Table 2 Summary of included studies (N = 21)

Aarons 2006 [81]

TfL, TrL

IC

N/a

Obs

CMH

NI

TrL and TfL positively predicted providers' EBPA.

Aarons et al. 2012 [82]

TfL

OC, IC; L

N/a

Exp

CW

I

During EBP implementation, TfL positively predicted InnCli, and InnCli positively predicted providers' EBPA. When delivering services as usual, LMX mediated the relation between leaders' TfL and InnCli.

Aarons et al. 2015 [28]

TfL

OC

N/a

Obs

MH

NI

OrgCul suffered more where leaders rated their TfL more positively than providers rated them, in contrast to where leaders rated themselves lower than providers. OrgCul tended to be better when providers and leaders agreed the leaders’ TfL was high than when they agreed it was low.

Aarons et al. 2016 [83]

TfL, TrL; PaL, LB

N/a

Sus

Obs

CMH, CW, Oth

Sus

TfL positively predicted sustainment, and PaL negatively predicted sustainment. Leaders’ ongoing championing of EBP and practical support for providers facilitated sustainment.

Aarons et al. 2017 [45]

IL

OC

N/a

Obs

MH

NI

OrgCli of involvement and performance feedback were highest when leaders rated their IL low and providers rated leaders’ IL high. Involvement climate did not differ when leaders and providers agreed that IL was strong compared with when they agreed it was weak. Performance feedback climate was higher when leaders and providers agreed that IL was strong.

Brimhall et al. 2016 [84]

TfL

OC, IC

N/a

Obs

CMH

NI

Greater TfL indirectly influenced providers' perceptions of EBPs as less burdensome through higher EmpCli and lower DemoCli.

Bunger et al., 2019 [65]

LB

OC

N/a

Obs

CW

I

Leaders’ activities influenced aspects of ImpCli including conveying expectations, providing support, and rarely rewarding implementation. Leaders conveyed expectations through diffusing information, synthesizing information, mediating between agency strategy and day-to-day activities, and selling implementation. Leaders supported implementation through diffusing, synthesizing, and mediating. They conveyed rewards through diffusing.

Corrigan et al. 2002 [85]

TfL, TrL, PaL

OC, IC

N/a

Obs

H, MH

NI

TfL was positively associated with transformational OrgCul. LfL and passive MBE were negatively associated with a TrC. TfL was positively associated with TrC based on leader report, but negatively associated with TrC based on provider report. Passive MBE was positively associated with TrC based on provider report. TfL was negatively associated with burnout among providers and leaders.

Fenwick et al. 2018 [86]

TfL

IC, F

N/a

Obs

CMH

NI

TfL and LMX positively predicted providers' attitudes toward feedback. LMX mediated the relation between TfL and providers’ attitudes toward feedback.

Fleury et al. 2014 [87]

LB

N/a

GI

Obs

MH, Oth

P, I

Inaccessibility of leaders, leader turnover, and leaders’ poor communication were barriers to implementation.

Green et al. 2014 [88]

TfL

OC

N/a

Obs

CMH

NI

Leaders’ TfL positively predicted EmpCli.

Guerrero et al. 2014 [89]

LC

N/a

A

Obs

AH

NR

Leaders’ EBPA and readiness-for-change attributes positively predicted implementation of contingency management treatment. Leaders’ openness towards EBPs positively predicted implementation of medication-assisted treatment.*

Guerrero et al. 2020 [15]

IL

IC

A

Obs

AH

I

IL was positively associated with provider's EBPA. IL did not mediate the relation among top leaders’ TfL and A.

Mancini et al. 2009 [90]

LB

N/a

GI

Obs

H, MH

I

Leaders’ failure to empower staff, poor organizational skills, poor management of internal dynamics and workload, and turnover were barriers to high-F implementation. Leaders understanding of the model, effective management of team dynamics, holding staff accountable, advocating on behalf of provider teams, empowering staff, conveying a sense of mission to the provider team, and equitably distributing the workload facilitated implementation.

Moser et al. 2005 [91]

LB, LC

N/a

GI

Obs

MH

P, I

Leaders’ turnover, lack of familiarity with the intervention, and lack of investment in implementation were barriers to implementation. Leader familiarity with the intervention facilitated implementation.

Powell et al. 2017 [92]

TfL, IL

IC

N/a

Obs

CMH

I, Sus, NI

TfL-idealized influence positively predicted providers’ knowledge of EBP. TfL-individual consideration negatively predicted providers' EBPA. Proactive IL positively predicted providers’ EBPA. Perseverant IL negatively predicted providers’ EBPA.

Rapp et al. 2010 [93]

LB

N/a

GI

Obs

MH

I

Leader behaviors were the greatest barrier to implementation, including: not setting expectations; only providing consultation on service-delivery when challenges arose; lacking prescriptions or structure to providers' practice; being overly conflict-avoidant; lacking meaningful feedback for providers; having only superficial knowledge of clinician practice; relying on coaxing and persuasion with no consequences for poor performance; poorly leading group supervision, which was dominated by administrative tasks; lacking follow-through; having competing responsibilities; lacking knowledge of EBP skills and feeling inadequate at supervising practice.

Savill et al. 2018 [94]

LB

N/a

GI

Obs

MH, SMH

I

Leaders facilitated implementation by working to incorporate EBP procedures into existing workflows (i.e., assessment checklists and forms) and meeting regularly with senior administrators and staff to monitor and troubleshoot implementation difficulties.

Van Erp et al. 2007 [95]

LB, LC

N/a

GI

Obs

MH

I

Lack of time for leaders to manage the intervention was a barrier to implementation. Leaders' strong personal commitment demonstrated by their dedication and enthusiasm to implement the intervention facilitated implementation.

Van Erp et al. 2009 [96]

LB

N/a

GI

Obs

H, MH

I

Leaders’ inability to administer the implementation process and to realize necessary conditions for implementation were barriers to implementation. Leader motivation facilitated implementation.

Williams et al. 2020 [97]

IL

OC

A

Quas

CMH

Mul

Increases in IL had a significant indirect effect on increases in clinicians’ EBP use via improvement in EBP ImpCli.