identified as no effect studies. The results of the new search added evidence rather than changing the levels of evidence found in the original search. With the updated literature, there is moderate evidence that patient handling training alone has no effect on musculoskeletal injuries [49, 52]. There is also moderate evidence that cognitive behavioral training alone has no effect on musculoskeletal injuries [49, 53]. Further information on the updated search is available from the first author.

Discussion

Our systematic review sought to answer the question: “Do occupational safety and health interventions in health care settings have an effect on MSK health status?” Across the 16 high and medium–high quality studies (19 including the update), we found a moderate level of evidence supporting a broad range of occupational safety and health interventions that reduce MSK injuries and the consequences of having an injury in health care settings. A moderate level of evidence means that a majority of high and medium–high studies found similar effects on MSK outcomes. Additionally, we found no evidence that any intervention had a negative or deleterious effect.

Next, we considered the effectiveness of specific intervention types. A moderate level of evidence was found for a positive intervention effect on improving MSK health for:
  • multi-component patient handling interventions (MCPHI)

  • physical exercise interventions.

Finally, two additional findings were determined as a result of the updated search. A moderate level of evidence was found that neither patient handling training nor cognitive behavior training alone have an effect on improving MSK health.

Many interventions were unique, meaning only one study per intervention was found. Therefore, the available evidence was insufficient to make statements about intervention effectiveness in health care settings for: back school; exercise & patient handling & stress management; exercise & patient handling training; intensive off-site MSK injury prevention programs; participatory ergonomics; equipment & patient handling training; equipment & equipment training; broad-based MSK injury prevention program; ergonomic devices & consultation; shower trolleys; integrated disability management; and pre-employment screen & RTW policy.

As for the interventions with moderate evidence, the review results are comparable to findings from other reviews that MCPHI were effective in reducing MSK injuries [10, 11]. Neither Hignett (2003) nor Dawson (2007) identified key characteristics of a MCPHI. Our results highlight three key components: an organizational policy aimed at reducing injuries associated with patient handling, purchase of lift or transfer equipment and a broad-based ergonomic training including safe patient handling and/or equipment usage. Because the studies found were only medium–high quality, we can only state a moderate level of evidence. Also similar to our review, multiple reviews have reported that there was little evidence that patient/material handling training alone was effective [1012, 54]. In our review, an insufficient number of studies tested equipment purchases alone even though biomechanical studies exist supporting the importance of equipment design in improving MSK health [55, 56]. Additionally, to say that one component could be as effective as the three combined, the performance of each component on reducing injuries would have to be compared to the combined MCPHI which to our knowledge, has not been done.

We found moderate evidence supporting physical exercise interventions to reduce MSK symptoms and disorders. There was substantial heterogeneity in the physical exercise interventions ranging from muscle training [40] to endurance training [42]. The heterogeneity could be related to the changing recommendations made by professional associations (e.g. Physical Therapy Associations) about the most appropriate type of exercises (e.g. cardiovascular vs. strength) over the 27 years during which the research was conducted and published. Most studies used self-reported MSK symptoms or disability measurements. We cannot predict whether an exercise intervention would result in either a change in the incidence of clinical MSK symptoms or injury reporting; both are important outcomes for stakeholders and policy holders.

Our review included studies in different health care settings including long-term care organizations and hospitals. As previously mentioned, differences in exposure might exist between settings due to characteristics including organization, economics and patient population. We were unable to stratify effects between hospitals and long-term care settings because only one study that was included in evidence synthesis took place in a nursing home setting [13].

The high and medium–high quality studies reviewed shared common traits, regardless of the intervention or outcome. They all had concurrent comparison groups or a statistical comparison group. Each study was designed to limit threats to internal and external validity. However, few studies in evidence synthesis used similar MSK outcomes. This challenged the review team when trying to integrate findings and made it impossible to calculate pooled effect sizes for interventions.

The review included a number of strengths. Stakeholders provided input to insure that messages would be relevant to occupational safety and health practice. The literature reviewed included articles from content experts and multiple languages other than English (i.e. French, Spanish and Swedish). Consensus decisions were made regarding administrative MSK outcomes which allowed these types of studies to be included in the review. The review team supports a number of review process improvements. First, more non-English articles and gray literature should be included in the process. Second, other review teams should continue to develop transparent and peer-reviewed methods to evaluate studies using administrative MSK outcomes. In the future, article authors should be contacted, if necessary, to clarify findings in the published studies. When possible, studies in which between-group comparisons were not made should be re-analyzed to provide evidence that can be included in data synthesis. In order to produce effect sizes, a full data set should be obtained from researchers, if possible.

While insufficient evidence was found for many intervention types, the term “insufficient” refers to the low number of studies, not the intervention quality. In fact, the review team supports more high quality studies of these interventions. Researchers, funders, employers and organized labor should attend to the effects and study quality reported in Table 9 as one way to gauge where to invest in future research.

Recommendations for Practice

Prior to making policy and best practice recommendations, the review team felt there should be a stronger level of evidence. Such recommendations require consistent findings from at least three high quality studies. The review did not find this level of evidence. With only a moderate level of evidence, the review team considered it feasible to recommend “practices to consider.”

The first practice to consider is multi-component patient-handling interventions. The intervention components include: an organizational policy aimed at reducing injuries associated with patient handling, purchase of lift or transfer equipment and a broad-based training on safe patient handling or equipment usage. Because the group of three intervention components is bundled, the team cannot advise stakeholders on whether one specific intervention on its own may be as good as the bundle.

The second practice to consider is exercise training programs, both aerobic and strength building. A further advantage to physical exercise is that it improves general health and reduces the risk of many chronic diseases. The review team would encourage convening a group who specializes in exercise training techniques to assist researchers in developing well designed interventions. The specialty group would provide a valuable resource in identifying, differentiating and assessing the heterogeneous exercise interventions. These findings may seem to support a recent article by Bigos that reported exercise interventions were effective at reducing back pain in working adults [57]. The review team would caution that the results of the Bigos review do not add further evidence for our review findings. The Bigos research question, study inclusion criteria, and methods of reporting results were more restrictive than our review. In addition, the Bigos review only included those studies where participants were free of back pain at study inception. We believe the further evaluation of primary, secondary and tertiary interventions is important as the prevalence of low back pain in nursing personnel has been reported at rates between 30 and 60% [36].

An additional message following our updated review is that workplaces should not consider implementing patient handling or cognitive behavioral training as isolated interventions. These interventions, when implemented alone, have been found to show no effects on MSK outcomes.

A strong level of evidence was not found supporting the MSK health benefits of interventions implemented in hospitals, long-term care facilities and other health care establishments. However, the team does strongly encourage continuing the evaluation of MCPHI with clearly defined components. While our review only found a moderate level of evidence for MCPHI, a systematic review of Injury/Illness and Loss Control Programs found strong evidence that multi-component programs had a positive effect in reducing the severity and costs of workplace injuries in general [58]. While the Brewer review did not focus solely on MSKs and healthcare, it further demonstrates that multi-component interventions can be evaluated when interventions are defined in terms of programs, policies, and practices [58]. Stakeholders informed us that changes in programs, policies and practices most often occur at the same time so it makes sense to evaluate them as a multi-component whole rather than analyzing the components as independent parts.

The review team was surprised to find only moderate evidence given the current level of investment in lifting equipment. The results of the review seem to reinforce the statement that “commonly accepted wisdom is not necessarily evidence” [59]. The challenge to researchers is to complete well designed studies that provide strong enough evidence to accept or reject the commonly accepted wisdom.

Acknowledgments

The authors wish to thank: Dan Shannon for obtaining bibliographic information and other materials; Jim Collins for providing references and input; Jane Gibson, Anita Dubey and Kiera Keown, and Jaime Wertman for their editorial advice; and Shanti Raktoe for administrative support. Jessica Tullar and Shelley Brewer were supported by an Occupational Injury Prevention Training Grant (T42 OH008421) from the U.S. National Institute for Occupational Safety and Health. The Prevention Reviews Program at the Institute for Work & Health has been supported by a grant from the Ontario Workplace Safety and Insurance Board. For the full report please visit the IWH web site at www.iwh.on.ca.

Copyright information

© Springer Science+Business Media, LLC 2010