Examining Caregiver Decision-Making using Human Factors Analysis and Classification System (HFACS), AcciMap, and Risk Evolution, Detection, Evaluation, and Control of Accidents (REDECA) Frameworks

Pediatric caregivers are a critical part of chronic pediatric care. They are the first to observe, monitor, and act as their child's medical condition progresses from benign to potentially life-threatening. Caregivers are thus in the unenviable position of being first responders without much formal training. Training, if any, is often truncated and limited to operating "therapy" devices like home mechanical ventilators (HMVs) or infusion pumps for home total parenteral nutrition (TPN). Little consideration is given to potential health risks, decision making, or even day to day activities like bathing. This paper examines caregiver errors using three frameworks derived from Reason's Swiss Cheese Model: HFACS, AcciMap, and REDECA. The Swiss Cheese model considers error as the effect of multiple layers of defense, organizational to individual, inadequately preventing a near miss event from escalating to an adverse event. HFACS examines errors using four primary factors - 1) Unsafe Acts, 2) Preconditions for Unsafe Acts, 3) Unsafe Supervision, and 4) Organizational Influences. AcciMap considers errors a result of poor interactions across organizational boundaries. REDECA examines how automation risks escalate; and how AI can help detect, prevent, and/or mitigate these risks. Our goal is to study caregiver error using these three frameworks to develop methods to identify sources of caregiver errors.