1014 — Results of a stepped-wedge trial of expert-coaching to increase implementation of the transradial approach in cardiac catheterizations
Lead/Presenter: Christian Helfrich,
All Authors: Helfrich CD (Seattle Denver COIN), Beaver K (Seattle-Denver COIN) Petrova VV (Seattle-Denver COIN) Doll J (Seattle-Denver COIN) Taylor L (Seattle-Denver COIN) Naranjo DE (Salt Lake City VAMC) Neely E (Seattle-Denver COIN) Sulc C (Seattle-Denver COIN) Sayre G (Seattle-Denver COIN) Seto A (Long Beach VAMC) Valle J (Seattle-Denver COIN) Gootee JW (VA Southeast Louisiana Health Care) Waldo S (Seattle-Denver COIN) Pancholy SB (Geisinger Commonwealth School of Medicine) Swaminathan RV (Durham VAMC) Vidovich MI (Jesse Brown VAMC) Maynard C (Seattle-Denver COIN) Rao SV (Durham VAMC
We tested the effectiveness of team-based coaching to increase transradial access (TRA) in cardiac catheterizations. TRA is a safer, patient-preferred and cost-saving alternative to transfemoral access (TFA) for performing cardiac catheterizations. However, TRA is technically challenging, and there is a well-documented learning curve for most operators, in which it takes between 30 and 50 cases before they become proficient. The Promoting Action on Research Implementation in Health Services (PARIHS) framework proposes that new, complex evidence-based practices can be facilitated by a trusted, knowledge expert who understands the context and evidence as experienced by clinicians implementing practice changes. We developed a coaching intervention entailing hands-on support from expert cardiologists and catheterization-laboratory nurse managers to overcome the steep TRA learning curve.
We conducted a cluster-randomized, stepped-wedge trial. Participating laboratories were randomized to receive coaching in one of three cohorts, each four months apart. Coaching included in-person training at a coachâ€™s laboratory followed by a coaching visit approximately a month later. The primary outcome was procedure-level TRA odds. The primary data source was the Clinical Assessment Reporting and Tracking Program, which houses data on cases performed in the 82 Veterans Administration catheterization laboratories. The analysis used 2017-2020 data. Laboratories were eligible if they performed > = 100 catheterizations with < 50% TRA in the year prior to coaching. Procedure-level TRA versus TFA was modeled using mixed-effects logistic-regression. Models tested the effect of coaching on TRA odds at 5-8 months and 9-12 months post coaching. We conducted telephone interviews following the coaching intervention.
We randomized eight sites, but three withdrew due to staff turnover; we subsequently added two sites, thus seven sites received coaching. We performed intent-to-treat (ITT) analysis on the eight randomized sites, and as-treated (AT) analysis on the seven that received coaching. Between 2011 and 2018 across all VA catheterization laboratories, TRA increased for diagnostic catheterizations (17.6% to 60.5%; p < 0.01) and interventions (14.5% to 51.8%; p < 0.01). At 5-8 months, in ITT, coaching was not significantly associated with TRA for diagnostic catheterizations (OR = 1.30, 0.75-2.22) or interventions (OR = 0.94, 0.46-1.93); in AT at 5-8 months, coaching was not significantly associated with TRA for diagnostic catheterizations (OR = 1.32, 0.99-1.77) or interventions (OR = 1.03, 0.68-1.55). At 9-12 months, coaching was not associated with any TRA implementation. Post-coaching interviews reported a range of lessons from coaching, such as proper set-up; expanding the types of cases performed (e.g., right-heart catheterization); and using ultrasound to make it easier and faster to gain arterial access. The participants also reported that it was helpful having the coaches visit the participating sites to reinforce what the catheterization teams were doing and providing encouragement.
TRA increased significantly, but there was no association between the designed coaching intervention and increased TRA adoption. Participants reported receiving the types of supportive problem solving we expected would mitigate the learning curve, and coaching might be expected to help sites not already making significant progress implementing TRA.
These findings demonstrate the importance of testing implementation strategies, like expert facilitation, using rigorous designs that control for secular trends. Methodologically, these findings contribute to the literature on stepped-wedge trials in implementation science and highlight important challenges related to timing of interventions.