FormalPara Clinician’s capsule

What is known about the topic?

Patient safety events are common during medical care, yet are largely underreported.

What did this study ask?

We sought to engage Emergency Department staff and leadership to develop requirements for an improved patient safety event reporting system.

What did this study find?

Frontline Emergency Department staff view clear definitions, transparency, and simplicity as key features of patient safety reporting systems.

Why does this study matter to clinicians?

Understanding perceptions of patient safety event reporting systems allows for meaningful improvements that may enhance use of the system

Introduction

Patient safety events (PSEs) are unwanted or unexpected events that occur during medical care, and include near misses, where the incident does not reach the patient; no harm events, where the incident reaches the patient but does not cause harm; and adverse events, where the incident reaches the patient and results in injury, harm, disability, or death [1, 2]. High cognitive loads and frequent interruptions make emergency departments (EDs) uniquely error prone environments, threatening patient safety and reducing care quality [3,4,5,6,7]. Furthermore, EDs are overcrowded and understaffed, increasing risks to patients and clinical burden on staff [8,9,10].

In 2007, British Columbia’s (BC) health authorities began implementing Patient Safety and Learning System (PSLS), a privately developed, web-based incident reporting software, to capture PSEs and identify system factors to prevent their recurrence [11]. Clinicians report in PSLS voluntarily using electronic forms that capture information about different incidents (e.g., falls), including a description of the event and its impact, contributing factors, patient and reporter information, and follow-up actions taken. Staff must submit incident reports, as defined in the Disclosure of Patient Safety Incidents policy, to PSLS per the Incident Management policy [12, 13]. Training and reporting guidelines are available on the staff Intranet. After clinicians submit reports, supervisors are responsible for reviewing and coordinating the investigation of incidents, including following up with the reporter either one-on-one, in meetings, or through group communications to close the loop on reporting [13].

Studies indicate underreporting is commonplace in voluntary incident reporting systems despite institutional policies [14,15,16,17]. Fear of punitive action, poor safety culture, and time required to complete forms are primary deterrents to reporting [2, 15, 18,19,20,21]. Unique challenges in the ED may compound these issues [3,4,5,6,7]. While previous investigators have studied barriers to PSE reporting and suggested that end-user engagement may improve reporting platforms, few studies have sought to understand limitations of and improve upon existing PSE reporting processes from the perspective of ED staff and departmental leadership at an acute care setting in Canada [11, 22,23,24,25]. The primary objective of this study was to explore the beliefs and perceptions of ED staff and leadership of the current PSE reporting system and identify features that they believe are important in such a system.

Methods

Study design, setting, and time period

We conducted a qualitative study using semi-structured focus groups, interviews, and workshops at Vancouver General Hospital (VGH), a large academic urban hospital with an annual ED census of 95,000 patients, between August 2017 and May 2019 in a research office on site [26]. 68 physicians, 60 residents, and 260 nurses work in the ED.

We approached this study through the philosophical and analytical framework of constructivist grounded theory, recognizing the researcher’s role in the co-construction of knowledge [27]. We used the Consolidated Criteria for Reporting Qualitative Research (COREQ) checklist to ensure a robust study [28]. We received ethics approval from the University of British Columbia Behavioural Research Ethics Board (H16-01238).

Recruitment

VGH ED nurses, physicians, and departmental leaders were eligible to participate. Using convenience sampling, we approached eligible individuals at VGH to participate in a focus group or interview. We sought to sample a representative distribution of clinical staff based on staffing in the ED and aimed to continue recruitment until we reached data saturation, but pragmatically recruitment ceased when no additional individuals volunteered within a reasonable timeframe. A pre-existing professional relationship with participants who worked in the ED facilitated recruitment. We sent study information to those who agreed to participate via email, which informed prospective participants of the study purpose. At the completion of each focus group or interview, we invited participants to attend a workshop.

Data collection

We prepared two semi-structured focus group and interview protocols, one for frontline staff and one for departmental leadership, informed by a review of relevant literature completed prior to study initiation. We used our protocol to guide the sessions, capturing emerging topics and ideas to explore in subsequent sessions (Appendix 2).

The principal investigator (TS), an emergency medicine resident, received methodological guidance and training from a social scientist on the study team (EB) with qualitative expertise. The principal investigator (TS) led the focus groups and interviews. A research assistant (DP) attended some sessions to co-facilitate and take notes that highlighted emerging topics and areas for further exploration in subsequent sessions. At the beginning of each session, we reiterated study goals and introduced the researchers. We continued each focus group or interview until we reached data saturation, ranging from 45 min to 1.5 h. Toward the conclusion of the focus group and interview portion of the study, we felt we achieved saturation as no new concepts or ideas began to emerge.

We developed a list of ideal features for a PSE reporting based on focus group and interview findings, which we presented to workshop participants. We used feedback to iteratively refine and finalize the list of ideal features. We audio recorded all interviews and focus groups, which a professional transcriptionist transcribed for analysis.

Data analysis

We coded and analyzed transcriptions using NVivo 12 (QSR International, version 12, 2020). We employed an inductive, grounded method to develop the coding structure and utilized the constant comparative method to incorporate relevant data within the final coding structure [29]. We referred to field notes to identify emerging themes from the sessions. Three researchers (TS, DP, SSS) met intermittently to refine the coding structure until the final structure was agreed upon (Appendix 3). We then recoded all transcripts using the final coding structure, which organized participant comments along the following themes: factors affecting PSE reporting behaviour; medium of PSE reporting; and, components of an ideal system. Two researchers (TS, DP) coded and compared 20% of the transcripts to ensure that the application of coding structure was consistent. We used thematic analysis to establish themes [30, 31].

Results

Focus group, interview, and workshop participant characteristics

Fifty participants attended at least one focus group, interview, or workshop. Each focus group had 2–5 participants, while interviews were one-on-one. We held two large group workshops with 28 and 9 participants, respectively. Nine individuals participated in both a focus group or interview and a workshop. We gathered information on participant’s training and years in practice for analytical purposes (Table 1).

Table 1 Characteristics of focus group and workshop participants
Full size table

Beliefs and perceptions of PSEs and current reporting practices

In focus groups and interviews, we sought to understand beliefs and perceptions about current PSE reporting practices. We describe our three main findings with supporting quotations.

PSEs are common but poorly defined

Staff were unaware of guidelines that define PSEs and differed in what they believed constituted a reportable event (Table 2). Participants generally agreed that events that led to direct harm and near misses should be reported, however, some believed that their colleagues might not perceive near misses as worth reporting.

Table 2 Representative participant quotes regarding definition of PSEs
Full size table

Participants identified several examples of reportable events (Appendix 1). There was consensus that medication errors, falls, and violence constitute reportable PSEs, representing three of the four most commonly discussed PSEs. These are easily identifiable events, and have been the focus of patient safety initiatives, particularly among frontline nurses. As one departmental leader noted: “we really were encouraging reporting of falls.” Participants also frequently cited operational challenges as PSEs, but were unsure if a delay in care should trigger a report.

There was less consensus about whether to report system-level issues, like overcrowding. Physicians were more likely to view these as PSEs, but rarely reported them in PSLS. Participants acknowledged that conditions that reduce care quality may constitute PSEs, but their frequency made it difficult for staff to differentiate between suboptimal conditions of a busy ED and a PSE. One physician commented: “it happens so frequently it's hard to call it [a PSE] because it becomes normal but it's not okay.” Among all groups, identifying reportable events did not necessarily translate into a completed report.

Patient Safety Learning System is underutilized and perceived as not meeting clinician needs

Most participants were aware of PSLS as a PSE reporting tool, and some had used PSLS, although no resident physicians had. Many of those who had used it had negative experiences citing rigid, time-consuming forms with unfamiliar fields that do not fit clinical workflows (Table 3).

Table 3 Participant quotes about beliefs and perceptions of PSLS
Full size table

Nurses expressed frustration with the redundancy of PSLS that necessitated duplicate charting. Physicians viewed PSLS as more of a risk management tool than a mechanism for quality improvement. Staff did not see how PSE reporting in PSLS translated into an improved work environment or patient care, creating a sense of futility. Although departmental leadership used PSLS to follow up on reported events, they acknowledged its lack of functionality for frontline staff.

Participants want to and do report PSEs through informal channels

Staff viewed informal PSE reporting as a professional responsibility and a tool to facilitate change by directing attention to emerging problems, informing education, and engaging staff in improvement efforts. These factors motivated staff to report PSEs through informal channels that evolved organically to meet these needs, including emails, verbally, and in nursing notes (Table 4). The perceived duty to report did not extend to PSLS, which staff bypassed for all but the most serious events.

Table 4 Real-world PSE reporting practices
Full size table

Perceived features of an ideal PSE reporting system

Informed by participants’ beliefs and perceptions about PSEs and current reporting practices, we subsequently captured perspectives about the features of an ideal PSE reporting system. We describe the three main features below, and summarize supporting quotations in Table 5.

Table 5 Features of an ideal PSE reporting system
Full size table

Provide clear definition of PSEs

Participants acknowledged a need for shared understanding and clear definition of what constituted reportable PSEs among staff and departmental leadership to support reporting. Participants suggested that including event type classification on PSE reporting forms may increase clarity about what constitutes a reportable event, and that broader communication from departmental leaders may increase PSE identification and reporting among staff.

Provide transparency about what reporting does

Staff want to share information that will lead to improved patient safety and care quality, and departmental leadership want to receive this information, but many believed that current PSE reporting channels lack sufficient transparency to demonstrate improved outcomes following reporting. Participants want transparency about what reporting does and were interested in being able to indicate their preferred level of feedback within the reporting form to clarify expectations and increase reporting.

Simplify the reporting form

Participants felt that PSE reporting should be simple and easy to use with free-text options and few mandatory fields, and that leveraging existing informal reporting channels may reduce duplication. Many acknowledged that detailed reporting requirements are a time burden and barrier to reporting, yet some managers noted the need for sufficient data to determine next steps, follow up, and generate feedback. Participants agreed that an ideal reporting form should balance simple data entry while enabling further escalation and investigation if needed.

Discussion

Interpretation of findings

In this study, we explored ED staff and leadership beliefs and perceptions of PSE reporting. As in other similar settings, PSEs in our ED are likely underreported. Beliefs and perceptions about the PSE reporting system, including low awareness of the system itself, are contributing factors to its underuse. Despite a willingness to report PSEs and receive PSE reports, the current system fails to meet clinicians’ needs and lacks clarity around reporting processes and actions taken. Participants identified features of an ideal PSE reporting system, which may increase usage of and engagement with the current system in capturing PSEs, if incorporated. Creating clear definitions of PSEs and reporting guidelines may support clinician understanding of reportable event types. Establishing closed feedback loops may clarify how and why clinicians’ reports affect decision-making. Streamlining and simplifying the reporting process may reduce the burden of documentation among busy frontline providers. Ultimately, the ideal PSE reporting form must balance leadership’s information needs for quality improvement with frontline ED staff’s limited time.

Comparison to previous studies

Several past studies described barriers to PSE reporting that are consistent with our study. A 2011 study identified six barriers to PSE reporting, which included inaccessibility of reporting forms [18]. A related study on adverse drug events observed that providers documented events in charts to support continuity of care but never reported them to external agencies through formal reporting channels, citing time constraints and duplication [19].

Previous studies have also identified features that facilitate PSE reporting that are consistent with our study, including clear reporting guidelines, usable reporting systems, and feedback loops to create learning opportunities [2, 20]. Previous investigators suggested engaging frontline healthcare providers in design processes to improve reporting and enable systematic data collection required to advance patient safety in EDs [32]. In keeping with prior suggestions, our study adds to the literature a nuanced understanding of frontline clinicians’ and departmental leaderships’ perceived ideal features of a PSE reporting system, which aims to address reporting barriers and underreporting in the ED context.

Strengths and limitations

A strength of this study is recruitment of participants with diverse clinical backgrounds, time in practice, and roles in the ED, including ED leadership, resulting in a diversity of perspectives. We sampled to thematic saturation by completing focus groups and interviews until no new themes emerged, and we returned concepts to staff multiple times to refine themes and ensure we interpreted data correctly.

The context of the study may have affected participant responses. Some participants knew the interviewer. Furthermore, the group setting may have been intimidating to some participants, causing them to be less forthcoming with their answers.

Individuals who agreed to participate are more likely to be interested in PSE reporting, and thus may have been more familiar with the PSE reporting process, leading to bias. We sampled a small proportion of staff from a single ED, which may limit the generalizability and transferability of our findings to other settings; however, this was not a primary goal of our study. Residents, who were slightly over-represented, were less familiar with PSE reporting processes, which may have affected findings.

Being an emergency physician, the principal investigator undertook the project with an understanding of responsibilities toward patient safety and expected use of PSLS. We mitigated the risk of bias and increased the credibility of our study by including team members with different disciplinary backgrounds in the research project. In particular, having three researchers develop the final coding structure and two researchers coding 20% of the transcripts reduced the risk of subjectivity bias and increased reliability.

Clinical implications

Current PSE reporting systems do not meet the needs of the frontline staff who are responsible for reporting, and consequently, fail to meet the data needs of departmental leadership. This study has allowed us to understand the needs of ED staff regarding PSE reporting. If this data is used to streamline PSE reporting processes it may lead to increased patient safety and quality of care.

Research implications

Future work should seek an improved understanding of which features of PSE reporting systems are associated with increased use by ED staff, and which features support the identification of local priorities and action items.

Conclusion

This study reveals that ED staff view clear definitions, transparency, and simplicity as key features of a PSE reporting system. While operationalizing these findings may support PSE reporting in the ED, other systemic barriers must be addressed to generate comprehensive patient safety data in the ED that can be used for surveillance, performance assessment, and to identify new or uncommon sources of harm. Only then will the ED foster an environment of patient safety, clinical excellence, and staff satisfaction.