What does the facilitation implementation strategy look like in the real world? Evidence from the ADEPT study Shawna N. Smith, Ph.D. University of Michigan 9 th Annual Conference on the Science of Dissemination & Implementation in Health 14 December 2016
Overview & orientation • 45 community-based sites working with an external facilitator (EF) to implement a mental health collaborative care model (CCM) for patients with mood disorders • Using unique data on EF tasks and interactions with sites, we describe what facilitation looks like in the “real world”
Barriers to CCM adoption in community-based practices • Consumers/Famil Consumers/Families ies – Lack of car Lack of care coor e coordination dination – Most privately insur Most privately insured seen in small practices ed seen in small practices • Pr Providers oviders – <30% of MSWs have access to EBPs <30% of MSWs have access to EBPs (Insel, 2009) – Inconsistent EMR use Inconsistent EMR use • Systems Systems – Primarily FFS, few bundled payment models Primarily FFS, few bundled payment models – ACOs--mental health car ACOs--mental health care? e? (O’Donnell, 2013)
Facilitation as a tool for addressing barriers Multi-faceted implementation strategy that focuses on “helping rather than telling”; PARiHS framework Facilitators aid sites by: • Assessing barriers to uptake • Guiding problem-solving in context of need for improvement • Ensuring communication and priority alignment with site leadership Rycroft-Malone et al., 2002, 2004; Kirchner et al., 2014
Flavors of Facilitation External facilitator (EF): • Location: Off-site, research team member • Topical focus: Benchmarking, coaching • Role: Confidante, outside observer Internal facilitator (IF): • Location: On-site, direct report to leadership • Topical focus: Leveraging, rapport-building, internal recognition, sustainability • Role: Role: Inside expert, champion Kilbourne AM et al. 2013; 2014; Goodrich et al. 2012
Research questions Using data from ADEPT: • [How] do sites make use of EF? • What does content of EF look like? • Does EF content differ by presence of IF?
The ADEPT Study Study Start Phase 2 Follow Up Month 18 and 24 Month 6 Month 12 Run-In Phase Assessments Assessment Assessment All sites offered REP to implement EBP; Con+nue REP Con+nue REP Pa+ents start EBP by Responders (A) k=5 sites Add External Month 3 Facilita+on REP+EF Con+nue REP+EF (B) Cont REP+EF k=7 sites k=23 sites Non- R Non- responders Add IF Con+nue REP+EF/IF Responders REP (C) k=11 sites (<50% of k=79 R pa+ents Responders sites Con+nue REP Add Internal & Con+nue REP receiving >=3 (D) External k=7 sites LG sessions) Facilita+on k=45 sites REP+EF/IF Non- Con+nue REP+EF/IF Cont REP+EF/IF k=22 sites (E) responders k=15 sites Con+nue REP Con+nue REP Con+nue REP Responders (F) k=20 sites Kilbourne et al, 2014; Funding: NIMH R01 MH99899
Life Goals CCM Decision Decision Support Support Self-management Self-management Access/Continuity Access/Continuity Life Goals Sessions: Life Goals Sessions: Pract Practice Guidel ice Guidelines: ines: CVD Risk, Symptoms CVD Risk, Symptoms Best practice txt, Best practice txt, Healthy Behaviors Healthy Behaviors metabolic syndr metabolic syndrome ome for mental health for mental health Provider Engagement Pr ovider Engagement conditions conditions Car Care management: e management: Registry tracking (Symptoms, QOL, functioning ) Registry tracking (Symptoms, QOL, functioning ) General Medical Provider Liaison General Medical Pr ovider Liaison Kilbour Kilbourne et al. Psych Serv 2008; 2013 ne et al. Psych Serv 2008; 2013
The ADEPT Study 45 ADEPT sites received EF for at least 6 months: 45 ADEPT sites • 5 sites received EF only for 6 months 12 • 7 sites received EF only for 12 months EF only sites 11 • 11 sites received EF only EF only for 6 months delayed followed by EF+IF for 6 months EF+IF sites • 7 sites received EF+IF for 6 months 22 EF+IF only • 15 sites received EF+IF for 12 month sites
Data & Methods Data: Study EF logged all tasks, categorizing: • Mode (email, phone, in person) • Personnel interacted with • Duration (5 minute intervals) • Primary content focus Methods: • Describe mode, duration and content of EF tasks • Examine differences in EF tasks across sites receiving EF and EF+IF sites, with focus on strategic content
EF Tasks Initiation & benchmarking Assess context, identify needs & barriers Coaching Develop rapport, work with providers to address barriers Leveraging Identify local leadership priorities, identify additional site champions to promote LG Implementation plan development Create measurable goals, specify tasks & timeline Link to outside resources Connect with similar sites, REP technical assistance Education Requirements & expectations of personnel Ongoing marketing Develop business plan, focus on sustainability Reinforcement Positive recognition and support Consultation Discuss issues with facilitation experts Background Site research, often non-interactive Kirchner et al, 2013
EF Tasks: Strategic Initiation & benchmarking Assess context, identify needs & barriers Coaching Develop rapport, work with providers to address barriers Leveraging Identify local leadership priorities, identify additional site champions to promote LG Implementation plan development Create measurable goals, specify tasks & timeline Link to outside resources Connect with similar sites, REP technical assistance Education Requirements & expectations of personnel Ongoing marketing Develop business plan, focus on sustainability Reinforcement Positive recognition and support Consultation Discuss issues with facilitation experts Background Site research, often non-interactive
Results 1,037 tasks were logged by EF for 45 sites between January 2015 and July 2016 Median EF time logged per site: 36 minutes/month [IQR: 25-43] Mode: 664 (64%) email (mean duration: 6 minutes) 353 (34%) phone (mean duration: 23 minutes)
Results: EF tasks, by month 250 120 Number of Tasks Mean Time Per Site 100 200 80 Mean Time Per Site Mean Time Per Site Number of Tasks asks 150 Number of T 60 100 40 50 20 0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 Mont Month
Results: EF vs. EF+IF sites Who did the EF interact with? 80 EF (N=564) EF+IF (N=473) 70 60 cent of interactive tasks Percent of interactive tasks 50 40 30 Per 20 10 0 Admin Supervisors Providers IFs Other
Results: EF vs. EF+IF sites How long did the EF spend per site? • Average EF sites: – Overall mean: 260 minutes – Controlling for time: 12.6 minutes/day • Average EF+IF sites: – Overall mean: 429 minutes – Controlling for time: 21 minutes/day *Difference not significant at p<0.05 after controlling for time
Results: EF vs. EF+IF EF (N=564) EF+IF (N=473) Initiation & benchmarking 42% -- 17% -- Coaching 0.6% 1.1% 0.7% 0.9% Leveraging 0.2% 0.4% 3% 4%* Implementation plan development 3% 6% 6% 8% Link to outside resources 0.4% 0.8% 1% 1.5% Education 25% 51%* 35% 44% Ongoing marketing 0.6% 1% 3% 4%* Reinforcement 11% 23% 28% 35%* Consultation 8% 17%* 2% 2% Background 10% -- 4% -- * Difference significant at p<0.05 or better
Results: EF vs. EF+IF EF (N=564) EF+IF (N=473) Initiation & benchmarking 65 minutes 51 minutes Coaching 1 4 Leveraging 2 13* Implementation plan development 13 43* Link to outside resources 1 2 Education 70 166* Ongoing marketing <1 9* Reinforcement 28 59 Consultation 26* 7 Background 73 37 EF sites spent an average of 16 minutes on strategic initiatives, EF+IF sites an average of 64 minutes . * Difference significant at p<0.05 or better
Conclusions • Use of EF and EF+IF strategies in community-based implementation studies is feasible, scalable feasible, scalable – EF logged 25-43 minutes/month with each site • EF tasks at sites with & without IF were largely similar , not significantly different in time commitment • Sites with EF+IF did spend more time on strategic tasks than EF sites, which may lead to longer-term sustainability of LG adoption
Next Steps • Differences in study outcomes by EF & EF+IF – ADEPT primary aim: EF vs. EF+IF as best ‘step-up’ option for sites non-responsive to REP? – Delayed effect of adding IF in Phase 3 rather than starting with EF+IF in Phase 2? • Mechanisms of effectiveness – Relationship between EF interaction content/time and effectiveness? – Combine EF and IF activity logs to data to better elucidate mechanisms • Cost-effectiveness of EF vs. EF+IF – Is IF worth the additional cost compared to EF alone?
Recommend
More recommend
