Table 1 Study characteristics

Protocol-Guided Rapid Evaluation of Veterans

Experiencing New Transient Neurological Symptoms (PREVENT)

Bravata 2020; Non-randomised cluster trial, Intervention performance

Bravata 2022, Mixed methods, Evaluation of QI sustainment

Damush 2021a, Mixed methods, Evaluation of implementation bundle

Damush 2021b, Mixed methods, Evaluation of intervention acceptability

Penney 2021, Mixed methods, Stakeholder perception of external facilitation

Rattray 2020, Observational qualitative, Evaluation of virtual “Hub” support of QI

USA

Patients with inpatient treatment of TIA. 42 hospitals

Intervention:

N = 162 patients from 6 hospitals

Control: N = 973 patients from 36 hospitals

External: research team, including QI nurse and QI physician

Remote: facilitator support via phone, email or Skype

In-person: in-person one-day kick-off meeting

Facilitation dose: any interaction with external facilitator by phone, email, Skype, or in-person during one-year active facilitation period. Median of 24 facilitation episodes (mean of 8 education sessions, 10 quality process monitoring, 12 planning, 11 networking and 23 nurse external facilitator interactions) across 6 hospitals

Intervention: QI to improve TIA quality of care: quality of care reporting system, performance feedback, clinical progression, professional education, electronic health record tools, virtual collaborative. Online QI plans monitored in web-based audit and feedback system, with goal setting and monthly collaborative conferences for one year

Facilitation strategy: external facilitation study team (primarily QI nurse and QI physician) supported education, team planning, goal setting; provided feedback; provided strategies for reflection and evaluation on practice

One-day kick-off meeting, attended in-person or via teleconference, to develop targeted site-specific action plan with short- and long-term activities and plans

Post-active implementation for one year, external facilitation no longer initiated by study team (sustainment)

Comparator: matched CG site based on TIA patient volume, facility, and TIA quality of care¹

For 6 IG sites: pre-stepped wedge allocation usual care

(1) At 1 year: increase in without-fail rate (patients received all 7 guideline-recommended TIA processes of care), p = 0.01

During sustainment: NS difference between groups

(2) Increased measure of TIA quality (TIA processes of care received, as proportion of processes eligible), p = 0.01

During sustainment: NS difference between groups

(3) NS change in anti-coagulation for atrial fibrillation

(4) NS change in anti-thrombotic medication administered

(5) NS change in brain imaging

(6) NS change in carotid artery imaging

(7) NS change in moderate/high potency statin therapy

(8) NS change in HT control

(9) NS change in neurological consultations

(10) NS change in 90 d recurrent stroke rate

(11) NS change in 90 d recurrent vascular event rate

(12) NS change in 90 d mortality rate¹

(13) In 6 IG hospitals: hospitals implemented 15–39 activities; mean 26.5 activities across sites. Pre-post difference NR

(14) In 6 IG hospitals: increased team organisation score for improving TIA care, from 1 to 4 to 8, out of 10; mean 6.67 across sites. Between group differences NR

(15) In 6 IG hospitals: 4 of 6 sites with ≥ 15 point improvement in without-fail rate

(16) Variability in external facilitation activity: increased facilitation activity related to increased site implementation activity, or to prompt action during site inactivity⁵. Facilitators supported critical junctures of success/failure at two sites. Loss of external facilitation support at sustainment led to decreased motivation, team activity and coordination, if PREVENT not implemented into routine practice

Stroke123

Cadilhac 2019, Prospective before-and-after study, Intervention effectiveness

Thayabaranathan 2021, Process evaluation of external facilitator role

Australia

Adult patients admitted for stroke or TIA. 19 hospitals

Intervention:

N = 2682

Control:

N = 5596

External: external staff with nursing or allied health QI officer

Remote: Telephone or email support

In-person: in-person 1 × 3 h workshop. Face-to-face peer support after workshop as needed

Facilitation dose: 1 × 3 h on-site workshop; ongoing phone, email or face-to-face support. Median 22/mean 30 h of external facilitator support across all hospitals

Intervention: organisational INT to improve quality of stroke care, with 2 stages: financial incentives to establish or enhance stroke units (time 1), financial incentives with QI external facilitator (time 2)

Facilitation strategy: external facilitator delivered workshop to review hospital performance, provided site-specific strategies, developed action plan using PDSA to improve clinical guideline adherence. Ongoing peer support, via phone, email or face-to-face. Additional support for remote education, networking and shared learning

Comparator: Pre-INT usual care

(1) Increased composite score (received all 8 guideline-recommended stroke processes of care) post-QI facilitation stage (post-time 2), p < 0.001. After adjustment of patient confounders, p = 0.007

Increased composite score post-INT (baseline to 13 mo post-INT), p < 0.001. After adjustment of patient confounders, p = 0.005

(2) Increased patients treated in stroke unit over time (baseline to 13 mo post-INT), p < 0.001

(3) NS difference in receiving thrombolysis for ischemic stroke

(4) Increased patients with early mobilisation over time (baseline to 13 mo post-INT), p < 0.001

(5) Increased swallow screening or assessment over time (baseline to 13 mo post-INT), p < 0.001

(6) Increased early aspirin delivery over time (baseline to 13 mo post-INT), p = 0.047

(7) Increased discharge with anti-HT medication over time (baseline to 13 mo post-INT), p < 0.001

(8) Increased discharge with anti-thrombotic medication over time (baseline to 13 mo post-INT), p < 0.001

(9) Increased discharge to community with care plan over time (baseline to 13 mo post-INT), p < 0.001

(10) No association between amount of external facilitation and absolute change in composite score. 13 of 19 hospitals with increased mean composite score⁸. 9 of 18 hospitals with increased absolute change in composite score

Shared Team Efforts Leading to Adherence Results (STELAR)

Cadilhac 2022, Stepped-wedge cluster RCT, Intervention effectiveness

Cadilhac 2017, Mixed-methods controlled before–after observational study, Pilot test of intervention

Australia

Adult patients admitted for stroke or TIA

Intervention RCT: N = 9 hospitals, 2146 patients

Control RCT: N = 9 hospitals, 1001 patients

Intervention pilot: N = 2 hospitals, 438 episodes of care

Control pilot: N = 2 hospitals, 419 episodes of care

External: research team, including implementation researcher and nurse researcher

Remote: Telephone or email support

In-person: 1 × 2 h workshop

Facilitation dose: 1 × 1 h videoconference, 1 × 2 h in-person workshop, ongoing phone or email support for 2 mo

Intervention: two-stage INT to improve adherence to acute stroke care clinical indicators. Audit with feedback; tailored site-specific action plan; nomination of clinical champion

Facilitation strategy: external facilitators delivered two workshops for audit feedback, education on clinical guideline adherence, design of site-specific action plan. Ongoing remote support via telephone or email

Comparator: Pre-INT usual care

(1) RCT: increased composite score for all site-specific plan on guideline-recommended stroke processes of care [below] post-INT, p = 0.016

Pilot: increased composite score post-INT, p < 0.001¹⁰. Sustained improvement at one year post-INT sustainment, p = 0.08

(2) RCT: increased odds of treatment in stroke unit, p < 0.05

(3) RCT: NS difference in swallow screening or assessment before oral intake

(4) RCT: NS difference in early mobilisation

(5) RCT: NS difference in discharge with anti-HT medication

(6) RCT: increased odds of discharge with anti-HT medication, p < 0.05

Pilot: increased odds of discharge with anti-HT medication, p = 0.001. NS difference at one year post-INT sustainment

(7) RCT: increased odds of discharge with care plan, p < 0.05

Pilot: increased odds of discharge with care plan, p < 0.001¹⁰. NS difference at one year post-INT sustainment

(8) RCT: NS difference in receiving thrombolysis, for ischemic stroke

(9) RCT: NS difference in receiving thrombolysis at ≤ 60 min of admission, for ischemic stroke

(10) RCT: NS difference in receiving anti-platelet medication at ≤ 48 h of admission, for ischemic stroke

(11) RCT: NS difference in discharge with lipid-lowering medication, for ischemic stroke

(12) RCT: NS difference in 30 d mortality rate

(13) RCT: NS difference in 90 d mortality rate

(14) RCT: NS difference in disability status at 90–180 d follow up

Jolliffe 2020, Non-randomised three-arm cluster-controlled feasibility study, Feasibility, efficacy and acceptability of two implementation packages

Australia

Stroke patients with upper limb stroke rehabilitation with OT or PT. 6 hospital ward or community neurological rehabilitation sites

Intervention A:

N = 20

Intervention B: N = 17

Control: N = 18

External: research team

Remote: Intervention delivered via Trello for group B and participants were encouraged to post any comments/questions to the research team

In-person: 6 × 45 min face-to-face education sessions and fortnightly/monthly coaching/mentoring for Intervention group A

Facilitation dose: fortnightly/monthly coaching with facilitator; and 6 × 45 min face to face education, for 3 mo

Intervention group A: Facilitator-mediated implementation of rehabilitation guidelines. Includes point of care videos, face to face education, online education modules, written manuals, fortnightly/monthly coaching and mentoring, auditing and feedback, access to physical resources, environmental alterations for encourage patient practice

Intervention group B: self-directed implementation of rehabilitation guidelines. Includes point of care videos, online education modules, posters with guideline recommendations, written manuals, physical resources, patient handouts. INT content delivered online

Comparator: usual care

(1) Increased adherence to rehabilitation OT/PT guideline: IG 1 vs IG 2, p < 0.0001; IG 1 vs CG, p < 0.0001; NS IG 2 vs CG

(2) NS change in Box and Block Test, between IG 1 vs IG 2, IG 1 vs CG, IG 2 vs CG

(3) Improved Fugl-Meyer Upper Extremity Assessment upper limb outcomes in IG 2 vs. CG, p = 0.027. NS between IG 1 vs IG 2, and IG 1 vs CG

(4) NS change in self-reported minutes of weekly therapy, between IG 1 vs IG 2, IG 2 vs CG, IG2 vs CG

Carers Count

Levy 2022, Interventional study and mixed methods evaluation, Intervention development and implementation

Australia

Stroke survivors and carers. 1 stroke rehabilitation ward

N = 30 stroke survivors, 30 carers

External: research team

In-person: fortnightly meetings and ongoing staff training

Facilitation dose: fortnightly meetings with team members and management; and ongoing staff training, for 5 mo

Intervention: Facilitation of new exercise-based group of stroke survivors and carers

External facilitator: fortnightly meetings with allied health team. Conducted on-going allied health staff training. Tailored INT to address barriers. Developed implementation scale-up, education resources, manuals. Involved consumer representatives in planning and implementation. Supported electronic medical record data collection

Comparator: pre-trial usual care

Post-trial only, focused on INT instead of facilitator involvement

(1) ≥ 80% agreement on group attendance benefits, satisfaction with staff support in group, understanding of post-stroke physical ability, confidence in post-discharge stroke management, socialisation with group peers

(2) 100% agreement on exercise-based group not increasing post-stroke stressors

(3) Qualitative outcomes: exercise-based group as ‘something to look forward to’, positive shared experience by stroke survivor and carer, decreased isolation, perceived benefits

Moore 2020, Pre-post study, Evaluation of intervention implementation, sustainability and clinician adherence

Canada

Subacute stroke patients with goal to improve walking. 1 subacute inpatient stroke rehabilitation facility

N = 157 patients

External: research team – university and translational researchers

Remote: 5–60 min monthly phone conversations for 15 mo

In-person: 1 × 3 h + 1 × 1 h education sessions for stroke staff; 5 × 1.5 h non-core stroke staff training education sessions

Facilitation dose: various knowledge translation interventions, including weekly team meetings, for 12 mo; education to stroke team and non-core staff; phone call/s

Intervention: multicomponent, iterative implementation plan with facilitation, leadership and knowledge translation interventions, to implement a gait assessment battery into routine clinical practice

Facilitators: researchers, with clinical team, identified site-specific barriers and enablers, then co-designed tailored knowledge translation INT, with education, leadership support, process changes, audit and feedback, equipment purchase, environment modification. Weekly feedback in team meetings. Phone calls to problem solve, discuss adherence, barriers and knowledge translation INTs

Comparator: pre-trial usual care

(1) Increased adherence to Gait and Balance assessment, from 46% at baseline to 85% at 6 mo and 95.2% at 48 mo. Significance NR

(2) Increased use of 10 min walk test, p = 0.03

Stroke Canada Optimization of Rehabilitation by Evidence-Implementation Trial (SCORE-IT)

Munce 2017, Qualitative descriptive, Stakeholder evaluation of barriers and facilitators

Salbach 2017, Quantitative process evaluation of SCORE-IT

Canada

NB: main trial results of SCORE-IT have not been published

Stroke rehabilitation patients. 20 sites

Intervention:

N = 9 sites, 169 patients

Control:

N = 8 sites, 143 patients

Internal: nurse and physical therapist

External: research team

Remote: teleconference, by external facilitators for internal facilitators

In-person: internal facilitators on-site, 4 h/wk for 16 mo

Facilitation dose: 1 × 2d training workshop for internal facilitators; teleconference support for internal facilitators; internal facilitator on-site for 4 h/wk for 16 mo

Intervention: facilitated knowledge translation INT, with internal facilitators at each INT site for 4 h/wk

Internal facilitator: promoted guideline implementation at INT site. Received two-day workshop, including change management, implementation strategies, developing implementation plan to address barriers, and incorporating behaviour change strategies. Contacted other facilitators via teleconference and web-based platform to share successful implementation strategies

Stroke teams provided with stroke rehabilitation guideline, evidence-based treatment protocols, posters, reminder cards

External facilitator: research team provided advice and support to internal facilitators via teleconference

Comparator: passive knowledge translation INT

Received stroke rehabilitation guideline without treatment protocols, handbook, and educational DVD on post-stroke standardised assessment tools. Opt-in facilitator role by motivated staff members

(1) Facilitator staff as leaders/champions supported implementation and provided continuity for trial procedures/tasks in face of high staff turnover. Lack of access to facilitator staff was a barrier: lack of champion, impeded continuity in face of staff turnover, no sustainability for knowledge translation interventions

(2) Increased implementation of sit-to-stand training, p = 0.028, and walking practice, p = 0.043, in IG. Increased implementation of standing balance training, p = 0.037, in CG. Decreased implementation of stretching training and sitting balance training in IG, p < 0.05

Implementation of Partners of Aphasic clients Conversation Training (ImPACT)

Wielaert 2018, Interventional study with mixed methods evaluation, Evaluation of intervention uptake, barriers/facilitators and implementation methods

Wielaert 2017, Qualitative, Exploration of client and stakeholder experiences

Wielaert 2016, Pre-post study, Assessment of stakeholder eligibility

The Netherlands

Partners of patients with aphasia. 7 rehabilitation sites and 3 nursing homes

N = 10 sites

Internal: speech language therapists, 2 per site

External: research team

Remote: telephone consultations, phone contact, email contact, quarterly newsletters

In-person: 4 × 1d meetings, 2 × 2 h outreach visits, 1 × 1d outreach meeting to site

Facilitation dose: 4 × 1d meetings and 2 × 2 h outreach visits for internal facilitators, by external facilitators. 1 × outreach meeting to site by external facilitator. Telephone consultations, phone and email contact, quarterly newsletters, for internal facilitators by external facilitators, across 13 mo

Intervention: multicomponent implementation INT of conversation partner training in aphasia (PACT). Financial support for local coordinators; internal facilitator training; aphasia-friendly education materials; feedback on recruitment, PACT training, implementation issues; phone and email reminders with newsletters

Internal facilitators: implemented PACT on-site – advocated for PACT with multidisciplinary team, provided and engaged clients in PACT. Two meetings for skill training in PACT. Two meetings on developing implementation plans, including goal setting and PACT analysis

External facilitators: trained internal facilitators in 4 meetings. Delivered three outreach visits – 2 for internal facilitators, 1 for multidisciplinary team and manager. Phone and email PACT supervision for internal facilitators after 4th meeting

Comparator: pre-trial usual care

(1) 7 centres with uptake of PACT in care pathway at 8mo post-INT

(2) Components necessarily for INT implementation: financial support; education to deliver PACT for internal facilitators; outreach visit and PACT presentation by external facilitator for multidisciplinary team and manager

iWalk study

Salbach 2022a, Before-and-after study, iWalk toolkit evaluation

Salbach 2022b, Mixed methods process evaluation of intervention implementation

Salbach 2021, Realist evaluation, Site-specific context-mechanism-outcome synthesis

Canada

Physical therapists. 9 hospitals

Intervention:

N = 375 hospital visits, 33 physical therapists, 7 practice leaders

Control:

N = 347 hospital visits

Internal: physical therapist, manager or practice leader providing post-stroke acute or rehabilitation care

External: physical therapist expert on the research team

Remote: email or phone contact

In-person: 3 × 1 h learning sessions by internal facilitator

Facilitation dose: 3 × 1 h learning sessions by internal facilitator; and email or phone support from external facilitator, over 21 mo

Intervention: implementation of iWalk toolkit with 10 m walk test and 6 min walk test in stroke rehabilitation. Toolkit: educational guide, educational video, mobile app

Internal facilitator: set up walkways for walk tests, organised and facilitated 3 learning sessions within a 5-month period

External facilitator: email or phone support

Comparator: pre-trial usual care

(1) Lack of internal facilitator in non-neurology services (e.g. palliative care, emergency care) did not have authority to implement iWalk toolkit: no walk test walkways, no motivation to attend learning sessions

(2) In acute care: internal facilitators reminded physical therapists to use iWalk toolkit and coordinated learning sessions. Internal facilitators being involved in clinical practice gave them more influence over physical therapists

(3) In rehabilitation care: internal facilitators adapted learning session materials, offered off-site attendance by teleconference, or organised on-site training for off-site therapists

(4) Internal facilitators supported daily interaction with physical therapists as reminders of iWalk toolkit use; addressed local policy that limited walkway set up; adapted recommended practice to local context, patient population and negative outcome expectations

(5) Increased administration of 10 m walk test, p < 0.05

(6) Increased administration of 6 min walk test, p < 0.05

(7) Increased odds of administering 10 m walk test once during hospital stay, p < 0.05

(8) Increased odds of administering of 6 min walk test once during hospital stay, p < 0.05

Triage, treatment and transfer of patients with stroke in emergency department trial (the T³ Trial)

Middleton 2019, Cluster RCT, Intervention effectiveness

McInnes 2020, Qualitative process evaluation of factors influencing protocol uptake

Australia

Stroke patients admitted to ED. 26 EDs with stroke units

Intervention:

N = 13 EDs, 677 patients

Control:

N = 13 EDs, 920 patients

Internal: site clinical champion, as local opinion leader

External: nurse researcher

Remote: teleconference, phone and email follow up

In-person: 2 × 1 h workshops; 1 × 30 min education session; engagement every 6 wks, alternating between site visit and remote teleconference call

Facilitation dose: 2 × 1 h face-to-face workshops to identify barriers to implementation and develop action plans, 1 × 30 min education session. Engagement every 6 wks, alternating between site visit and teleconference call every 3 mo. Follow up of internal facilitator-initiated emails and phone calls

Intervention: assess stroke patients for thrombolysis eligibility; treat with thrombolysis; monitor fever, BGL and swallow function; transfer to stroke unit. Delivered via workshops, clinician education, reminders (lanyards, ED posters) and site engagement with visits, telephone and email

External facilitator: led workshops with clinicians; supported development of site-specific action plan; delivered education to clinicians and clinical champions; maintained site engagement

Internal facilitator: led local clinical change; delivered education for clinicians and new staff

Comparator: usual care

(1) NS difference in death or functional dependency status at 90 d. After adjustment for stroke onset to ED admission time, NS

(2) NS difference in functional dependency status at 90 d

(3) NS difference in health status at 90 d

(4) NS difference in patients screened for thrombolysis eligibility

(5) NS difference in receiving thrombolysis for ischemic stroke

(6) NS difference in 4-hrly temperature monitoring in ED

(7) NS difference in 6-hrly BGL monitoring in ED

(8) NS difference in swallow screening or assessment in 24 h of ED admission

(9) NS difference in transfer to stroke unit within 4 h of admission

(10) Internal facilitator needed to unite ED and stroke clinicians and lead implementation. Lack of authority of, or respect towards, internal facilitator as barrier to medical staff change

(11) Internal facilitator, as nurses, reported lack of medical staff in internal facilitator roles as barrier to engaging medical staff in implementation and practice change