Advertisement

Occupational Health Science

, Volume 2, Issue 4, pp 437–450 | Cite as

Identifying Safety Peer Leaders with Social Network Analysis

  • David A. Hurtado
  • Lisset M. Dumet
  • Samuel A. Greenspan
  • Yaritza I. Rodríguez
  • Gregory A. Heinonen
Brief Research Report

Introduction

Injuries rates in healthcare are higher than in other industries (50.2 vs. 38.8 per 10,000 workers) (Dressner 2017). Safe patient handling programs reduce the risk of patient-assist injuries, an issue that accounts for 45% of workers’ compensation claims. Safe patient handling programs implement engineering (e.g., introduction of patient-transfer equipment), administrative (e.g., no-manual-lift policy implementation), and educational (e.g., peer coaching) control measures (Nelson and Baptiste 2006). A study across 141 Veterans Administration facilities over three years found that 23% of the variation in injury reduction was attributed to actions performed by peer leaders (Powell-Cope et al. 2014). Program effectiveness ranges from non-significant number of lost workdays in a five-year period (Nelson et al. 2006), negative benefit-to-cost ratios over seven years (Tompa et al. 2016), to decrements in injury rates of 50–65% in one to seven years (Tompa et al. 2016; Zadvinskis and Salsbury 2010). Considering the wide range of program effectiveness, identifying optimal peer leaders might be a significant strategy to boost the impact of safe patient handling programs (Matz 2005). However, evidence-based methods to detect peer leaders are still lacking (Ploeg et al. 2010).

Why Peer Leaders Matter for Safe Patient Handling

Although management and supervisors sanction engineering controls (e.g., purchase of patient-transfer equipment) and administrative controls (e.g., development of patient-transfer policies or protocols), peer leaders are important to catalyze the integration of these components (Nelson and Baptiste 2006). Logistically, safe patient handling often requires two or more workers to operate patient assist or transfer equipment adequately (Baptiste et al. 2006). An eight-year study documenting successes and barriers of the development and maintenance of a peer led safe patient handling program found that co-worker and supervisor support were integral to accommodating the uniqueness of each patient’s mobility needs (Olinski and Norton 2017). By providing support to others, peer leaders contribute to overcoming barriers for nurses who wish to engage in safe patient handling (Krill et al. 2012). Additionally, peer leaders need to work closely with supervisors as workers use patient-transfer equipment more frequently if supervisor support is high (Caspi et al. 2013). Peer leaders contribute to injury prevention via two pathways: 1) provision of co-worker support including information, role-modeling with feedback (e.g., how to use equipment, how to report incidents) (Alamgir et al. 2011; Lawton and Parker 2002; Spruce 2015) and 2) proactive liaison with stakeholders to spearhead engineering or administrative control measures (Allen et al. 2015; Krill et al. 2012).

Peer leaders can act as “boots on the ground” only if their co-workers recognize in them the attributes that are needed to effect change. Although leadership theories (e.g., transformational leadership) are more suitable for supervisors who have the formal power and authority to produce organizational change, theories that explain the informal influence of peer leaders are rarer. Nevertheless, the role of peer leaders in safe patient handling programs is similar to the notion of opinion leaders featured in the Diffusion of Innovations theory (Rogers 2010). Opinions leaders are outstanding individuals whose social influence derives from their perceived expertise and credibility (Greenhalgh et al. 2004). Opinion leaders spark organizational change by leveraging their relationships with their peers (Dearing 2009), contrary to supervisors with formally ascribed power and authority.

Approaches to Identify Peer Leaders for Safe Patient Handling

Safe patient handling program guidelines define peer leaders as those highly motivated individuals who are committed to workplace safety, with appropriate technical and communication skills (Matz 2005). Detecting individuals with these characteristics is critical because systematic reviews have indicated that motivation and unit safety norms are salient determinants of equipment use (Rickett et al. 2006). However, program guidelines offer no indication of a technique to discover which workers are best positioned to champion safety at their units. Although many safe patient handling studies have not described how peer leaders were selected, common approaches include self-identification (Alamgir et al. 2011) or supervisor-recommendation (Kutash et al. 2009; Springer et al. 2009). However, these two approaches may conflate social influence with self-confidence or a more favorable relationship with supervisors (Iyengar et al. 2011).

Social Network Analysis (SNA) is an alternative method to identify peer leaders. SNA is a technique that examines advice-seeking interactions among peers to detect influential people for specific matters (Valente and Davis 1999). SNA has seldom been used to guide occupational health interventions (Zohar and Tenne-Gazit 2008), but public health interventions implemented by SNA-identified peer leaders to accelerate the diffusion of health and safety behaviors have reduced the intention to smoke (Valente et al. 2003), substance consumption (Valente et al. 2007), sexually transmitted diseases (Schneider et al. 2015) or have incremented sunblock uptake (Tsai et al. 2016). Peer leaders would be those group members that facilitate or block the flow of resources such as information or advice through direct interactions with other members or indirectly through multiples degrees of separation (Cho et al. 2012). Peer leaders have the potential to reach the most group members with implications for the spread of health and safety practices (Borgatti 2006).

Study Aims

This study applied SNA as a proof of concept to identify peer leaders, defined as those individuals who emerged from advice-seeking interactions about safe patient handling. Based on the construct of network centrality, SNA-identified peer leaders are individuals deemed by their co-workers as sources of safe patient handling advice as well as individuals who seek advice of others and connect network members. Possessing these network characteristics are imperative given that patient handling often requires two or more workers to operate patient assist or transfer equipment adequately (Baptiste et al. 2006). If SNA is a valid approach, then detected peer leaders should exhibit the attributes that program guidelines have delineated for peer leaders such as technical proficiency, commitment to workplace safety, communication, and peer support. This leads to the first hypothesis:
  • Hypothesis 1: SNA-identified peer leaders will obtain higher scores in safety attributes (e.g., equipment utilization, co-worker and supervisor support for safety) and interpersonal attributes (e.g., relations with supervisor, teamwork) than non-leaders.

Furthermore, this study compared the safety and interpersonal attributes of SNA identified peer leaders vs. supervisor identified peer leaders to establish if SNA has merits over the alternative approach.
  • Hypothesis 2: Workers identified by SNA as peer leaders will obtain higher scores in safety variables (e.g., equipment utilization, co-worker and supervisor support for safety) and interpersonal attributes (e.g., relations with supervisor, teamwork) than supervisor-identified peer leaders.

Given that there could be overlap between SNA and supervisor-identification of peer leaders, this study also tested whether workers identified as peer leaders by both methods would have better scores in safety and interpersonal attributes than workers deemed as peer leader by each single method. This examination would reveal the potential benefits of combining methods to identify peer leaders.
  • Hypothesis 3: Overlapping workers identified by SNA and supervisors as peer leaders will obtain higher scores in safety variables (e.g., equipment utilization, co-worker and supervisor support for safety) and interpersonal attributes (e.g., relations with supervisor, teamwork) than peer leader identified by either SNA or supervisors.

Methods

Participants

This cross-sectional study was conducted at a critical access hospital in Oregon in 2016. Eligible participants were Registered Nurses (RNs) and Certified Nursing Assistants (CNAs) from the Medical and Intensive Care Units (n = 38, response rate of 77.5%). Research assistants administered a survey on-site during paid time at the beginning or end of the shift. Five supervisors (charge nurses) also completed a short survey. Non-clinical staff were excluded. The Institutional Review Board at Oregon Health and Science University approved this study.

Measures

Identification of Peer Leaders with Social Network Analysis

In order to identify peer leaders, the survey included the following statement: “please choose from an employee roster all of your unit co-workers you would consult or ask advice about safe patient handling”. Peer nominations were examined through several measures of network centrality, a construct about the prominence of individuals in a network, indicating potential to diffuse resources such as information or advice in a group (Borgatti et al. 2013). A worker was identified as a peer leader if he or she scored at or above the 80th percentile of at least four of the following seven measures (Valente and Pumpuang 2007).
  1. (1)

    In-degree: the proportion of nominations that each worker received. This variable is a direct proxy of influence or popularity and it is often used in peer-led interventions (Tsai et al. 2016; Valente et al. 2003).

     
  2. (2)

    Out-degree: the proportion of peers that each worker nominated. Out-degree indicates which peers serve as sources of advice.

     
  3. (3)

    In-beta power: the proportion of nominations received weighted by nominators’ in-degree. A worker could exert indirect influence if he or she was nominated by a worker who already had multiple peer nominations.

     
  4. (4)

    Out-beta power: the proportion of nominations sent weighted by nominators’ out-degree. A worker could exert indirect influence if he or she nominates a worker who has multiple sources of peer advice.

     
  5. (5)

    Two-step in-reach: the proportion of peers a worker through two serial direct nominations. This measure indicates accessibility of any peer in the network.

     
  6. (6)

    Two-step out-reach: proportion of peers reached by a worker through two serial direct nominations. This measure signals a peer’s potential to diffuse advice.

     
  7. (7)

    Betweenness: proportion of times a worker lies in the shortest path connecting two peers. This measure indicates a worker ability to broker or interconnect advice.

     

Supervisor Identification of Peer Leaders

Supervisors (n = 5) were asked to select from the employee roster all workers they would consult or ask advice about safe patient handling. Under this approach, any worker selected was considered a peer leader.

Safety Attributes

Based on safe patient handling program guidelines for peer leaders, the following safety attributes were selected to determine if detected peer leaders would exhibit appropriate technical skills, if they were committed to promoting and advancing workplace safety (Matz 2005), and if they influenced social factors which can impact equipment use such as peer support (Alexopoulos et al. 2011) and safety climate (Lee and Lee 2017).
  1. (1)

    Equipment utilization (10 items): this was the primary validation measure because it pertains to the aspect for which advice was requested and because it demonstrates appropriate technical skills. The survey included a list of patient-transfer equipment (e.g., lifts, lateral transfer aids) available at the hospital, and participants self-reported the typical daily frequency of using each piece of equipment. Responses were coded with a scale from never (0 points) to ten times or more per day (7 points). An index was created adding individual scores for six different pieces of equipment.

     
  2. (2)

    Safety participation (3 items): this variable was measured with a validated scale about participants’ commitment and involvement with workplace safety (e.g., “I put in extra effort to improve the safety of the workplace”) (Neal and Griffin 2006).

     
  3. (3)

    Safety compliance (3 items): this variable pertained to adherence with safety procedures (e.g., “I use all the necessary patient handling equipment to do my job”). Both Safety participation and safety compliance scales were computed with the average of three questions that were coded from completely disagree (1 point) to completely agree (5 points). Both scales had observed Cronbach’s α > .88.

     
  4. (4)

    Supervisor support for safety (6 items): this variable pertained to the shared perceptions of supervisor’s role in improving safety climate. Only items about direct supervisors’ support for safety were included (e.g. “my direct supervisor refuses to ignore patient handling rules when work falls behind schedule”) (Zohar and Luria 2005). This variable was measured with six questions, coded with a five-point Likert scale from 1 (completely disagree) to 5 (completely agree) with an observed Cronbach’s α = .91.

     
  5. (5)

    Peer support for safety (2 items): this variable measured by two questions from the Nordic Psychosocial Questionnaire (e.g., “can you get support and help from other nurses?”), coded from very seldom (1 point) to very often or always (5 points) (Ørhede et al. 2000). This scale had appropriate internal reliability with a Cronbach’s α = .88.

     

Interpersonal Attributes

Also informed by safe patient handling program guidelines for peer leaders, the following interpersonal attributes were selected to determine if detected peer leaders would exhibit appropriate interpersonal and support skills (Matz 2005).
  1. (1)

    Leader-Member Exchange (LMX) (7 items): was included to assess peer leaders’ relations with their respective supervisors. LMX pertains to workers’ interactions with their supervisors, measured with six questions (e.g., “I would characterize the relationship I have with my manager as extremely effective”) (Bauer and Green 1996), coded from completely disagree (1 point) to completely agree (5 points). This scale had appropriate internal reliability with a Cronbach’s α = .92.

     
  2. (2)

    Teamwork (11 items): Questions from a nursing-specific team-work questionnaire included several facets of teamwork including perceived effect of teamwork (5 items) (e.g., “the team approach to patient care helps team members make better decisions”), communication (3 items) (e.g., “good communication among team members after a procedure is important for patient safety”) and level of teamwork (3 items (e.g., “my unit encourages teamwork and cooperation among its members”) (Kaissi et al. 2003). These subscales were coded with a four-point Likert scale from 1 (strongly disagree) to 5 (strongly agree) with Cronbach’s α between .76 and .85.

     

Socio-Demographic and Occupational Attributes

Self-reported characteristics were recorded, including age (in years), sex (male or female), race/ethnicity (Non-Hispanic white or other), Body Mass Index (BMI), job title (Registered Nurse or Certified Nursing Assistant), tenure (in years), and regular shift (day vs. night).

Statistical Analyses

Univariate and bivariate statistics were computed with UCINET (Borgatti et al. 2002). Nominations were compiled in an adjacency matrix (n x n), where n denotes the number of nodes (e.g. a nurse) in the network. Each row and column label corresponded to either the nominating nurse or nominee. Entries were binary so that a value of 1 indicates a peer nomination. A sociogram (Fig. 1) was also created to show the sources and direction of safe patient handling nominations.
Fig. 1

Sociogram depicting peer-based advice-seeking nominations about safe patient handling in a sample of patient-care workers (n = 38). The head of the arrow signals the direction of the nomination. Size of the figures reflect averages of self-reported equipment use. Peer leaders identified with SNA are shown in black diamonds (n = 3). Peer leaders identified by supervisors are illustrated with grey triangles (n = 3). Workers identified by both SNA and supervisors are depicted with grey squares (n = 5)

Pearson correlations between network centrality measures and safety and interpersonal attributes were computed to assess the salience of each of the seven peer leader definition criteria. Hypothesis 1 was tested comparing safety and interpersonal attributes of peer leaders identified by SNA vs. the rest of participants. Hypothesis 2 was examined contrasting safety and interpersonal attributes between non-overlapping SNA-identified peer leaders and supervisor-identified peer leaders. Hypothesis 3 was tested comparing safety and interpersonal attributes between overlapping workers who were identified by both techniques (SNA and supervisors) and peer leader identified by either SNA or supervisors. Comparison between groups was conducted with a Kruskal–Wallis, a non-parametric alternative test to a one-way ANOVA or with a Fischer’s exact test for small samples with fewer than five members. Given that the small sample size was not powered to detect statistically significant differences, Hedges’ g (i.e., differences in group means divided by weighted standard deviations -g-) were computed as primary effect size measure (negligible differences or g ≤ 0.3 were ignored) (Sullivan and Feinn 2012).

Results

Table 1 displays the characteristics of peer leaders identified by SNA (n = 6) and the rest of their co-workers. The proportion of women, CNAs and night workers were statistically significantly lower among peer leaders than in the rest of the staff. Peer leaders identified with SNA outscored the rest of their co-workers (Hypothesis 1) in equipment use (g = 0.82), safety participation (g = 0.67) and peer support for safety (d = 0.34). Likewise, peer leaders outscored their co-workers in effect of teamwork (g = 0.48), and team communication (g = 0.47).
Table 1

Comparison between SNA identified peer leaders vs. unit staff

 

Overall Sample

Unit Staff

SNA identified Peer Leaders

 

(n = 38)

(n = 32)

(n = 6)

M/%

SD

M/%

SD

M/%

SD

g/χ a

Sociodemographic variables

 Age (yrs.)

40.97

12.56

40.66

12.91

42.8

11.41

 

 Female (%)

86.8

 

93.8

 

50.0

 

8.50**

 Non-Hispanic White (%)

78.9

 

78.1

 

83.3

  

 BMI

30.59

6.6

31.62

6.63

25.05

2.44

−1.01**

 Registered Nurse (%)

55.3

 

50

 

83.3

  

 Tenure (yrs.)

5.89

8.48

5.74

9.01

6.44

6.73

 

 Day Shift (%)

50

 

45.8

 

66.7

  

Safety attributes

 Equipment use

7.63

5.19

6.95

5.06

11.25

4.66

0.82

 Safety participation (1–5)

3.81

0.52

3.75

0.49

4.11

0.62

0.67

 Safety compliance (1–5)

3.95

0.51

3.93

0.54

4.06

0.25

 

 Supervisor support for safety (1–5)

3.32

0.95

3.36

0.95

3.08

0.99

 

 Peer support for safety (1–5)

4.29

0.69

4.25

0.71

4.5

0.63

0.34

Interpersonal attributes

 Leader-Member Exchange (1–5)

3.66

0.61

3.64

0.64

3.79

0.37

 

 Effect of teamwork (1–4)

3.39

0.49

3.35

0.49

3.60

0.53

0.48

 Team communication (1–4)

3.44

0.41

3.41

0.40

3.61

0.44

0.47

 Level of teamwork (1–4)

3.29

0.30

3.29

0.31

3.28

0.25

 

Centrality variables

 In-degree

0.18

0.08

0.16

0.06

0.28

0.04

2.19***

 Out-degree

0.19

0.2

0.14

0.17

0.43

0.19

1.57**

 In-power

0.14

0.07

0.13

0.06

0.21

0.04

1.46**

 Out-power

0.11

0.12

0.08

0.11

0.24

0.11

1.38**

 In-reach

0.53

0.12

0.51

0.12

0.64

0.08

1.22**

 Out-reach

0.54

0.37

0.48

0.36

0.86

0.25

1.26**

 Betweenness

0.02

0.04

0.01

0.02

0.09

0.04

2.41**

a Hedges’ g (standardized mean differences); differences ≤0.3 not shown; **p < .05; ***p < .001; p < .1. Comparison between groups performed with Mann-Whitney U test for not normally distributed samples. Fisher’s exact test was used to conduct a chi-square test with one or more cells with an expected frequency of five or less

Figure 1 displays nominations regarding advice on safe patient handling. The head of the arrow signals the direction of the nomination. The size of the nodes mirrors the self-reported frequency of equipment use. Unique peer leaders identified by SNA (n = 3) are shown with black diamonds in Fig. 1. Peer leaders exclusively identified by supervisors (n = 5) are shown in grey triangles in Fig. 1. Three peer leaders were identified by both SNA and supervisors (37.5%) shown in grey squares in Fig. 1. Table 2 shows that SNA-identified peer leaders outscored supervisor identified peer leaders in equipment use (g = 0.66), safety participation (g = 0.80), perceived effect of teamwork (g = 0.91) and team communication (g = 1.1) (Hypothesis 2). However, SNA identified peer leaders had worse perceptions of supervisor support for safety than supervisor identified peer leaders.
Table 2

Comparison between Social Network Analysis (SNA) identified peer leaders vs. Supervisor-identified peer leaders vs. Peer Leaders identified by both techniques

 

A. SNA- identified peer leaders (unique)

B. Supervisor-identified peer leaders (unique)

C. Peer Leaders identified by both techniques

D. Unit Staff

Kruskal -Wallis/ Fisher’s exact test a

A vs B

A vs C

B vs C

(n = 3)

(n = 5)

(n = 3)

(n = 27)

M/%

SD

M/%

SD

M/%

SD

M/%

SD

p value

g b

g

g

Sociodemographic variables

 Age (yrs.)

47.7

11.4

29

2.9

35.5

9.2

42.5

12.9

0.09

2.1**

0.8

−0.8

 Female (%)

33.3

 

80

 

66.7

 

96.3

 

0.02

   

 Non-Hispanic White (%)

100

 

80

 

66.7

 

77.8

     

 BMI

23.2

1.7

33.2

8

26.9

1.2

31.3

6.5

0.04

−1.1

−1.6

0.7

 Registered Nurse (%)

100

 

40

 

66.7

 

51.9

     

 Tenure (yrs.)

6.7

8.1

1.0

0.1

6.2

6.9

6.7

9.7

 

0.9**

  

 Day Shift (%)

100

 

25

 

33.3

 

50

     

Safety attributes

 Equipment use

11.7

6.1

8

2.8

10.8

4

6.8

5.4

 

0.7

 

−0.7

 Safety participation (1–5)

4.1

0.8

3.5

0.4

4.1

0.5

3.8

0.5

 

0.8

 

−1.0

 Safety compliance (1–5)

4.1

0.2

4.1

0.4

4

0.3

3.9

0.6

    

 Supervisor support for safety (1–5)

2.9

1.1

3.3

0.7

3.2

1.1

3.4

1

 

−0.4

  

 Peer support for safety (1–5)

4.5

0.9

4.7

0.4

4.5

0.5

4.2

0.7

    

Interpersonal attributes

 Leader-Member Exchange (1–5)

3.7

0.3

3.4

0.9

3.9

0.5

3.7

0.6

   

−0.5

 Effect of teamwork (1–4)

3.9

0.1

3.4

0.5

3.3

0.7

3.3

0.5

 

0.9

0.8

 

 Team communication (1–4)

3.8

0.4

3.3

0.3

3.4

0.5

3.4

0.4

 

1.1

0.6

 

 Level of teamwork (1–4)

3.2

0.2

3.3

0.3

3.3

0.3

3.3

0.3

    

a Comparison between groups with Kruskal –Wallis test as a nonparametric alternative to the one-way ANOVA, and comparison of more than two independent groups. Fisher’s exact test was used to conduct a chi-square test for one or more cells with an expected frequency of five or less

b Hedges’ g (standardized mean differences comparing two groups); differences ≤0.3 not shown; **p < .05; ***p < .001; p < .1. Mann-Whitney U test was used to compare two groups for not normally distributed sample

Peer leaders identified by both SNA and supervisors had similar scores in safety and interpersonal attributes than those uniquely identified by SNA, but outscored those peer leaders identified only by supervisors in equipment use (g = 0.7), safety participation (g = 1.0) and LMX (g = 0.5) (Hypothesis 3). Peer leaders uniquely identified by SNA had higher scores in perceived effect of teamwork (g = 0.8) and team communication (g = 0.6) than peer leaders identified by both methods.

Table 3 shows the correlations between centrality measures and safety and interpersonal attributes. Equipment use was significantly associated with out-degree (r = .35), out-power (r = .35), out-reach (r = .36) and betweenness (r = .41). Safety participation was associated with out-power (r = .37) and out-reach (r = .38).
Table 3

Pearson’s correlations between utilization of safety, interpersonal and centrality variables (n = 38)

 

A.

B.

C.

D.

E.

F.

G.

H.

I.

J.

K.

L.

M.

N.

O.

P.

A. Equipment use

1

               

B. Safety participation

.24

1

              

C. Safety compliance

.19

.37*

1

             

D. Supervisor support for safety

−.1

.24

.45**

1

            

E. Peer support for safety

.31

−.04

.34*

.33*

1

           

F. Leader-Member Exchange

05

.18

.24

−.01

−.02

1

          

G. Effect of teamwork

.19

.01

.54**

.01

.34*

.18

1

         

H. Team communication

.22

.01

.03

−.32

.06

.27

.62**

1

        

I. Level of teamwork

.15

−.04

.36*

−.02

.15

−.07

.57**

.52**

1

       

J. In-degree

.16

.14

.02

−.02

.05

.17

.01

.10

−.21

1

      

K. Out-degree

.35*

.27

.16

.04

.03

.01

.13

−.01

−.02

.16

1

     

L. In-beta power

.02

−.01

−.1

−.06

−.08

.17

−.09

.04

−.31

.92**

.07

1

    

M. Out-beta power

.35*

.37*

.15

.05

−.05

−.02

.11

.01

−.02

.03

.91**

−.1

1

   

N. Out-reach

.36*

.38*

.21

.28

.2

−.08

.09

−.07

−.06

.02

.76**

−.15

.81**

1

  

O. In-reach

−.05

−.02

−.03

−.01

−.06

.17

−.09

−.04

−.27

.78**

.19

.91**

−.03

−.1

1

 

P. Betweenness

.41*

.19

.14

.01

.28

.14

.18

.10

.02

.48**

.79**

.36*

.63**

.58**

.41*

1

*p < 0.05; **p < 0.001

Discussion

This study showcased Social Network Analysis (SNA) as a proof of concept to identify peer leaders for safe patient handling in a community hospital. Following the notion of opinion leaders per the Diffusion of Innovations, peer leaders were defined as those perceived both as a top source of safe patient handling advice and who could also reach and interconnect advice from most workers. Based on ideal attributes of peer leaders as described in program guidelines, detected peer leaders also demonstrated appropriate safety and interpersonal attributes, an empirical basis to support their peer leadership. Workers uniquely identified as peer leaders with SNA outscored those uniquely identified as peer leaders by supervisors, an alternative approach to often used to identify such leaders. Moreover, combining SNA with supervisor-identification of peer leaders did not appear to have additional merits over using only SNA. The magnitude of the differences in safety and interpersonal attributes ranged from medium to large effect sizes.

SNA-identified peer leaders use patient-transfer equipment more frequently and reported greater commitment to promoting workplace safety than the rest of the co-workers. Likewise, SNA-identified peer leaders exhibited a tendency for teamwork, advanced training, increased availability during the day shift where transfers are more common, and physical traits (e.g. height) that may be useful for patient transfers. SNA-identified leaders were more critical about supervisors’ role in supervisor support for safety than other staff. This finding suggests that SNA identified peer leaders may be more aware of actual safety practices than their supervisors, a conclusion shared by other studies that have document how formal supervisors may overestimate the safety situation of a hospital (Hurtado et al. 2017). However, since peer-based nominations were explicitly about safe patient handling, other non-measured relational facets between peer leaders and supervisors could explain this negative difference.

Despite the fact that peer leaders detected by SNA had greater in-degree –a proxy of peer influence (Tsai et al. 2016)- than the rest of their co-workers, this measure was not correlated with safety and interpersonal attributes. For instance, worker #60 as the most central node in Fig. 1 was not selected as a peer leader. This worker had a high in-degree but a low out-degree, meaning that although several peers considered him or her as an expert, this worker did not have many other peers available for safe patient handling support. Since innovations usually spread faster through those individuals who can reach or bridge many other network members (Holliday et al. 2016), in-degree alone was not enough as a criterion for peer leadership selection. The collaborative nature of safe patient handling, which often requires multiple people for lift-assisted transfers, plausibly explain why outward network variables such as out-reach or betweenness were correlated with equipment use and safety participation. Lastly, SNA-identified peer leaders may be desirable to target for future safety initiatives to help ensure they are working in conjunction of emerging safety initiatives rather than using their peer influence in less-desirable ways (e.g., unintentionally spreading misinformation).

Limitations

This exploratory study had a small sample size; however, information was maximized with SNA, a suitable technique to comprehend dynamics among group members. Although self-reported data had appropriate psychometrics, self-reports are susceptible to recall biases, calling the need for studies to complement self-reported data with administrative records, observations or real-time counters in the case of equipment use. Measuring equipment use in real time was not feasible and limited to self-reports. Additionally, measures of trust and availability were not available, and critical to understand the attributes that make ideal peer leaders who exert pro-social social influence. The statistical and construct validity of the results lie in the magnitude and direction of the differences in safety and psychosocial attributes. However, prospective or interventional research must examine the predictive validity of this approach. A practical limitation is the complexity of SNA, so at the very least, programs should inquire about workers’ opinions about reputable peers to complement other approaches such as supervisor or self-identification of peer leaders.

Conclusions

SNA identified workers who were central for the diffusion of safe patient handling advice, and who demonstrated adequate characteristics to exert safety influence. SNA-identified peer leaders outscored supervisor-identified peer leaders as well as their co-workers in relevant safety and interpersonal attributes. SNA detects naturally-occurring peer leaders which could be further trained and enrolled in an occupational health and safety program that leverages on their recognized social influence. Analysis of peer-based supportive interactions is relevant for safe patient handling and other workplace programs aiming at instituting peer leaders to champion safer norms and practices.

Notes

Acknowledgements

This study was funded by the Oregon Institute of Occupational Health Sciences at Oregon Health & Science University. Special thanks to our organizational healthcare partner as well as to W. Kent Anger, PhD, Leslie B. Hammer, PhD and the Oregon Healthy Workforce Center.

References

  1. Alamgir, H., Drebit, S., Li, H. G., Kidd, C., Tam, H., & Fast, C. (2011). Peer coaching and mentoring: A new model of educational intervention for safe patient handling in health care. American Journal of Industrial Medicine, 54(8), 609–617.  https://doi.org/10.1002/ajim.20968.CrossRefPubMedGoogle Scholar
  2. Alexopoulos, E. C., Tanagra, D., Detorakis, I., Gatsi, P., Goroyia, A., Michalopoulou, M., & Jelastopulu, E. (2011). Knee and low back complaints in professional hospital nurses: Occurrence, chronicity, care seeking and absenteeism. Work, 38(4), 329–335.PubMedGoogle Scholar
  3. Allen, D., Weinhold, M., Miller, J., Joswiak, M. E., Bursiek, A., Rubin, A., et al. (2015). Nurses as champions for patient safety and interdisciplinary problem solving. Medsurg Nursing, 24(2), 107–111.PubMedGoogle Scholar
  4. Baptiste, A., Boda, S. V., Nelson, A. L., Lloyd, J. D., & Lee, W. E. I. (2006). Friction-reducing devices for lateral patient transfers. AAOHN Journal, 54(4), 173–180.  https://doi.org/10.1177/216507990605400407.CrossRefPubMedGoogle Scholar
  5. Bauer, T. N., & Green, S. G. (1996). Development of leader-member exchange: A longitudinal test. Academy of Management Journal, 39(6), 1538–1567.  https://doi.org/10.2307/257068.CrossRefGoogle Scholar
  6. Borgatti, S. P. (2006). Identifying sets of key players in a social network. Computational & Mathematical Organization Theory, 12(1), 21–34.CrossRefGoogle Scholar
  7. Borgatti, S. P., Everett, M. G., & Freeman, L. C. (2002). Ucinet for windows: Software for social network analysis. Harvard: Analytic Technologies.Google Scholar
  8. Borgatti, S. P., Everett, M. G., & Johnson, J. C. (2013). Analyzing social networks. Thousand Oaks (USA): SAGE Publications Limited.Google Scholar
  9. Caspi, C. E., Dennerlein, J. T., Kenwood, C., Stoddard, A. M., Hopcia, K., Hashimoto, D., & Sorensen, G. (2013). Results of a pilot intervention to improve health and safety for healthcare workers. Journal of occupational and environmental medicine/American College of Occupational and Environmental Medicine, 55(12), 1449.CrossRefPubMedCentralGoogle Scholar
  10. Cho, Y., Hwang, J., & Lee, D. (2012). Identification of effective opinion leaders in the diffusion of technological innovation: A social network approach. Technological Forecasting and Social Change, 79(1), 97–106.CrossRefGoogle Scholar
  11. Dearing, J. W. (2009). Applying diffusion of innovation theory to intervention development. Research on Social Work Practice, 19(5), 503–518.CrossRefPubMedPubMedCentralGoogle Scholar
  12. Dressner, M. A. (2017). "Hospital workers: An assessment of occupational injuries and illnesses,". Monthly Labor Review: U.S. Bureau of Labor Statistics.Google Scholar
  13. Greenhalgh, T., Robert, G., Macfarlane, F., Bate, P., & Kyriakidou, O. (2004). Diffusion of innovations in service organizations: Systematic review and recommendations. The Milbank Quarterly, 82(4), 581–629.CrossRefPubMedPubMedCentralGoogle Scholar
  14. Holliday, J., Audrey, S., Campbell, R., & Moore, L. (2016). Identifying well-connected opinion leaders for informal health promotion: The example of the ASSIST smoking prevention program. Health Communication, 31(8), 946–953.CrossRefPubMedGoogle Scholar
  15. Hurtado, D. A., Kim, S.-S., Subramanian, S., Dennerlein, J. T., Christiani, D., Hashimoto, D., & Sorensen, G. (2017). Nurses’ but not supervisors’ safety practices are linked with job satisfaction. Journal of Nursing Management.  https://doi.org/10.1111/jonm.12484.CrossRefPubMedGoogle Scholar
  16. Iyengar, R., Van den Bulte, C., & Valente, T. W. (2011). Opinion leadership and social contagion in new product diffusion. Marketing Science, 30(2), 195–212.CrossRefGoogle Scholar
  17. Kaissi, A., Johnson, T., & Kirschbaum, M. S. (2003). Measuring teamwork and patient safety attitudes of high-risk areas. Nursing Economics, 21(5), 211.PubMedGoogle Scholar
  18. Krill, C., Staffileno, B. A., & Raven, C. (2012). Empowering staff nurses to use research to change practice for safe patient handling. Nursing Outlook, 60(3), 157-162. e151.CrossRefGoogle Scholar
  19. Kutash, M., Short, M., Shea, J., & Martinez, M. (2009). The lift team's importance to a successful safe patient handling program. Journal of Nursing Administration, 39(4), 170–175.CrossRefPubMedGoogle Scholar
  20. Lawton, R., & Parker, D. (2002). Barriers to incident reporting in a healthcare system. BMJ Quality & Safety, 11(1), 15–18.CrossRefGoogle Scholar
  21. Lee, S.-J., & Lee, J. H. (2017). Safe patient handling behaviors and lift use among hospital nurses: A cross-sectional study. International Journal of Nursing Studies, 74, 53–60.CrossRefPubMedGoogle Scholar
  22. Matz, M. (2005). Unit-based peer safety leaders to promote safe patient handling. In A. Nelson (Ed.), Safe patient handling and movement: A practical guide for health care professionals (pp. 139–150). New York: Springer.Google Scholar
  23. Neal, A., & Griffin, M. A. (2006). A study of the lagged relationships among safety climate, safety motivation, safety behavior, and accidents at the individual and group levels. Journal of Applied Psychology, 91(4), 946–953.  https://doi.org/10.1037/0021-9010.91.4.946.CrossRefPubMedGoogle Scholar
  24. Nelson, A. L., & Baptiste, A. S. (2006). Evidence-based practices for safe patient handling and movement. Clinical Reviews in Bone and Mineral Metabolism, 4(1), 55–69.  https://doi.org/10.1385/BMM:4:1:55.CrossRefGoogle Scholar
  25. Nelson, A. L., Matz, M., Chen, F., Siddharthan, K., Lloyd, J., & Fragala, G. (2006). Development and evaluation of a multifaceted ergonomics program to prevent injuries associated with patient handling tasks. International Journal of Nursing Studies, 43(6), 717–733.  https://doi.org/10.1016/j.ijnurstu.2005.09.004.CrossRefPubMedGoogle Scholar
  26. Olinski, C., & Norton, C. E. (2017). Implementation of a safe patient handling program in a multihospital health system from inception to sustainability: Successes over 8 years and ongoing challenges. Workplace Health & Safety, 65(11), 546–559.CrossRefGoogle Scholar
  27. Ørhede, E., Hottinen, V., Skogstad, A., Knardahl, S., Elo, A.-L., Dallner, M., & Gamberale, F. (2000). User’s guide for the QPSNordic: General Nordic Questionnaire for psychological and social factors at work. Copenhagen: Nordic Council of Ministers.Google Scholar
  28. Ploeg, J., Skelly, J., Rowan, M., Edwards, N., Davies, B., Grinspun, D., et al. (2010). The role of nursing best practice champions in diffusing practice guidelines: A mixed methods study. Worldviews on Evidence-Based Nursing, 7(4), 238–251.CrossRefPubMedGoogle Scholar
  29. Powell-Cope, G., Toyinbo, P., Patel, N., Rugs, D., Elnitsky, C., Hahm, B., et al. (2014). Effects of a national safe patient handling program on nursing injury incidence rates. Journal of Nursing Administration, 44(10), 525–534.CrossRefPubMedGoogle Scholar
  30. Rickett, B., Orbell, S., & Sheeran, P. (2006). Social-cognitive determinants of hoist usage among health care workers. Journal of Occupational Health Psychology, 11(2), 182.CrossRefPubMedGoogle Scholar
  31. Rogers, E. M. (2010). Diffusion of innovations (5th ed.). New York: Simon and Schuster.Google Scholar
  32. Schneider, J. A., Zhou, A. N., & Laumann, E. O. (2015). A new HIV prevention network approach: Sociometric peer change agent selection. Social Science & Medicine, 125, 192–202.  https://doi.org/10.1016/j.socscimed.2013.12.034.CrossRefGoogle Scholar
  33. Springer, P. J., Lind, B. K., Kratt, J., Baker, E., & Clavelle, J. T. (2009). Preventing employee injury: Implementation of a lift team. AAOHN Journal, 57(4), 143–148.CrossRefPubMedGoogle Scholar
  34. Spruce, L. (2015). Back to basics: Implementing evidence-based practice. AORN Journal, 101(1), 106-114. e104.CrossRefGoogle Scholar
  35. Sullivan, G. M., & Feinn, R. (2012). Using effect size—or why the P value is not enough. Journal of Graduate Medical Education, 4(3), 279–282.CrossRefPubMedPubMedCentralGoogle Scholar
  36. Tompa, E., Dolinschi, R., Alamgir, H., Sarnocinska-Hart, A., & Guzman, J. (2016). A cost-benefit analysis of peer coaching for overhead lift use in the long-term care sector in Canada. Occupational and Environmental Medicine, 73(5), 308–314.  https://doi.org/10.1136/oemed-2015-103134.CrossRefPubMedGoogle Scholar
  37. Tsai, J., Valente, T. W., Miller, K. A., De La Haye, K., Pickering, T. A., & Cockburn, M. G. (2016). Friendship networks and sun safety behavior among children. Network Science, 4(03), 314–335.CrossRefGoogle Scholar
  38. Valente, T. W., & Davis, R. L. (1999). Accelerating the diffusion of innovations using opinion leaders. The Annals of the American Academy of Political and Social Science, 566(1), 55–67.CrossRefGoogle Scholar
  39. Valente, T. W., & Pumpuang, P. (2007). Identifying opinion leaders to promote behavior change. Health Education & Behavior, 34(6), 881–896.  https://doi.org/10.1177/1090198106297855.CrossRefGoogle Scholar
  40. Valente, T. W., Hoffman, B. R., Ritt-Olson, A., Lichtman, K., & Johnson, C. A. (2003). Effects of a social-network method for group assignment strategies on peer-led tobacco prevention programs in schools. American Journal of Public Health, 93(11), 1837–1843.CrossRefPubMedPubMedCentralGoogle Scholar
  41. Valente, T. W., Ritt-Olson, A., Stacy, A., Unger, J. B., Okamoto, J., & Sussman, S. (2007). Peer acceleration: Effects of a social network tailored substance abuse prevention program among high-risk adolescents. Addiction, 102(11), 1804–1815.CrossRefPubMedGoogle Scholar
  42. Zadvinskis, I. M., & Salsbury, S. L. (2010). Effects of a multifaceted minimal-lift environment for nursing staff: Pilot results. Western Journal of Nursing Research, 32(1), 47–63.  https://doi.org/10.1177/0193945909342878.CrossRefPubMedGoogle Scholar
  43. Zohar, D., & Luria, G. (2005). A multilevel model of safety climate: Cross-level relationships between organization and group-level climates. Journal of Applied Psychology, 90(4), 616.CrossRefPubMedGoogle Scholar
  44. Zohar, D., & Tenne-Gazit, O. (2008). Transformational leadership and group interaction as climate antecedents: A social network analysis. Journal of Applied Psychology, 93(4), 744.CrossRefPubMedGoogle Scholar

Copyright information

© Springer Nature Switzerland AG 2018

Authors and Affiliations

  • David A. Hurtado
    • 1
  • Lisset M. Dumet
    • 1
  • Samuel A. Greenspan
    • 1
  • Yaritza I. Rodríguez
    • 1
  • Gregory A. Heinonen
    • 1
  1. 1.Oregon Institute of Occupational Health SciencesOregon Health & Science UniversityPortlandUSA

Personalised recommendations