This is a stepped-wedge cluster randomized effectiveness-implementation hybrid study aimed at determining the effect of dissemination of a QI bundle on the time to treatment of SE among hospitalized, non-critically ill children. The primary study endpoint is to decrease the time from the SE diagnosis to treatment with the first dose of a benzodiazepine (BZD) as measured during hospitalization, which will decrease chances of morbidity and mortality.

The overall study design is a stepped-wedge cluster randomized trial. A stepped-wedge is a unidirectional crossover design in which clusters switch treatments at different time points, enabling statistically rigorous assessment of interventions while reducing ethical and resource limitations for quality improvement studies.[1] Seven centers will be randomly assigned to implement the QI bundle at staggered 1-month intervals after a baseline period. Only the timing of the dissemination visit will be randomized; all sites will receive the same materials and perform the same activities (e.g. focus group, process map development, simulations).

This study is an effectiveness-implementation hybrid study. The effectiveness-implementation hybrid design is ideal for assessing both clinical interventions and implementation.17 Importantly, our interventions have strong face-validity, are evidence-based, low-risk and low-cost. For instance, the price to change BZD formulations is nominal (~$1 per dose), and other QI bundle interventions are primarily focused on frontline staff process changes. While testing the effect of the QI bundle on time to BZD treatment, we will utilize this framework and mixed methods analysis to measure implementation and identify barriers and facilitators. Thus, this proposal is of high-value, as it will provide randomized multicenter efficacy data while providing understanding for broad implementation following study end.

For the effectiveness component, we will utilize a stepped-wedge cluster randomized design. Based on our preliminary data, the intraclass correlation coefficient is estimated around 0.5, an overall sample size of 60 episodes will be adequate to achieve powers greater than 90%. A potential pitfall of the stepped wedge design is the potential confounding effects of temporal trends. We will mitigate this by tracking data at the primary site, which will be in the sustain phase throughout the study entirety. We reduce temporal effects of site enrollment (e.g., staffing changes, temporal trends in hospital admissions) by enrolling sites at different times throughout the calendar year.

The implementation component was designed utilizing the Practical, Robust Implementation and Sustainability Model (PRISM).[2] PRISM provides a scientifically rigorous and structured approach to implementation strategy development through domains focusing on (1) program, (2) external environment, (3) implementation and sustainability infrastructure and (4) recipients.[2] These elements are addressed as follows:

Program. Organizational readiness across the study sites will be critical for success. Our proposal is based on evidence derived from a successful single-center study. The QI bundle is low-cost and all interventions are currently FDA-approved. Our innovative de-implementation of time-consuming, low-value workflows (e.g. IV medication administration) will decrease care complexity in the initial stages of SE treatment. The QI bundle components are trialable, adaptable and reversible. Treatment results are immediately observable by stakeholders through individual outcomes (SE cessation) and shared measure and feedback data reports. We highlight improved outcomes and safety as organizational, caregiver, and patient priorities to achieve broad buy-in.

External Environment. Regulatory and professional organization priorities support our area of study and primary efficacy outcomes. Rapid treatment of SE has been identified as a quality measure by the AAN11, and guidelines for such care have been published by the AES.[3] Additionally, our study proposal further aligns with NAEC, which mandates a focus on rapid treatment of SE through pathway requirements across 260 hospitals. Data from our proposal will serve as a framework for accomplishing the goal of rapid SE treatment, which is inconsistently met at present.[4]

Implementation and sustainability infrastructure. Our infrastructure will utilize proven features associated with successful implementation projects.[5,6] Co-investigators experienced in working within pSERG will provide a bridge to the local QI and clinical teams, engaging stakeholders at all 3 organizational levels (frontline staff, mid-level management and senior administration). Measure and feedback will be emphasized through control chart data and implementation reports. Furthermore, the adaptable protocol allows for site-specific implementation strategies for the QI bundle as well as iterative PDSA development in the Sustain and Independent phases in order to address local drivers.

Recipients. Positive organizational characteristics are supported by LOS from senior hospital administrators and nurse managers. Furthermore, the co-investigators have previously collaborated with pSERG from their respective centers. Each site has access to data through Export, Transform, Load (ETL) data queries and EHR. The diverse demographics of patients is aided through intentional site selection and inclusion of nearly all ages of children. Importantly, our proposed interventions of performing basic seizure first aid and using non-IV forms of BZD aligns with those of patients and families in the ambulatory setting.[7]

Inclusion Criteria:

• SE episode occurs in a male or female child aged between > 30 days to < 19 years

• Seizures meeting AT LEAST ONE of the following criteria:

  1. continuous clinically apparent seizure lasting greater than 5 minutes
  2. continuous clinically apparent seizure of any duration receiving BZD
  3. repeated seizures without return to neurological baseline within 5 minutes

Exclusion Criteria:

• SE episode occurs in a child with infantile spasms
• SE episode occurs in a child with electrographic-only seizures without clinical signs other than encephalopathy

Participants will be masked as to whether the implementation bundle has been dissemination or implemented within the site they receive treatment.