Table 1: Evidence summary - healthcare workers and SPH.

First author and year Summary of study Evidence
Intervention involving just device use
Trinkoff et al. 2003 [42] Randomly selected working nurses surveyed (N=1163) via an anonymous mailed survey in the US. Mechanical lifting device availability was associated with a significantly reduced likelihood of neck and back MSDs; back injury was less likely when lifting teams were available. It could be argued that data collection by anonymous mailed survey may bias the study population towards nurses with injury or concerns relating to SPH.
Evanoff et al. 2003 [7] Review of injury rates and lost workdays both before and after mechanical lift in 4 acute care hospitals and 5 long-term care facilities (36 intervention units) in the US. The study population was followed for 5 years (including 2-3 years post-intervention). MSD rates, lost workday injuries, and days lost due to injury were reduced post-intervention. Higher self-reported lift use frequency in the long-term care facilities (also greater MSD rate reductions, compared with hospitals). Correlation seen between higher self-reported lift use and reductions in injury and lost day injury rates.
Li et al. 2004 [41] Effectiveness of mechanical patient lifts studied pre- and post-intervention in workers at a US community hospital (N=36). Reduced injury rate and symptoms relating to MSDs reported post-intervention; statistically significant improvement in MS comfort (p<0.05) reported for all 9 body parts surveyed - this included shoulders, lower back and knees. The study limitations include the pre-post study design, and a small sample size.
Chhokar et al. 2005 [45] Assessment of overhead lifts in an extended care facility in Canada, with analysis of injury trends over 3 years pre-intervention, and 3 years post-intervention. Significant and sustained decrease found post-intervention in patient handling injury-related direct costs (p=0.034) and number of days lost (p=0.024). The number of patient-handling claims decreased from 65 pre-intervention to 47 post-intervention, and the number of all MS injury claims decreased significantly (p=0.006). Difficult to differentiate between the effects of device implementation, and of an injury reduction and training programme overall. No formal, randomised control group included.
Miller et al. 2006 [44] Assessment of ceiling lift introduction. Matched pre- and post-intervention questionnaires comparing patient handling injuries were completed by front-line care staff from intervention (n=17) and control (n=15) facilities in Vancouver, Canada. Staff perception of injury risk was significantly lower (p<0.05) when ceiling lifts were used compared with manual transfer methods; 82% of staff believed that lifting patients was easier to perform using ceiling lifts compared with other methods such as floor lifts.
Koppelaar et al. 2012 [46] Ergonomic device use assessed by 186 nurses on 735 patient handling activities in 17 nursing homes in the Netherlands, in real time, using a structured patient handling programme centred on the presence of a ward Ergo or Transfer Mobility coach (a specialist trained in ergonomic principles who is responsible for supporting this process). Reported ergonomic device use of 69%; use of anti-embolism stocking slide and lifting devices associated with lower frequency of exerted forces, and adjustable bed and shower chairs with shorter awkward back posture duration. Device use also associated with less forceful movements and fewer awkward back postures; using lifting devices during patient transfers reduced exerted forces by two-thirds. The nursing homes included were not selected at random, and targeted for coach employment.
Knibbe et al. 2012 [47] Analysis of 213 horizontal patient transfers (87% heavy/very heavy) in the Netherlands. Introduction of stretcher slings increased the number of safe transfers performed from 29.5% to 83.6%, reducing the number of nurses required.
Burdorf et al. 2012 [43] A literature analysis of patient lifting device use in healthcare settings. A Markov decision analysis model was used to assess studies on the effect of manually lifting patients, and on introducing lifting devices on lower back pain and related injury claims. Implementation of lifting devices was found to reduce lower back pain and related claims, but the best scenario only showed a maximal reduction of 14% in lower back pain prevalence, and of MSD injury claims from 5.8 to 5.6 per 100 work years. The projection from this analysis was that complete elimination of manual patient lifting would reduce lower back pain by 10.5%, and MSD injury claims to 4.3 per 100 work years.
Holtermann
et al. 2015 [48]
A prospective cohort study of female healthcare workers (elder care services; N=1,478), with no reported lower back pain at baseline. Questionnaires on assistive device use were sent to participants; screening was post-response. In those workers occasionally using assistive devices, the multi-adjusted OR for developing infrequent lower back pain was 1.21 (95% CI 0.90-1.62), compared with an OR of 1.78 (95% CI 1.19-2.66) for those rarely using assistive devices. A significant trend was found between frequency of assistive device during patient handling and the risk for infrequent lower back pain (p < 0.01), with rare use associated with an increased infrequent lower back pain risk; no increased risk for frequent lower back pain was found.
Intervention involving device use and/or ergonomic strategy
Garg et al. 1992 [13] Prospective, epidemiologic study reviewing an ergonomic intervention strategy to reduce back stress in healthcare providers. Conducted in two units of a US nursing home (140 beds; 57 nursing assistants). Decrease in incidence rate for back MSDs from 83 per 200,000 work-hours pre-intervention, to 47 per 200,000 work-hours post- intervention.
Lynch and Freund 2000 [14] Analysis of a one-year Back Injury Prevention Programme at an acute care hospital in the US, with ergonomic evaluation of patient handling, pilot testing and purchase of new equipment, a train-the-trainer programme, and training of approximately 50% of nursing staff (374 nurses, and other patient handling staff). Efficacy was evaluated using self-reported knowledge, work practices and back pain in a subset of trainees and controls. Risk factor knowledge was increased post-implementation; a marginal increase was seen in mechanical device use for patient transfer, while repositioning of patients in bed was significantly decreased (p=0.017). Back MSDs were reduced by 30% from the average of the previous 3 years, with the number of reported injuries reduced in the quarter immediately following the training programme.
Brophy et al. 2001 [15] Study reporting on implementation of a 5-step ergonomics programme in a US nursing home over a 7-year period, with the aim of changing behaviour across all levels. Health and financial outcomes were compared pre- and post-intervention. Mean number of MSDs was significantly reduced – from 15.7 per 100 pre-intervention, to 11.0 per 100 post-intervention (p=0.05). Number of lost workdays was also significantly reduced, from 1476/year pre-intervention to 625/year post-intervention (p=0.05).
Yassi et al. 2001 [5] Three-armed RCT (control, safe lifting and no strenuous lifting) on the effectiveness of training and equipment in reducing MSD injury in staff undertaking patient handling (N=346 [nurses and unit assistants]) in a Canadian acute care hospital. Training in handling techniques, patient assessment and back care received by both intervention arms; the ‘no strenuous lifting’ arm also aimed to eliminate manual patient handling by using mechanical and other assistive equipment. The frequency of manual handling tasks was significantly decreased in the ‘no strenuous lifting’ arm (p<0.001); self-perceived work fatigue, back and shoulder pain, safety, and frequency and intensity of physical discomfort associated with patient handling tasks improved in both intervention arms - with greater improvement in the ‘no strenuous lifting’ arm.
Johnsson et al. 2002 [49] A training programme in patient handling and moving skills was evaluated (N=51 healthcare providers in Sweden), based on video-recordings before and after training, plus questionnaire both before and 6 months post-training (physical exertion, job strain, and MSD problems). No significant decrease in MSDs was found post-training, but physical exertion following bed-to-chair transfer was reduced after training programme completion, and 98% of participants positively reported on their participation in training, which also led to improved work technique.
Owen et al. 2002 [6] An intervention study and follow up in two US hospitals (ergonomic programme with assistive patient handling devices [n=37] and control [n=20]) on perceived exertion felt by nursing staff undertaking patient handling tasks. Data collection forms were used to assess injury data and lost and restricted workdays following programme implementation. Mean differences in perceived exertion to shoulder and to lower back were statistically significant between the control and experimental hospitals (p<0.001), with a concomitant decrease in the number of back MSDs (from 20 pre-intervention, to 12 post-intervention) and of lost/restricted work days (from 64 pre-intervention, to 3 post-intervention).
Collins et al. 2004 [16] Six-year intervention trial of a back injury prevention programme in 1728 nursing personnel, in six US nursing homes. Significant reduction in incidence of patient handling-related injury (p<0.05), with a reduction also in lost work day injuries (from 5.8 to 2.0 per 100 personnel) following programme implementation.
Hartvigsen
et al. 2005 [50]
Evaluation of an education and low-tech ergonomic intervention programme in 345 Danish nurse’s aides and home care nurses. The ergonomic programme was no more effective than basic education in reducing or preventing lower back pain, but any education in training transfer techniques was associated with an improvement in back pain, and was considered by participants as helpful.
Fujishiro et al. 2005 [17] Evaluation of MSD rate changes between baseline (1 year pre-intervention) and post-intervention (up to 2 years) in 100 work units in 86 healthcare facilities in the US (73 nursing homes, 10 MR/DD facilities and 3 hospitals) in a programme for ergonomic consultation and financial support for ergonomic device purchase, with a 2-year data collection follow-up period. Comparison with BLS data used instead of controls. Median MSD rate decreased post-intervention to 6.64 per 200,000 employee-hours, from 12.32 per 200,000. Each intervention type (reduction of bending, elimination of lifting, reduction of lifting, and a combination of the three) was associated with reduced MSD rates, and post-intervention MSD rates were considerably lower than comparable BLS-reported national rates.
Nelson et al. 2006 [12] Implementation of an ergonomics programme incorporating evidence-based practice, technology and safety improvement, evaluated prospectively in 23 high-risk units (19 nursing home care units and 4 spinal cord injury units) in 7 facilities in the US. Injury rates, lost work days, modified work days, job satisfaction, staff and patient acceptance, programme effectiveness, and programme costs/savings were compared over a nine-month pre-intervention period, and a nine-month post-intervention period, in 875 nursing staff. MSDs and the number of modified duty days taken per injury decreased significantly (p=0.036 and p=0.02, respectively), with post-intervention injury rate reduced in 15 of the 23 units. An 18% decrease was seen in the total number of lost work days. The number of ‘unsafe’ patient handling practices performed daily was self-reportedly decreased significantly (p=0.027), with equipment being rated by nurses as the most effective programme element, followed by a No Lift Policy.
Charney et al. 2006 [51] Data on patient handling injuries was compared before and after implementation of a ‘zero-lift’ programme in 31 US hospitals (replacement of manual patient lifting, transferring and repositioning with mechanical lifting/other devices). Patient handling injury claims were decreased by 43% (from 3.88 per 100 full time equivalents [FTEs] pre-intervention, to 2.23 per 100 FTEs post-intervention). Lost time frequency decreased from 1.91 FTEs pre-intervention, to 1.03 per 100 FTEs post-intervention.
Muto et al. 2008 [52] A non-randomized intervention study in care staff in two schools for children with disabilities. Intervention included use of nursing assistance tools (n=21) vs. control (n=20) groups. No significant difference in lower back pain prevalence was seen in either group. Upper arm pain decreased in the intervention group from 47.6% at baseline to 23.8% at end point (p=0.063); high lower back burden decreased in the intervention group from 57.1% at baseline to 33.3% at end point (p=0.063). No significant impact was seen on low back pain, and depression.
Garg & Kapellusch 2012 [18] Long-term effects of ergonomic programme implementation with patient handling device installation in six long-term care facilities and one chronic care hospital in the US, plus a control group. Injury data was collected for an average of 38.9 months pre-intervention and 51.2 months post-intervention. Patient handling injuries decreased by 59.8% post-intervention (p <0 .001), and lost workdays by 86.7% (p<0 .001). All devices were rated as less stressful on the low back (p <0.001), shoulders (p≤0.008), and wrists (p≤0.005). Post intervention, no problems in performing their tasks (including patient lifting) were experienced by pregnant or older participants, or those with back problems, and the lifts were rated as more comfortable and safe by the majority of patients.
Kuijer et al. 2014 [53] Evidence-based practice guideline developed by a team of occupational health and safety professionals, based on a systematic literature review.      • Training and advice may theoretically reduce low back load, but this is unlikely to be achieved and upheld in practice
     • Patient lifting devices can remove the need for manual lifting, but low back loading can still occur
     • Lifting belts can help to reduce low back loading
     • Lifting teams can reduce the number of patient lifts performed
     • Team lifting does not increase the risk for lower back pain