Abstract
The objective of this study was to use the critical decision method (CDM) to map the decision-making process of parent caregivers (PCGs) responding to patient safety events at home and to identify the work system factors that influence this process. PCGs (N = 13) were asked to share about a specific patient safety event that had occurred while caring for the CMC at home. PCGs were asked to describe the event in chronological order, after which their decision making at each timepoint was probed. The final process map included nine steps, the success of which were shaped by work system factors. This map of PCG decision making can be used to train policymakers, durable medical equipment companies, and clinical providers on the work that PCGs perform at home, and to evaluate whether a family has what they need to maintain home-based safety.
Keywords
Introduction
Children with medical complexity (CMC) are pediatric patients with multiple, disabling chronic conditions, affecting multiorgan systems and requiring access to specialized care (Children’s Hospital Association, 2013). Family caregivers, most often parents, are tasked with implementing complex treatment regimens at home, including using advanced medical devices, managing tracheostomies, and administering high-risk medication regimens (McCann et al., 2012). Although home is the preferred long-term care setting for CMC, aspects of the home, including those related to supplies and equipment, availability of home nurses, the training received by parent caregivers in hospital, and caregiver fatigue, increase risk for safety events (Elias et al., 2012). Safety events at home have been linked to hospitalizations, emergency department visits, and unplanned medical procedures for CMC (Foster et al., 2020). Despite the centrality of home-based care to the wellbeing of CMC, little is known about patient safety in the home for CMC, with a specific gap in knowledge regarding how parent caregivers (PCGs) respond to patient safety events. The objective of this study was to use the critical decision method (CDM) to map the decision-making process of PCGs responding to patient safety events at home, and to identify the work system factors that influence this process.
Background
CDM is an approach to cognitive tasks analysis that has been used to understand how those with expertise in a domain make high-stakes and time-pressured decisions, including in healthcare delivery systems (Gazarian, 2013; Klein et al., 1989). In the context of home healthcare, PCGs are the resident experts, as they have typically managed the CMC’s illnesses since birth and have become adept at identifying problems, leveraging available resources, and facilitating communication across the care team (Hirt et al., 2023). To design interventions for PCGs, it is necessary to first understand the decision-making process they engage in the face of patient safety events. Further, it is crucial to leverage a human factors approach to understand the systemic factors that support or hinder decision making.
Approach
We identified PCGs and their children through electronic patient registries, non-profit organizations, and word of mouth communication. PCGs met inclusion criteria if they cared for a child between the ages of 0 to 17 with chronic conditions affecting multiorgan systems resulting in dependence on implanted medical devices, such as enteral tubes, tracheostomy tubes, or central lines, for medication administration. There were no additional exclusion factors.
CDM interviews were conducted over Zoom and lasted for 60 to 75 min. PCGs were asked to focus the interview on a specific patient safety event that had occurred while caring for the CMC at home. Example events included a G-tube, GJ-tube, or tracheostomy being dislodged; medication overdoses or missed medications; and the dysfunction of life-sustaining equipment, such as ventilators and feeding pumps. PCGs were first asked to describe the event in chronological order while a notetaker created a timeline of the safety event. After confirming that the timeline was correct, PCGs were asked how they made decisions at each point, including the cues or information to which they attended; the bases for decisions made, including courses of action that were rejected; and, if they were dissatisfied with their response, what hypothetical training or experience could have helped (Klein et al., 1989).
Two coders (AJ and SS or AJ and MF) read each transcript and extracted all information relevant to decision making to a decision analysis table (e.g., Table 1), which included a row for each decision point and a column to document the cues, background knowledge, situation assessments, rationale, goals, and emotions associated with each decision point (Wong, 2003). Next, an inductive qualitative analysis was used to identify the common decision points across participants and the systemic influences that were pertinent at each decision point (Russ-Jara et al., 2023). The final process map was confirmed through discussion until team-wide consensus was reached. In addition to the coders, the research team included a human factors engineer (NW), a pediatrician (CF), and an expert in qualitative analysis (CH). Parents of CMC who were not study participants reviewed the process map and provided feedback on its pertinence to their lived experience.
Example Excerpt from Decision Analysis Table.
Outcome
PCGs’ (N = 13) average age was 37.9 years old (SD = 4.8), while children’s average age was 5.6 years (SD = 3.0). All 13 children received some or all of their medication by tube (gastrostomy, gastrostomy jejunostomy, or nasogastric) and were fed in all or part by tube. The most common conditions were seizure disorders (N = 8), cerebral palsy (N = 8), and brain injury (N = 5). The final process map depicted PCGs’ decision-making process during patient safety events (Figure 1). The process involved nine possible steps. In the figure, the sociotechnical system factors affecting each step are written in italics. Hyphenated lines indicate that a step was not used in all cases. Four of the steps—assessing the problem, identifying possible solutions, assessing available time, and attempting potential solutions—formed a cycle that could be repeated multiple times. Risk or harm mitigation could begin at any point during this four-step cycle. The cycle resolved either by finding a satisfactory solution at home, seeking acute care, or doing both.

The decision-making process of PCGs of CMC responding to patient safety events at home.
Conclusion
In this study, we explored how PCGs make time-sensitive decisions during patient safety events and the sociotechnical system factors that influence each step of the process. Our process aligns with the decision ladder framework (Rasmussen, 1986), which has been used to model time-sensitive decision making among experts in military, industrial, and hospital settings. Our results suggest that PCGs follow a similar process but face home-specific systemic factors such as the proper functioning of equipment, the availability of supplies, and the effectiveness of their training prior to discharge. Each systemic factor identified in the process map represents an opportunity for intervention. For example, PCG experience and training was found to influence most steps in responding to patient safety events. This underscores the need for robust training for PCGs prior to CMC discharge from the hospital. This map of PCG decision making can also be used to train policymakers, durable medical equipment companies, and clinical providers on the work that PCGs perform at home, and to evaluate whether a family has what they need to maintain home-based safety.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This project was sponsored by the Agency for Healthcare Research and Quality (AHRQ) Grant R18 HS29638-01. The views expressed in this paper reflect those of the authors and not necessarily those of AHRQ.
