Causal analysis is a core function of safety programs. Although established protocols exist for conducting root cause analysis for serious safety events, there is limited guidance for apparent cause analysis (ACA) in health care. At our institution, through a novel facilitated ACA approach, we aim to improve safety culture and provide a clear approach to address precursor safety events and near-miss safety events. We define facilitated ACA as limited investigation (scope and duration) of a safety event that resulted in little to no harm. These investigations require fewer resources and focus on preventive strategies. Our facilitated ACA model, with an operational algorithm and structured process, was developed and implemented at our tertiary-care, freestanding, urban pediatric hospital in 2018. Sixty-four ACAs were completed, and 83% were identified with the algorithm. Process measures, including time from event reporting to ACA launch (median 3 days; interquartile range 2–6 days), are tracked. Patient safety consultants averaged 5 hours to complete a facilitated ACA. A median of 3 disciplines or departments participated in each facilitated ACA. Through an iterative process, we implemented a structured process for facilitated ACA, and the model’s strength includes (1) right event, (2) right team, (3) right analysis, and (4) right action plans. This novel facilitated ACA model may support organizational cause analysis and improve safety culture with higher-reliability processes.
Clinicians work in complex environments with a high risk of unintended harm to patients. The issue of patient harm has received national attention since the Institute of Medicine (now the National Academy of Medicine) published the report “To Err Is Human: Building a Safer Health System.”1 Despite renewed focus on pediatric patient safety, harm to patients has not shown a significant decline,2 and errors continue to affect as many as one-third of hospitalized children.3,4 To organize a strategy to combat unintended harm to children, the American Academy of Pediatrics policy statement on principles of pediatric patient safety focuses on 3 key domains to minimize pediatric medical errors: describing the state of pediatric patient safety, understanding the science behind the culture of safety, and implementing strategies to improve patient safety.5 In addition, recently published priorities for pediatric patient safety emphasize the importance of developing high-reliability solutions to improve safety culture.6 The novel facilitated apparent cause analysis (ACA) approach presented in this article is an example of how to operationalize the American Academy of Pediatrics domains by using high-reliability solutions and thus advance our culture of safety and improve patient care.
Many high-reliability industries have developed strategies to minimize human error. Inspired by their work, our patient safety team used models from these industries to better understand the science of safety culture. Causal analysis (for example, root cause analysis [RCA]) is a foundational tool in patient safety programs within health care.7–9 Cause analysis methodologies have been effective in industries outside of health care, such as construction and aviation,10,11 in which multidisciplinary teams are tasked with analyzing harmful events and constructing action plans to prevent reoccurrence of errors. Analysis of preventable medical errors identifies a root cause and aims to build a safer, more reliable system to reduce human error. Many hospitals effectively use incident reporting along with event investigations to actively learn from preventable medical errors12 ; these event reports provide an opportunity for organizations to move toward high-reliability solutions and drive improvements in safety culture.
RCA has been discussed extensively in the literature; however, to the best of our knowledge, there is no standardized definition of an ACA. At Children’s National Hospital, in partnership with patient safety organizations and thought leaders such as Healthcare Performance Improvement, LLC13 (now acquired by Press Ganey, Cleveland, OH), our institution adopted the following definition of ACA: limited investigation (scope and duration) of an event that resulted in little to no harm (performed instead of RCA). Table 1 shows a comparison of ACA and RCA at our hospital. In contrast to RCAs, ACAs require fewer resources and are focused on preventive strategies.14 In the safety literature, there is a well-recognized classification system for types of safety events,13 specifically serious safety events, precursor safety events, and near-miss safety events. In this classification system, serious safety events usually undergo an RCA, and precursor safety events and near-miss safety events may be considered for evaluation by ACA.
Although established protocols exist for conducting a resource-intensive RCA,12 there is limited guidance in the use and facilitation of ACAs. Therefore, it is important to build on the ACA model as an additional tool to address pediatric safety improvement priorities. An informal retrospective review of previous ACAs at our hospital identified many events that were attributed to human error without further probing into the system in which the error was made. This patient safety initiative aimed to revise our ACA process to actualize the new paradigm of safety. This process is designed to provide a standardized and reliable approach to cause analysis that moves beyond human errors and creates systemic solutions.
Approach for Facilitated ACA
Hospital Setting
Children’s National Hospital is a freestanding, tertiary-care children’s hospital in an urban setting. Children’s National Hospital is a 323-bed hospital with pediatric and cardiac units and NICUs, caring for ∼15 000 inpatients and ∼120 000 emergency department encounters per year. Approximately 10 000 safety event reports per year, including inpatient and ambulatory setting reports, are entered by using the safety event reporting system and reviewed by the hospital’s patient safety team. The intent in this article is to focus on the ACA process and how this was modified at Children’s National Hospital into a facilitated method for cause analysis. The initial ACA approach was introduced in July 2016 at our institution. Through iterative improvements, the first structured and facilitated ACA was launched in May 2018.
Brief Summary of the Initial ACA Method
To better understand our current facilitated ACA model, we share a brief summary of our approach to the initial ACA method. In the past, the hospital-based safety team reviewed each safety event report and determined which events were appropriate for an ACA. The team emailed a document to the unit and/or care team leader, which included questions regarding the background and circumstances surrounding the event, units and/or care teams involved, as well as a request for review and an action plan. The documents were returned to the Patient Safety Office within 2 weeks of initiation and filed for record keeping with limited discussion and follow-up on action plans. Although local leaders were frequently able to identify highly reliable interventions to prevent reoccurrence of these safety events, the degree to which these interventions were implemented was variable. Because the end product of the initial ACA focused on actions (such as education and training), the outcomes were often reactionary and without a clear plan for developing higher-reliability interventions.14 Therefore, it was deemed necessary to revise the ACA model to further improve harm reduction in our institution.
Facilitated ACA Model
Redefine the Goal of the ACA Process
To identify highly reliable actions to improve system safety, it is of paramount importance to first construct and validate the cause of an event. However, our review of the initial ACA process revealed that in many events, human error was determined to be the sole cause of harm. Because the process ended with this determination, there was a limited number of system-wide analyses and, therefore, no opportunity for system-level improvement. Classification of these initial ACAs as being due to human errors conflicts with the axiom in modern safety science that human errors are not the root cause of the event but that the errors were made because of gaps in the process or system.12,15–17 The guiding principle of the facilitated ACA model focuses on the concept that human error should not be the final result of an ACA evaluation but rather should lead to system improvement that will ultimately prevent this type of error in the future. Table 2 summarizes the features of the facilitated ACA model.
Reviewing Safety Event Reports and Creating an Algorithm
The patient safety team reviews all submitted event reports and triages the report to either (1) the local team or unit leadership or (2) the chief quality and safety officer and/or risk management team. Local teams have the primary responsibility for follow-up on all reported safety events; therefore, the patient safety team does not manage the majority of events beyond the initial triage. However, incidents deemed concerning for a safety event are triaged as an ACA or RCA. The hospital’s RCA process is modeled after the RCA 2 framework developed by the National Patient Safety Foundation.12 RCAs are primarily conducted on events that are classified as serious safety events by the hospital’s Safety Event Classing Committee by using Safetyn Event Classification methodology.13
To develop consistency, the team created an algorithm to prioritize events that are triaged to a facilitated ACA (Fig 1). The algorithm is intended to optimize resource use, maximize learning from the incident report, and reduce duplicative review of events, which could result in redundant or conflicting follow-up actions.18 The algorithm also provides recommendations for the best course of action for a given event. Using the shared mental model established with the algorithm, a subset of the incident reports is evaluated by using the facilitated ACA process. Three example cases are shared in Table 3.
The first screening element in Fig 1 determines if the safety event report relates to an actual event that occurred or is a description of a risk perceived by the report writer. Perceived risks cannot be described by using a sequence of events and are better analyzed by using a prospective tool, such as Failure Modes and Effects Analysis. Events with high-level harm and high complexity, such as those involving multiple departments and disciplines, are evaluated for potential RCAs. Events that may be the result of a known treatment complication are screened by using the Known Complication Test13 to determine if risk was minimized to its fullest extent and to confirm that proper treatment was administered. Events that involve technical equipment or engineering concerns are referred to the appropriate technical departments. Next, we apply the just culture methodology.19 At our hospital, just culture means that staff are accountable for their actions but are not blamed for system failures beyond their control; this policy was adapted from a published decision tool.20 Certain events, such as central line–associated bloodstream infections, fall, and ventilator-associated pneumonias, are already part of established quality-improvement priorities, with operational definitions and metrics being evaluated by standing committees. These events are not evaluated through the facilitated ACA process because established groups are designated to review those events on a scheduled basis. Use of the algorithm presented in Fig 1 narrows the incident reports to events that are suitable for a facilitated ACA.
Pre-ACA Meeting
The hospital-based safety team reaches out to the local leader before launching the ACA to obtain background information surrounding the event. Once this discussion occurs, an ACA is proposed and launched.
Face-to-Face ACA Meeting
A mutual decision is made between the patient safety team and the unit and/or care team leader regarding disciplines that are essential for the ACA discussion. The meeting is scheduled for a 90-minute time block when all parties can be present. We adhere to 90 minutes, acknowledging that the ACA is improperly scoped if this time allotment is inadequate. Effort is made to limit prework for the members who will be present at the ACA meeting as a compromise for their attendance and attention.
To launch the meeting, the patient safety team leader explains the ACA structure and process with focus on the system and precursors to human error. The first 30 to 45 minutes of the meeting are spent describing the sequence of events by the unit and/or care team leader(s). The patient safety team leader restates the sequence of events as a readback to ensure mutual understanding and to create a shared mental model. Attendees enumerate the problems that led to the safety event. The patient safety team leads the dialogue with respect to the identified problems and assists in the development of action items.
As a general requirement, all proposed action items must have ownership within the group. If the desired final outcome is outside the independent operational control of individuals in the room, the action items are scoped as the first practical steps toward a larger improvement. After the 90-minute meeting, the patient safety team creates a formal document that summarizes the meeting and accountability for the action plans. There are no deadlines for the items; however, there is a 30-day follow-up between the ACA participants and the patient safety team to review the implementation time line, discuss the progression of improvements, and identify barriers.
Post-ACA Meeting
The first follow-up component is a conference call with ACA meeting attendees 30 days after the initial meeting. The patient safety team leader facilitates the call to monitor progress with the action items and plan additional interventions. If barriers arise in implementing actions, the call provides an opportunity to discuss issues and escalate challenges to leadership. After the ACA actions have been in progress for 30 days, the ACA is formally presented to our Patient Safety Committee. In addition, the unit and/or care team leaders present ACAs with broad generalizability at hospital-wide quarterly event reviews, sharing identified safety issues and improvements undertaken in their own departments.
Patient Safety Team: Structure, Training, and Time Commitment
The patient safety team, including a director of patient safety and 2 patient safety consultants, reports to an executive director and the chief quality and safety officer. Patient safety consultants are the primary facilitators of the ACA process and receive training in meeting facilitation, team dynamics, change management, and conflict resolution before independently facilitating ACAs. Specialized training for consultants includes human factors, high-reliability action planning, human performance, safety science, and cognitive psychology. Subject matter experts are used for clinical cases that may be outside the consultants’ expertise. Before conducting the ACA process independently, safety consultants are trained by their peers and the director of patient safety. They perform ACAs with supervision until they are able to effectively facilitate a multidisciplinary team in developing highly reliable and attainable action items.
Patient safety consultants average ∼5 hours of time per facilitated ACA. In the facilitated model, 15 minutes are spent reading the event report and checking for follow-up documentation, 30 minutes are spent contacting and discussing the ACA with local leaders, 60 minutes are spent scheduling the meeting and booking an accessible space, 90 minutes are spent facilitating the ACA meeting, and 45 minutes are spent arranging and conducting follow-up and processing documentation.
Iterations are focused on improving logistics and supporting meaningful dialogue, which would ultimately allow for sustainable system changes. The team has created a manageable documentation form, expanded the meeting format to 90 minutes, and transferred meeting scheduling to the patient safety team to decrease the burden on local management and increase attendance.
Tracking Measures
To evaluate the facilitated ACA model, the patient safety team has created measures focused on process, content, and spread. Process measures include the rate of ACA per month and time to ACA launch. Hospital leadership has established these process measures to track ACAs and assess the time commitment and resource use for the patient safety team. Content measures include event categories and harm classification as well as the number of participants and attendees at ACA meetings. Spread measures include how findings were disseminated to the Children’s National Hospital community.
Tracking Experience With Facilitated ACA
The patient safety team tracks measures related to the facilitated ACA process, content, and spread. Median time from event reporting to ACA launch is 3 days (interquartile range [IQR] 2–6), from launch to meeting is 11 days (IQR 8.75–17), and from event to meeting is 15 days (IQR 10–22.25). As is expected on the basis of the algorithm, the majority of facilitated ACA cases include events with minimal to no harm. Events are classified by the event reporting system, and the leading content areas for facilitated ACAs are medication and fluids, provision of care (for example, communication issues or delays in care), and safety and/or security. Additional measures of the ACA process included team engagement and broadened discipline involvement. Overall, there were 318 participants in the new model ACAs; this represents 190 unique individuals and an average of 5.5 attendees (median = 5) per facilitated ACA. These individuals represented >60 different departments and disciplines, and a median of 3 disciplines and/or departments was involved in each facilitated ACA. A marker of spread evaluated the number of attendees present in the various forums where ACAs were presented by the unit-based leadership. There were 27.5 attendees (range 22–36) in the Patient Safety Committee when 2 to 5 ACAs were presented and discussed monthly. Quarterly hospital-wide ACA reviews were conducted in the auditorium and facilitated by the chief quality and safety officer; there were ∼100 attendees at each of these sessions.
Core Elements of the Facilitated ACA: Right Event, Right Team, Right Analysis, and Right Action
Our vision to move beyond human error as the sole cause of a safety event and focus on system changes that support high-reliability interventions drove our efforts to develop the facilitated ACA model. In our experience, previous processes were reactive to safety events, whereas the facilitated model takes a more proactive approach; the new approach employs a thorough and complete event review, which results in high-reliability actions. Through our iterative learnings, we established a framework of attributes for an effective ACA focused on (1) identifying the right event, (2) assembling the right team, (3) conducting the right analysis, and (4) focusing on the right action plan. Table 4 shares examples that highlight this framework.
The Right Event
The decision to launch an ACA is based on the content of the safety event report and discussion with local leaders. On the basis of our team’s experience, the right event for a facilitated ACA will have a definite sequence of events, involve a limited number of functional units, and benefit from multidisciplinary engagement. The ACA focuses on identifying causes of system failures in a time-limited setting. To accomplish this task, the facts of the event must be known and verifiable so we can work to identify system gaps. This definition allows for a robust analysis of events such as medication administration errors and procedural delays. However, it leads to difficulty in examining events such as hospital-acquired infections because a definitive causal chain may be impossible to establish. Effective ACAs examine events in which an ideal set of process steps can be compared with what actually occurred, thus allowing for easy identification of system gaps.
Our team sought to identify events for a facilitated ACA that would both require and benefit from multidisciplinary engagement. By targeting events involving multiple disciplines, the goal was to initiate an in-depth analysis with the involved local unit and/or care team leaders. For example, after initiation of the facilitated approach, the team observed a shift in the general event types undergoing ACA away from laboratory events and toward medication safety events. This trend reflects the desire to target events involving multiple units and/or care teams. Previously, only one unit or care team was identified as the lead; this may have limited the development of system-based, higher-reliability solutions. At our institution, laboratory events are routinely tracked, trended, and individually analyzed by existing laboratory quality assurance processes. Errors in processing send-out laboratories are more easily addressed within the laboratory unit. However, medication safety events often involve multiple disciplines, such as nursing, pharmacy, and providers; in turn, they often benefit from a multidisciplinary approach. With multiple departments now completing the ACA form, the patient safety team has a broader perspective on these events. Broader perspective increases the potential effectiveness of resultant action plans.
The Right Team
In an effective ACA, the full complement of disciplines with appropriate knowledge and system oversight is present. The assembly of an ACA team is based on the initial information in the safety event report. The patient safety team and local leaders must use this information to anticipate the general direction of the required analysis and therefore the most appropriate and relevant meeting attendees. Although local leaders are expected to investigate the event before the ACA team meeting, additional facts and perspectives are frequently raised during the ACA meeting itself. Given the necessity of completing a thoughtful analysis in a limited time frame, assembling an appropriately broad set of stakeholders to participate in an ACA must be balanced with the need to maintain a manageable team size. However, assembling the right team can be challenging because facts may arise during the ACA meeting that require the participation of a discipline that was not invited initially. This may delay the process and require an additional meeting to be scheduled.
In bringing together the right team, it is crucial to have an effective facilitator. The ACA facilitator is part of the patient safety team and is trained in relationship building, strong communication, managing expectations, and conflict resolution along with skills in performance improvement and safety methodologies. The facilitator begins every ACA with confidentiality rules, an overview of the event, and intent to focus on system (rather than individual) failures. Although no specific training is required for ACA participation, the most effective ACA teams are composed of individuals who have an intuitive grasp of the profound influences of the system and/or organization on the likelihood of error. This paradigm is embodied in the ever-growing concept of just culture in health care.19 ACAs are most effective when attendees have a working knowledge of just culture and subscribe to the belief that it is necessary to identify system changes to improve safety rather than focusing on individual blame. At our hospital, just culture means that staff are accountable for their actions but are not blamed for system failures beyond their control. These concepts are woven into the introduction that the patient safety lead provides for the team.
Right Analysis
Another core feature of facilitated ACA is using the right analysis, and this is led by the patient safety facilitator. It is important that the right sequence of events is established before identifying where in the process the error occurred. Tools like asking the 5 “whys” may help identify a system issue rather than a more superficial human error. Specifically, the facilitator will help ensure that a human error is not the identified cause; instead, the team will try to identify error precursors. However, using the right analysis may be challenging because some events are better suited to other types of in-depth analysis. For example, events involving processes of high complexity may be better suited to an RCA because more intensive resources may be required to fully understand the systems involved.
The Right Actions
Without the development of strong action items, the ACA’s effectiveness will be limited. Clinically meaningful action items often require significantly greater resources when compared with the weaker action items of education or retraining. There is certainly a role for education, but in striving for reduction in the risk of error, it is important to identify highly reliable interventions and push the system to more sophisticated systems science.14,15,18 A responsibility of the patient safety facilitator is to help the team brainstorm high-reliability actions to create a safer system. These actions target system-focused improvements rather than people-focused improvements.21 As such, the time-limited model of the facilitated ACA presents a challenge in creating meaningful action items that can require project planning, approvals, and resource allocation. To navigate this issue, ACA action items focus on the first practical steps toward larger improvements, and participants are asked to describe action items that are within their control. Rather than promising the delivery of a large initiative, the team documents a first practical step, such as proposing a practice change with a division chief or requesting a price estimate. By framing action items as the first practical steps toward larger improvements, the ACA builds accountability and provides direct guidance. The patient safety team continues to engage with the appropriate stakeholders to ensure that larger initiatives do move forward. Presenting the ACA at Patient Safety Committee meetings often leads to broader organizational adoption of the actions and identification of additional resources to assist in improvement by engaging different leadership levels.
The creation of the right action plan can be challenging because although the ACA team strives for strong practical actions that can address the cause of the event being examined, there may be barriers to the development of such actions (for example, information technology system architecture or available resources). However, the patient safety team is in place to address these barriers, shape the action plan, and move the action plan forward to achieve change.
Next Steps and Summary
We have received positive feedback from hospital unit and care team leaders involved with a facilitated ACA, particularly because they are given the opportunity to discuss solutions in more depth and have structured follow-up. As we continue to focus on improving patient safety culture at our institution, next steps will include a deeper understanding and measurement of the impact of an ACA and its action items. For example, after the actions from the facilitated ACA are completed, did a similar event recur? We also want to further investigate potential correlations between the facilitated ACA model and improvements in patient outcomes and/or hospital safety culture, as noted in culture survey scores.
In our institution’s journey to zero harm, we have moved from a paper-based e-mail query to an active, facilitated cause analysis format with multidisciplinary engagement. To maximize resources and ultimately impact safety culture, a facilitated ACA model was created with the explicit goal of providing more opportunities for analysis and improvements for events that may be overlooked or simply attributed to human error. We have seen tremendous acceptance from the local leaders that now feel empowered to proactively request engagement for an ACA when appropriate. The guiding principle of this facilitated ACA model is to move beyond human error and focus on improving systems to prevent similar errors from occurring in the future. We believe this tool may be helpful to other hospitals that seek to improve their safety culture.
Drs Parikh and Shah and Mr Hochberg contributed to the conception and design and the acquisition and interpretation of data, drafted the initial manuscript, and reviewed and revised the manuscript; Ms Merkeley, Ms Fahey, and Ms Lavette contributed to the conception and design and the acquisition of data and provided edits and revisions to the submitted manuscript; Dr Cheng contributed to the design, acquisition of data, analysis, and interpretation of data and provided edits and revisions to the submitted manuscript; and all authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work.
FUNDING: No external funding.
References
Competing Interests
POTENTIAL CONFLICT OF INTEREST: The authors have indicated they have no potential conflicts of interest to disclose.
FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.
