Share this article on:

Effect of a Simulation Educational Intervention on Knowledge, Attitude, and Patient Transfer Skills: From the Simulation Laboratory to the Clinical Setting

O'Donnell, John M. CRNA, MSN, DrPH; Goode, Joseph S. Jr CRNA, MSN; Henker, Richard CRNA, PhD; Kelsey, Sheryl PhD; Bircher, Nicholas G. MD; Peele, Pamela PhD; Bradle, Judith BA; Close, John MA, PMSD; Engberg, Richard BS; Sutton-Tyrrell, Kim RN, DrPh

doi: 10.1097/SIH.0b013e318212f1ef
Empirical Investigations

Introduction: Musculoskeletal injury in the workplace is the primary work-related factor in loss of nursing personnel from the workforce. Moving or transferring patients is the dominant contributing event. A simulation educational approach has not been closely studied in this area but may have advantages over traditional approaches. Specific aims were to (1) evaluate the effect of a simulation intervention on success of patient transfers in a clinical setting and (2) measure change in participants' knowledge and attitude as a result of the intervention.

Methods: A prospective, observational, longitudinal design was used. Baseline patient transfer observations were conducted on control and intervention units. An optimum task set was developed using hierarchical task analysis methods. Subjects (N = 71) completed pre- and postintervention knowledge and attitude assessments. The intervention consisted of simulated patient transfers using a mannequin, education, and training, followed by repeated simulated transfers using a mannequin with debriefing. Observations of patient transfers in patient care areas were repeated at 4 and 12 weeks.

Results: Patient transfer success improved from 66% at baseline to 88% at the 4-week measurement point (t = 7.447, P ≤ 0.0004). At 12 weeks, transfer success had decreased to 71%, with addition of new employees between weeks 4 and 12 confounding the 12-week measurement. Knowledge improved from a baseline of 65% to 95% postsimulation intervention (z = −6.634, P ≤ 0.0004). Attitude change was also evaluated with significance seen with 12 of 15 items (P ≤ 0.05).

Conclusions: A simulation intervention was successful in significantly improving knowledge and changing subject perceptions with regard to this task. Skills acquired through simulation successfully transferred to the clinical setting. Improvement in success for patient moves not trained in the simulation laboratory suggests that acquired skills were generalizable and supports application to different settings.

From the Department of Acute/Tertiary Care, University of Pittsburgh School of Nursing (J.M.O., R.E., J.S.G., R.H.); Department of Epidemiology, University of Pittsburgh Graduate School of Public Health (S.K., K.S.-T., P.P., N.G.B.); University of Pittsburgh School of Medicine (N.G.B.); University of Pittsburgh School of Dental Medicine (J.C.); The Winter Institute for Simulation, Education and Research (WISER) (J.M.O.); Department of Anesthesiology, University of Pittsburgh Medical Center (N.G.B., J.S.G., J.B.); and UPMC Health Plan (P.P.), Pittsburgh, PA.

Supported by the USAF, administered by the US Army Medical Research Acquisition Activity, Ft. Detrick, MD (Award # DAMD 17-03-2-0017).

The study was performed at the University of Pittsburgh WISER Institute and the University of Pittsburgh Medical Center Southside Hospital Institute for Research and Rehabilitation.

The authors declare no conflicts of interest.

Reprints: John M. O'Donnell, CRNA, MSN, DrPH, University of Pittsburgh School of Nursing, 336 Victoria Building, Pittsburgh, PA 15261 (e-mail:

Nurses and nurse aides represent the majority of healthcare providers with nurses comprising 55% of the total healthcare workforce.1 Back injury and other musculoskeletal injuries have been cited as the single largest contributor to the ongoing nursing workforce shortage.2 The US Department of Labor's annual reporting of nonfatal injuries within private industry consistently lists nurses and nurse aides among injury leaders in the workplace. In 2009, nurse aides were first in the US workforce and registered nurses were sixth in nonfatal occupationally related musculoskeletal injury.3 Incidence rates in 2009 (defined as injuries per 10,000 workers) among nurse aides and orderlies were 226.4/10,000, which is significantly greater than the rate of manual laborers (146.39/10,000).3 The issue of nurse and nurse aide injury remains a national priority because of both a national and international shortage of nurses, with projections indicating a shortfall of as many as a million nurses by 2020.4

Back belts and other lift-support devices have been evaluated but have not been demonstrated to consistently reduce injury.5–7 Lifting teams and comprehensive ergonomic assessment of work spaces combined with lift equipment purchase and “no-lift” or “minimal lift” policies have been advocated and form the core of a national campaign sponsored by the American Nurses Association.8–10 Despite these efforts, nurse and nurse aide injury rates remain high with manual patient transfer identified as the most important risk factor.2,11 Simulation represents a new approach to the problem of transfer-related injury. Simulation has been used to evaluate the spinal loading effect of manual patient transfer versus use of mechanical devices. Daynard et al12 found that spinal loading was reduced with mechanical devices but that lifts took a significantly longer period of time to perform. Simulations of workplace conditions (such as spinal loading models and situations requiring use of body mechanics) have been used in the industry to model mechanisms of musculoskeletal injury and identify prevention strategies.13

Patient simulation as an educational methodology has grown dramatically in popularity over the last decade. While the discipline of healthcare simulation can be defined in several ways, Tekian et al14 defined simulation as “a person, device or set of conditions which attempts to present evaluation problems authentically.” While the spectrum of applications within the healthcare setting is impressive, evaluation of the effect of simulation training on improved patient care processes is complicated by several obstacles. These include the infrequency of some clinical events, shortage of resources required for structured study in a clinical setting, and a lack of validated methodology and tools for collection of clinical data. Patient transfer is a viable simulation training target for several reasons. First, patient transfer is not a rare event, occurring hundreds of times per day in most hospitals. Second, patient transfer is a task that can be broken into component elements. Third, the high level of nursing and nurse aide injury related to this task makes it a priority target.3

In approaching this issue, O'Donnell and Goode15 identified the lack of a universally accepted approach to performing patient transfer while minimizing provider risk. Using the ergonomic method of hierarchical task analysis, these authors identified an optimum task set for patient transfer, created a patient transfer protocol (Table 1), wrote a corresponding simulation training program, and developed mobile data collection tools for observation of patient transfers in the clinical setting.15,16 The protocol was based on best evidence from the literature; reviewed by a panel of healthcare and ergonomic experts (physical therapists, occupational therapists, professional nurses, hospital administrators, ergonomic experts, and an occupational medicine physician); and refined through an iterative process of description, clinical observation, evaluation, and re-evaluation. Significant improvement in caregiver team transfer success was demonstrated in the simulation laboratory. Inter-rater reliability in use of the protocol was established.16

Table 1

Table 1

The purpose of this study was to determine the impact of a simulation training program emphasizing a patient transfer protocol on successful clinical adoption of the protocol. Specific aims were to (1) evaluate the effect of a simulation intervention on patient transfer success in a clinical setting and (2) measure change in participant's knowledge and attitude as a result of the intervention.

Back to Top | Article Outline


Setting and Sample

The setting was a University-affiliated rehabilitation institute including four clinical units divided between two hospitals. The three intervention units were physically connected at one hospital and employed 81 providers. The control unit was at a separate facility and employed 14 providers. Key demographic factors were compared between control and intervention unit personnel at baseline to assure equivalence. The intervention group was a convenience sample of nurses and nurse aides who were randomized to teams of three to four for the simulation intervention. Inclusion criteria for the intervention group included employment as a nurse or nurse aide on the intervention units and participation in direct patient care. Exclusion criteria included age younger than 18 years or inability to perform a patient transfer because of a physical limitation.

A total of 71 employees on the intervention units agreed to participate in the simulation intervention, representing 88% of the total employee group (71/81). All participants were either nurses (n = 48) or nurse aides (n = 23). Participants consented before receiving the simulation intervention and completing the knowledge and attitude tools. Control unit employees conducted patient transfers according to their normal practice and did not receive an intervention.

Back to Top | Article Outline

Simulation Intervention Design and Measurement

A prospective, observational, longitudinal design (Fig. 1) was used. After Institutional Review Board (IRB) approval, structured baseline observation of patient transfers was conducted on control and intervention units. The researchers were not permitted to interfere with patient care or control team composition, and as a result, some teams on the intervention units included or were entirely composed of untrained personnel. Furthermore, the IRB did not permit the researchers to identify participation as research subjects or track individual performances during observations. Because of these constraints, the unit of analysis was the team that is reasonable as patient transfers are typically a “team” activity. Transfers were observed at baseline and again at 4 and 12 weeks. At each observation point, transfer events were rated by trained observers supervised by ergonomic experts (occupational therapists certified as ergonomists). Each unique transfer event was rated with a score that represented the number of steps rated “completed”/the total number of steps rated using the patient transfer protocol. A score of “completed” indicated that all elements of the step were performed correctly according to a set of operational definitions.

Figure 1.

Figure 1.

The simulation intervention was conducted at the Peter M. Winter Institute for Simulation, Education and Research (WISER) of the University of Pittsburgh. After consent was obtained, baseline team transfer skill was determined by having teams perform two simulated transfers using the Tuff Kelly Transfer Mannequin (Laerdal Inc., Stavanger, Norway). Ergonomic experts scored all events using the SimMan software system. After the second transfer was performed, a structured video debriefing was conducted using the SimMan Debriefing Viewer Program. The SimMan log file display included evidence-based statements providing rationale for each transfer step as well as the rating of the expert. On-line patient transfer protocol materials were then reviewed with each team. The simulations, debriefings, and on-line material review constituted the intervention, which lasted for 1.5 hours. After this intervention, transfer skill was reassessed by having the same participant teams perform two different simulated patient transfers. A 10-item knowledge and 15-item attitude assessment were completed immediately pre- and postintervention by participants via wireless computer consoles with responses automatically uploaded into the WISER Simulation Information Management System (SIMS). These tools were specifically referenced to the optimal transfer steps described through hierarchical task analysis. The SIMS system allowed assessment data to be entered in a secure, password-protected manner. All subject data were deidentified by an honest broker before being released to the investigators for analysis.

Back to Top | Article Outline


The primary measurement tool for evaluating team performance of patient transfer was the “patient transfer protocol.” The construct validity of the patient transfer protocol tool can be evaluated in light of evidence of the five sources of validity defined by Downing,17,18 which include content evidence, response process, internal structure, relationship to other variables, and consequences.

Content evidence in support of the validity of the patient transfer protocol is established as follows. First, the transfer protocol defined the optimal task steps for patient transfer and was developed using hierarchical task analysis methods, which have been widely used in industry and are considered to be valid for analysis of individual to system level processes.19,20 Thus, use of hierarchical task analysis resulted in the development of a patient transfer protocol that accurately represented the clinical process. This task analysis development process was iterative and included observation, description, and testing to establish the optimum set of task steps (Table 1). Each step was cross-referenced to best evidence from the literature and accepted standards of care and then operationally defined by substeps derived from iterative review by the expert panel. Second, a 10-item multiple choice knowledge tool was given pre- and postsimulation intervention. Each item was associated with a transfer protocol step and referenced to best evidence or standards of care. All items were cross-referenced to content within the web-supported course material or to specific debriefing feedback programmed within the SimMan software.

Response process evidence in support of the validity of the transfer protocol included use of a standardized instructor manual to ensure that the simulation intervention and all assessments were conducted in the same manner for all participants. A test map was developed for multiple choice items with reference to each protocol step. All researcher activities in the simulation setting were followed according to a checklist, including reading of introduction and consent forms, obtaining consents, administering baseline knowledge and attitude assessments, randomizing participants to transfer teams, conducting two initial transfers, transfer debriefing and review of course material, conducting two final transfers, and administering postintervention knowledge and attitude assessments. Knowledge and attitude measure scores were reported via the WISER SIMS System. Analysis of the knowledge test items was conducted by a panel of healthcare experts. All trainee responses were deidentified and aggregate item analysis included measures of item difficulty.

Relationship to other variables was demonstrated in two areas. First, a 15-item instrument was administered, which examined participant attitudes or perceptions regarding change in their cognitive, affective, psychomotor, communication, and safety skills. These areas directly relate to domain areas needed to perform a patient transfer. Items were adapted with permission from attitudinal items used by the Office for Measurement and Evaluation at the University of Pittsburgh or WISER. Second, the percent of intervention unit participants trained in the protocol was compared with protocol success at each observation point.

Consequences of the intervention in support of validity were evaluated as follows. Participants (71/81 or 88% of intervention unit personnel) entered the intervention arm voluntarily. Table 3 demonstrates that participants had a high level of anticipation that the intervention would be valuable across multiple domain areas including improved understanding of how to prevent provider injuries. A second important consequence of the protocol was the recognition of value by the University of Pittsburgh Medical Center (UPMC) Center for Quality Improvement and Innovation and the UPMC Health Plan, which adopted the protocol for use in mandatory employee training across two community hospitals with subsequent plans to support implementation across an entire health system.

Table 3

Table 3

The final evidence in support of the validity of the tool included evaluation of inter-rater reliability during use of the tool. Ergonomic expert scores were compared with trained raters during simulated transfer events. Step scoring options included “completed” indicating successful and correct completion of all substeps or “not completed” indicating failure on any single substep. The inter-rater reliability scores from the simulated moves are characterized as substantial agreement to near perfect (κ = 0.43–0.83). The terms moderate, substantial, and near-perfect agreement in interpretation of the kappa statistics are based on the classic interpretation by Landis and Koch.21

Back to Top | Article Outline

Statistical Methods

SPSS 15.0 was used for all data analyses. Level of significance for all statistical tests was established a priori as 0.05. Descriptive statistics were used to summarize and compare baseline demographic variables of employees on both intervention and control units. Mean ± standard deviation and median values were calculated for age. The Mann-Whitney U test was used to evaluate for differences in age between the control and intervention groups. Differences between gender, job class, and years of clinical experience were evaluated using χ2.

Observations of patient transfer were conducted at three measurement points: baseline, 4 weeks, and 12 weeks. The unit of analysis on the patient care units was the transfer team with success measured by adherence to the patient transfer protocol as previously described. A distribution of patient transfer ratings was developed for each measurement point. The data collection site and the location (intervention control unit), the type of move performed, and a score for each step of the transfer were recorded.

A 2 × 2 × 3 univariate analysis of variance (ANOVA) was conducted to look at the differences in team performance over time (between subject effects). Three main effect variables were entered into the model. These included Group (control vs. intervention units), Move type (chair moves vs. bed moves), and Time (patient transfer performance baseline, 4 weeks postintervention, and 12 weeks postintervention). As each team was uniquely constituted for each event, within-subject measurements were not conducted in accordance with IRB restrictions.

Pre- and postintervention knowledge levels were measured using the same 10-item multiple-choice examination. Each item corresponded to one of the steps of the patient transfer protocol as well as best practices according to the expert panel and literature. Mean ± standard deviation scores were calculated for both the pre- and postintervention scores. The unit of analysis for this measure was the individual. Improvement in performance was evaluated using a Wilcoxon signed-rank test. The Kolmogorov-Smirnov test for normality was conducted and the data were not normally distributed, which was apparent by inspection of the variables histogram. Attitude items were rated with a 5-point Likert scale (strongly disagree = 1, strongly agree = 5). Multivariate ANOVA was used in analysis of the Likert scale attitude data.

Back to Top | Article Outline



Deidentified demographic data were obtained for respondents on both intervention (n = 74) and control (n = 14) units. Mean age of personnel in the intervention unit was 44.6 ± 11.9 years and on the control unit 44.4 ± 13.8 years (z = −0.097, P = 0.953). Gender, years of experience, and job title classifications were compared using the χ2 statistic. Control and intervention unit personnel were not significantly different with respect to these variables (P > 0.05) (Table 2).

Table 2

Table 2

Back to Top | Article Outline

Patient Transfer Skill Improvement in the Clinical Setting

The study was conducted from May 2006 to August 2006. The total number of transfer observations in the study was 306. At baseline, a total of 103 transfers were observed on the intervention units and 34 events on the control unit. Four-week postintervention, a total of 53 transfers were observed on the intervention units and 30 transfers on the control unit. At 12 weeks, a total of 54 transfers were observed on the intervention units and 32 transfers on the control unit. HP iPAQ handheld personal computers were programmed with the patient transfer protocol checklist as a simple graphical user interface (a program designed to allow a computer user to interact easily with the computer by making choices from menus or groups of icons) and used for data collection.

All three main effect variables (Group, Move type, and Time) in the 2 × 2 × 3 ANOVA model demonstrated significance (P ≤ 0.0004). Interaction effects were then analyzed with the Group × Time interaction significant (P ≤ 0.0004).

Protected t tests (equivalent to the Fisher least significant difference test) were used in post hoc pairwise comparisons. This analysis revealed that the change in transfer success (66%–88%) at the 4-week postintervention measurement point was significant (t = 7.447, P ≤ 0.0004) (Fig. 2). No other pairwise comparison of Group × Time was statistically significant, indicating that patient transfer success at the 12-week postintervention point had decreased toward baseline.

Figure 2.

Figure 2.

A variety of patient transfers were observed on the clinical units. They were stratified into bed-based moves versus chair-based moves (Move type). All transfer events during the simulation intervention were conducted as “bed” events due to mobility limitations of the Laerdal Tuff Kelly Transfer Mannequin. Improvements in chair-based move success in the clinical setting at 4-week postintervention were equal to improvements in bed-based move success, indicating that the protocol generalized to the nontrained move. While overall patient transfer success at 12 weeks regressed toward baseline (88%–71%), chair-based move success regressed from 86% to 54%, whereas the bed-based patient move success regression was smaller (89%–77%) (Fig. 3).

Figure 3.

Figure 3.

Success on each protocol step on the intervention units was analyzed for each measurement point. The protocol steps with the highest baseline success rates were Gather Equipment and Resetting the Environment. Steps with the lowest baseline success were Perform the Move followed by Assess the Patient, Reassess the Patient, and Communicate to Personnel (Fig. 4). At 4-week postintervention, every protocol step demonstrated improvement, which paralleled the simulation laboratory findings. All individual protocol steps were above the 80% success level with the exception of Communicate to the Patient and Perform the Move. At 12 weeks, the overall success percentage had regressed to 71%.

Figure 4.

Figure 4.

Back to Top | Article Outline

Knowledge Improvement

The 10-item multiple-choice knowledge instrument was administered pre- and postintervention to all subjects in the intervention arm of the study. Knowledge improvement was significant with mean knowledge scores improving from 65% (±18%) to 90% (±12%) (z = −6.634, P ≤ 0.0004).

Back to Top | Article Outline

Attitude Change

Pre- and postintervention attitude evaluation was conducted. Subjects responded on a 5-point Likert scale to 15 items. Mean scores for both pre- and postattitude measures were calculated. Multivariate ANOVA was used to evaluate within-subject effects; Wilk's lambda was significant (F = 2.94, P = 0.003). Post hoc univariate analysis demonstrated statistically significant change in 12 of 15 items (P ≤ 0.05) (Table 3).

Back to Top | Article Outline


In this report, we demonstrate that a simulation intervention can be used to improve success for a common clinical event: the patient transfer. Improvement in patient transfer success was measured using an optimal task set for patient transfer. Participants on the intervention units demonstrated significant improvement in transfer skill at the 4-week measurement point while there was no change noted in the control unit. This outcome demonstrated that the protocol was effective in changing clinical behavior among teams of nurses and nurse aides who received the training. This clinical improvement paralleled the simulation laboratory improvement in patient transfer success discussed earlier and first reported by O'Donnell et al.16

In addition to improved patient transfer skill, participants in this study experienced improvement in knowledge and change in attitude. The knowledge improvement was 25 percentage points and participants indicated that items were relevant and that the assessment should be used again in future courses. Participants demonstrated change in multiple attitudinal areas and felt that they had improved in identifying personnel needed for moves; knowledge of injury prevention; skill in injury prevention; overall skill in patient transfer; improved communication; and safety in patient transfer. In addition, participants reported that training program objectives were met; scenarios were realistic; scenarios were similar to clinical transfer events; injury prevention skills had improved; overall patient transfer skills had improved; confidence had increased; and effectiveness as team members had improved.

Three items that did not demonstrate statistically significant change in rating were “improved knowledge of equipment needed for transfer,” “anxiety related to being observed,” and “realism of simulation scenarios” during the transfer events. Equipment selection is not a focus of the curriculum although the on-line material does review this topic. Anxiety scores pre- to postintervention decreased. This decrease was not statistically significant, but it is important to note that there was a reduction in anxiety reflecting efforts to provide a nonthreatening and comfortable environment for subjects within the context of the research protocol. This effect was noticeable as participants became more comfortable with the simulation environment. Participants anticipated that the simulation scenarios would be realistic before the event (mean score 4.17/5), with no significant change in this perception after the training intervention.

Finally, postintervention improvement in success on untrained (chair) moves was similar to improvement on the trained (bed) moves at 4 weeks (Fig. 3). This finding indicates that participants applied the patient transfer protocol to moves they did not practice during the intervention. The improvement in untrained moves suggests that the protocol was generalized to the chair moves by participants. This is relevant in use of the protocol in training at institutions that use different equipment or perform other types of patient transfers. Furthermore, this has broader implications in developing future training and evaluation protocols for other healthcare tasks. This is an important finding as a number of other reports have described a broadened scope and diversity in the application of simulation methods, including alteration of attitude, skill attainment, diagnosis of system problems, enhanced knowledge, and evaluation of competence.22–38

Back to Top | Article Outline

Regression of Transfer Success

Regression of patient transfer success at 12 weeks occurred on the intervention units, with no significant change in success noted on the control unit. Two possible explanations were considered for this effect: (1) dilution of group skill by the addition of untrained personnel or (2) a fading of the intervention effect.

Back to Top | Article Outline

Dilution of Group Skill

To determine whether there was a dilution effect, we analyzed personnel turnover on both control and intervention units. The control unit experienced the following changes in their staff. Five providers left the unit and 6 were hired, bringing the total staffing to 15. None of the new employees had worked on the intervention units or received training in the protocol. The overall success rate on the control units did not substantially change between the three observation points.

Success on the protocol relative to staff turnover on the intervention units was then analyzed. Use of team and not individual performance data was a requirement of the IRB as a method to prevent workplace performance repercussions for individual team members. Thus, we were unable to track individual participants or look at the mix of trained versus untrained participants on teams. This was true at baseline as well due to the nonparticipation of 10 employees. Because of this limitation, team composition was dynamic and could have ranged from entirely trained to entirely untrained at any point in the study. At the 4-week measurement point, 89% of the staff on the intervention unit had received the training and transfer success was 88%. Between the 4-week and 12-week measurement points, a total of three employees left the unit and 13 employees were hired. The three employees who left had not received the training. The addition of the 13 new personnel therefore represented the addition (net) of 10 untrained providers (12.3%) and a new personnel total of 91 (78 [81 − 3] + 13). None of the new staff members were trained before the 12-week measurement point. The 17% regression in transfer success between 4 and 12 weeks must therefore be considered in light of a 12.3% dilution by untrained personnel. The reduction of transfer success could be predicted by taking this dilution into account. New employees would be expected to perform transfers at the baseline success rate of 65% (average of control and intervention units at baseline). These individuals would be expected to perform a proportionate number of the 53 moves we observed (12.3% of 53 = 6.5 moves). Trained employees would be estimated to have a success rate as high as 88% (best case scenario) and perform a proportionate number of the 53 moves (77.7% of 53 = 46.5 moves). The overall calculation would thus be ([0.123 × 0.65 × 53] + [0.777 × 0.88 × 53])/53 = 76.5% success rate. This rate is 5.5% higher than the 71% rate observed in the study.

Back to Top | Article Outline

Fading of the Intervention Effect

Fading of the intervention effect is an expected outcome in educational intervention studies. One way to address this issue is to deploy the intervention on a broader scale and control for dilution to evaluate for fade more directly. As noted in the validity discussion, the UPMC Health Plan evaluated and decided to adopt the protocol for use in transfer training as part of an overall hospital-based musculoskeletal injury prevention program (We've Got Your Back). All nurses and nurse aides in one community hospital (n = 293), including new hires, received the training. Results at baseline were lower (56% success) than either control or intervention units (mean 65%). No immediate follow-up was conducted in this hospital-wide training program. However, a 16-week observation of transfer success demonstrated an 83% success rate with an overall improvement in success from baseline of 27% compared with the 22% improvement from baseline on the intervention units at the 4-week observation point on the intervention units.

In summary, we believe that the regression on the intervention units noted at 12 weeks is most likely a combination of both dilution and a fading of the intervention. Our assumption in the above calculations for dilution is that trained participants would continue to perform at the level observed at 4 weeks (88%). This is unlikely as skill decay is expected after an educational research intervention.39–42 Even if performance did continue at this level, the observed success of 71% was lower, which suggests that fading may also have occurred.

Back to Top | Article Outline

Limitations of the Study

Inability to Track Individual Team Members

As discussed, the authors were constrained by the IRB from identifying individual providers during the clinical observations because of the potential risk of workplace repercussions. Thus, we were unable to track individual participants or look at the mix of trained versus untrained participants on teams with the consequence of being unable to determine how team composition (trained vs. untrained) impacted patient transfer success rates.

Back to Top | Article Outline

Technology Challenges

A second limitation to the study was the level of technology expertise required to participate. While 71 nurses or nurse aides participated in the simulation intervention, only 67 participants were able to complete both the pre- and post-test knowledge tool and the pre- and post-test attitude tools. Participants' lack of familiarity with use of technology in general and the wireless laptop, in particular, were identified as the probable causes. These issues occurred despite scripted orientation to the equipment and technology.

Back to Top | Article Outline


Over the past three decades, a variety of patient transfer and injury prevention programs have been published in the literature. However, none have combined an ergonomically derived patient transfer protocol with current simulation educational approaches. The purpose of this study was to determine the impact of a simulation training program emphasizing a set of optimum transfer task steps on patient transfer success in the clinical setting. The 4-week postintervention observation point demonstrated significant improvement in transfer success; however, the regression toward baseline at 12 weeks was confounded by the addition of new staff on the intervention units. In a follow-up implementation in a community hospital, we offered the training to all providers and then conducted a 16-week (4-month) observation. Our hope was to minimize the impact of dilution on transfer success by having all providers (including new employees) receive the training. The outcome was that observed transfer success at 16 weeks was 83% (an improvement of 27% from baseline in this facility). Future research with more frequent observation points is needed to determine the rate of skill decay and establish an optimum time interval between training cycles.

The development of the handheld computer data collection system was crucial to the success of this project. The graphical user interface is reprogrammable and will be adapted for future simulation projects requiring acquisition of clinical data. Other important characteristics of the system included simplicity of data entry, streamlined data uploading, ability to change answers during rating, and the use of removable storage chips, which support portability and data protection. Finally, the system is unobtrusive, allowing the observer to enter data without alarming providers, families, or patients in the clinical setting.

Improvement in provider success in transfer types (chair-based transfers) that were not part of the intervention indicates that the patient transfer protocol is generalizable to events other than those that were specifically trained during the intervention. This finding is particularly intriguing as it holds potential for deploying this transfer protocol across a variety of settings and transfer types. In addition, these results suggest that, with modification, the approach could be adapted to transfer situations outside the healthcare setting (ie, industry or the military) or for improving performance of other healthcare processes.

The public health implication of the ongoing loss of qualified nurses from our healthcare system due to injury is evident. In this report, we have combined a proven ergonomic approach with a simulation intervention. Use of simulation methods grounded in established ergonomic methodology provides a new approach to developing training and assessment protocols for nursing injury prevention as well as for other healthcare processes.

Ideally, follow-up investigation will focus on training an entire healthcare system with concurrent tracking of protocol adherence, provider injury rates, and costs. It is important that an economic analysis be conducted to characterize return on investment for the program. Additional follow-up goals include several long-term measurements, including transfer skill retention, reduction of provider injury rates, and decreased attrition of providers due to injury.

Back to Top | Article Outline


1. United States Department of Labor. NAICS 61 & 62: Education and Health Services. Bureau of Labor Statistics; 2006. Available at: Accessed October 7, 2010.
2. Edlich RF, Hudson MA, Buschbacher RM, et al. Devastating injuries in healthcare workers: description of the crisis and legislative solution to the epidemic of back injury from patient lifting. J Long Term Eff Med Implants 2005;15:225–241.
3. United States Department of Labor. United States Department of Labor. Nonfatal Occupational Injuries and Illnesses Requiring Days Away From Work, 2009. Washington, DC: USDL; 2010. Available at: Accessed December 4, 2010.
4. Dall T. What is Behind HRSA's Projected Supply, Demand and Shortage of Registered Nurses? Washington, DC: HRSA; 2006:1–34.
5. Gatty CM, Turner M, Buitendorp DJ, Batman H. The effectiveness of back pain and injury prevention programs in the workplace. Work 2003;20:257–266.
6. Minor SD. Use of back belts in occupational settings. Phys Ther 1996;76:403–408.
7. Mitchell LV, Lawler FH, Bowen D, Mote W, Asundi P, Purswell J. Effectiveness and cost-effectiveness of employer-issued back belts in areas of high risk for back injury. J Occup Med 1994;36:90–94.
8. Blakeney B. ANA launches back injury prevention campaign handle with care. Imprint 2003; 50:46–47.
9. de Castro AB. Handle with care: the American Nurses Association's Campaign to address work-related musculoskeletal disorders. Online J Issues Nurs 2004;9:3.
10. de Castro AB, Hagan P, Nelson A. Prioritizing safe patient handling: the American Nurses Association's Handle with Care Campaign. J Nurs Adm 2006;36:363–369.
11. Edlich RF, Winters KL, Hudson MA, Britt LD, Long WB. Prevention of disabling back injuries in nurses by the use of mechanical patient lift systems. J Long Term Eff Med Implants 2004;14:521–533.
12. Daynard D, Yassi A, Cooper JE, Tate R, Norman R, Wells R. Biomechanical analysis of peak and cumulative spinal loads during simulated patient-handling activities: a substudy of a randomized controlled trial to prevent lift and transfer injury of health care workers. Appl Ergon 2001;32:199–214.
13. American Conference of Governmental Industrial Hygienists. Ergonomic Interventions to Prevent Musculoskeletal Injuries in Industry. Chelsea, MI: Lewis Publishers; 1987.
14. Tekian A, McGuire CH, McGaghie WC. Innovative Simulations for Assessing Professional Competence: From Paper and Pencil to Virtual Reality. University of Illinois: Chicago, IL; 1999.
15. O'Donnell JM, Goode JS. Simulation in nursing education and practice. In: Riley R, ed. Manual of Simulation in Healthcare. 1st ed. Oxford, NY: Oxford University Press; 2008:241–276.
16. O'Donnell J, Bradle J, Goode J. Development of a simulation and internet based pilot intervention to evaluate adherence to a patient transfer protocol in the real world environment. Simulat Healthc 2007;1:62.
17. Downing SM. Validity: on meaningful interpretation of assessment data. Med Educ 2003;37:830–837.
18. Downing SM. Reliability: on the reproducibility of assessment data. Med Educ 2004;38:1006–1012.
19. Annett J, Cunningham D, Mathias-Jones P. A method for measuring team skills. Ergonomics 2000;43:1076–1094.
20. Stanton NA. Hierarchical task analysis: developments, applications, and extensions. Appl Ergon 2006;37:55–79.
21. Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics 1977;33:159–174.
22. O'Donnell J, Fletcher J, Dixon B, Palmer L. Planning and implementing an anesthesia crisis resource management course for student nurse anesthetists. CRNA 1998;9:50–58.
23. Liu D, Grundgeiger T, Sanderson PM, Jenkins SA, Leane TA. Interruptions and blood transfusion checks: lessons from the simulated operating room. Anesth Analg 2009;108:219–222.
24. DeCarlo D, Collingridge DS, Grant C, Ventre KM. Factors influencing nurses' attitudes toward simulation-based education. Simul Healthc 2008;3:90–96.
25. Spunt D, Foster D, Adams K. Mock code: a clinical simulation module. Nurse Educ 2004;29:192–194.
26. Ackerman A. Acquisition and Retention of CPR Knowledge and Skills for Junior Level Baccalaureate Nursing Students. Pittsburgh, PA: School of Nursing, Duquesne University; 2007.
27. Tsang JS, Naughton PA, Leong S, Hill AD, Kelly CJ, Leahy AL. Virtual reality simulation in endovascular surgical training. Surgeon 2008;6:214–220.
28. Draycott TJ, Crofts JF, Ash JP, et al. Improving neonatal outcome through practical shoulder dystocia training. Obstet Gynecol 2008;112:14–20.
29. Crofts JF, Bartlett C, Ellis D, Hunt LP, Fox R, Draycott TJ. Training for shoulder dystocia: a trial of simulation using low-fidelity and high-fidelity mannequins. Obstet Gynecol 2006;108:1477–1485.
30. Marsch SC, Tschan F, Semmer N, Spychiger M, Breuer M, Hunziker PR. Performance of first responders in simulated cardiac arrests. Crit Care Med 2005;33:963–967.
31. Hunt EA, Walker AR, Shaffner DH, Miller MR, Pronovost PJ. Simulation of in-hospital pediatric medical emergencies and cardiopulmonary arrests: highlighting the importance of the first 5 minutes. Pediatrics 2008;121:e34–e43.
32. DeVita MA, Schaefer J, Lutz J, Wang H, Dongilli T. Improving medical emergency team (MET) performance using a novel curriculum and a computerized human patient simulator. Qual Saf Health Care 2005;14:326–331.
33. DeVita MA, Braithwaite RS, Mahidhara R, et al. Use of medical emergency team responses to reduce hospital cardiopulmonary arrests. Qual Saf Health Care 2004;13:251–254.
34. Thompson S, Neal S, Clark V. Clinical risk management in obstetrics: eclampsia drills. Qual Saf Health Care 2004;13:127–129.
35. Hoffmann RL, O'Donnell JM, Kim Y. The effects of human patient simulators on basic knowledge in critical care nursing with undergraduate senior baccalaureate nursing students. Simul Healthc 2007;2:110–114.
36. Girzadas DV Jr, Clay L, Caris J, Rzechula K, Harwood R. High fidelity simulation can discriminate between novice and experienced residents when assessing competency in patient care. Med Teach 2007;29:472–476.
37. Tuttle RP, Cohen MH, Augustine AJ, et al. Utilizing simulation technology for competency skills assessment and a comparison of traditional methods of training to simulation-based training. Respir Care 2007;52:263–270.
38. O'Donnell JM, Beach M, Hoffman L. Human simulation training in the ICU: applicability, value, and disadvantages. Crit Care Alert 2005;13:41–48.
39. Barsuk JH, Cohen ER, McGaghie WC, Wayne DB. Long-term retention of central venous catheter insertion skills after simulation-based mastery learning. Acad Med 2010;85(10 suppl):S9–S12.
40. Niles D, Sutton RM, Donoghue A, et al. “Rolling Refreshers”: a novel approach to maintain CPR psychomotor skill competence. Resuscitation 2009;80:909–912.
41. Lammers RL. Learning and retention rates after training in posterior epistaxis management. Acad Emerg Med 2008;15:1181–1189.
42. Einspruch EL, Lynch B, Aufderheide TP, Nichol G, Becker L. Retention of CPR skills learned in a traditional AHA Heartsaver course versus 30-min video self-training: a controlled randomized study. Resuscitation 2007;74:476–486.

Simulation; Hierarchical task analysis; Back injury; Ergonomics

© 2011 Society for Simulation in Healthcare