1466-7657/asset/INR_large.jpg?v=1&s=f5c831ab5a26cf45b9c4a769482df46d4219d3d6)
The Hospital Survey on Patient Safety was used to identify opportunities for safety culture improvement in a 30-bed intensive care unit. Based on the survey results, a core team decided to focus on three safety domains: reporting errors, approachability of authority figures and handovers. The project team subsequently interviewed 39 intensive care unit staff members, gathering information on these three domains that will inform future safety efforts. Numerous barriers and facilitators to improvement were described. This mixed-methods approach could be applied in other hospitals seeking to quickly yet thoroughly understand how their safety culture can be improved. Developing local strategies to reduce these barriers may promote a safer patient experience at our hospital.
A rapid assessment of barriers and facilitators to safety culture in an intensive care unit
-
Funding
This project was funded by a grant from the Health Research and Educational Trust as part of the Project Protect Fellowship.
-
Conflict of interest
None to disclose.
Abstract
Introduction
Hospitals are complicated organizations involving multiple disciplines with overlapping relationships and responsibilities. In this complex environment, new initiatives can fail simply because of poor communication, distrust of management or conflicts among providers. Cultural, or adaptive, changes are therefore an important part of any successful implementation effort.
Many quality improvement efforts have evaluated a hospital's safety climate – a surrogate for hospital culture. Safety climate refers to frontline staff's shared perceptions of patient safety norms and behaviours in a specific clinical area. Studies have found an association between safer climates and a reduction in adverse events, such as accidental injuries and medication errors.
In this quality improvement project, we assessed whether we could rapidly identify opportunities to improve patient safety climate within an intensive care unit (ICU) through the use of a standardized survey and in-depth interviews of healthcare workers. The findings from our project were intended to inform local interventions to improve patient safety.
Methods
This study was performed at a hospital in the mid-western United States. This project was reviewed by the hospital's research officer and deemed to be quality improvement, not research.
The hospital's ICU is a 30-bed unit that cares for trauma, surgical and medical patients. During June 2014, all individuals who regularly work in the ICU were asked to complete an electronic version of the Hospital Survey on Patient Safety (HSOPS), which seeks feedback on safety issues, medical errors and event reporting. The survey is published by the Agency for Healthcare Research and Quality; it has been shown to be both reliable and valid (Colla et al. 2005).
After the survey was completed, a multidisciplinary core group was convened to review the aggregate HSOPS results. This group included two ICU nurse leaders, an ICU staff physician, an infection preventionist, the chief nursing officer and the hospital epidemiologist. The group identified three items from the HSOPS in which the unit score was low: error reporting, providing feedback to authority figures and handovers.
Next, ICU employees were interviewed about these three items of focus. Two nursing students were hired to conduct these confidential, semi-structured interviews. Since the nursing students were not part of the hospital or its ICU, it was felt that they would elicit more honest responses from employees. ICU employees were recruited to participate during their regular working shift. For each of the three items, open-ended questions were asked in line with a published ‘check-up’ tool (Sexton et al. 2007). Additional employees were recruited until theoretical saturation was reached.
All interviews were performed during October and November 2014. Each interview lasted 10–15 min. All interviews were audio-recorded, transcribed and de-identified.
Each interview transcript was reviewed for accuracy and completeness. Transcripts were analysed using emergent thematic analysis (Livorsi et al. 2015). Recurrent themes were organized using elements and sub-elements of the Promoting Action on Research Implementation in Health Systems (PARiHS) framework (Rycroft-Malone 2004).
Results
The survey was distributed to 270 ICU staff members. Forty-four ICU staff members (16%) participated in the HSOPS survey, including 14 nurses, 8 pharmacists, 6 physical/occupational/speech therapists, 5 respiratory therapists, 2 dieticians, 7 physicians and 2 other staff members. Table 1 shows three items on which the ICU scored poorly and which were the topic of the subsequent interviews.
| 1. Staff feel free to question the decisions or actions of those with more authority. |
| 48% chose ‘most of the time’ or ‘always’ |
| 2. When a mistake is made, but is caught and corrected before affecting the patient, how often is it reported? |
| 39% said ‘most of the time’ or ‘always’. |
| When a mistake is made that could harm the patient, but does not, how often is this reported? |
| 52% chose ‘most of the time’ or ‘always’. |
| 3. Important patient care information is often lost during shift changes. |
| 52% chose ‘agree’ or ‘strongly agree’. |
Thirty-nine ICU staff members participated in the interviews, including 20 nurses, 7 physicians, 7 physical/occupational/speech therapists, 3 pharmacists and 2 respiratory therapists. Barriers and facilitators to the three processes were identified, as described below (Table 2).
| Process | Barrier | Facilitator |
|---|---|---|
| Submitting incident reports |
|
|
| Approachability of authority figures |
|
|
| Handovers |
|
|
Barriers to submitting incident reports
The largest barrier to reporting errors was fear of repercussions. Nurses expressed concerns about being ‘shot down’ by physicians for having made a mistake or being seen as a ‘stickler’ or a ‘tattle-tale’ for reporting on a colleague. According to one speech therapist, the thought of submitting an incident report was ‘scary’, because the report would be a permanent record of the event. This fear stemmed, in part, from a perception that managers ‘don't treat all staff members the same’ or the acknowledgement that the staff members ‘do not already have a great relationship’ with their supervisor.
Establishing anonymity or providing positive reinforcement could help alleviate these fears. According to one nurse, ‘right now we have to write down our names and who is reporting it. If it was completely anonymous, then maybe people would be more willing to report’ (Nurse 3). Others highlighted the importance of positive feedback, or thanking those who do submit reports:
If somebody does fill out a report, are they thanked for contributing to the culture of safety at our hospital? Probably not. They are probably just risking everything and not ever thanked for anything. So, thanking people for calling attention to these matters [would be helpful] (Staff physician 5).
Several other barriers to reporting errors were identified. Some felt that the incident report form could be simplified or streamlined. There was confusion about what type of event justified an incident report, and the physicians-in-training were not even aware of how an incident report was submitted. A few participants did not feel motivated to submit a report, because they were not convinced they would receive any feedback or that the report would effect a positive change in the system. Others dismissed the importance of reporting errors, ‘I think sometimes people have the mentality, “Well, if it did not harm the patient, it is okay”. It is not okay, but that is the mentality’ (Nurse 14).
Approachability of authority figures
Participants generally interpreted ‘authority figures’ to be physicians, but some participants also referred to unit managers or senior hospital leaders as ‘authority figures’.
A major obstacle to approaching authority figures was the perception that their personality-type was difficult. These challenging personality types included physicians whose ‘egos are too strong’, nurse managers who are ‘intimidating’, and insecure individuals who ‘might take it [feedback] the wrong way or get upset’.
Some acknowledged that the approach to feedback may also be part of the problem. Nurses who are ‘sarcastic, are angry and use a tone’ prompt a negative response in the recipient. Doctors’ competency in talking to nurses was also questioned. Training in communication skills was cited as a potential strategy for improvement.
The reluctance to ‘speak up’ may reflect the overall negative culture of the unit. Participants expressed frustration that nothing ever changed in the unit:
All of us have been here so long that…you can say something and things just do not change…So there is no point really in questioning authority because nothing changes (Nurse 15)
This sense of hopelessness led to a cynicism that feedback was of no utility. Instead of sticking their neck out, nurses instead decided that it was easier to just remain silent.
To facilitate ‘speaking up’, nurses could engage senior nurses for assistance, but two physicians observed that this was infrequently done. Physicians felt that nurses were more likely to give input when the physician team purposefully included the nurse in team rounds. A nurse's level of experience also facilitated ‘speaking up’: ‘If you're more experienced and you've been here longer, you feel a lot more comfortable with questioning things’ (Nurse 10).
Handovers at nursing shift change
Interruptions were a commonly identified barrier to effective handovers. Several participants described their reports being disrupted by family members, telephone calls or physicians. The unit has no clerk or aide to assist in addressing these issues while the nurses are in-report, and the hospital has a 24-h visitation policy for family.
Some nurses felt that the use of a more standardized report would avoid lapses in passing on information. A few nurses cited their experience at other hospitals, explaining how a standardized written handover sheet facilitated the system-by-system review of the patient.
The place I came from before I worked here had a sheet of paper that hit system-by-system and then a little area that had common procedures or something that would have happened during the day, and that way you…are very routine….[and] you would not miss things (Nurse 1).
Nurses acknowledged the importance of self-accountability during their handovers. Accuracy was improved by ‘writing things down’ and being ‘a person who knows what to ask’.
Discussion
Though essential for patient safety, communication in health care is often inadequate and ineffective. Communication lapses were determined to be one of the major causes of adverse events reported to the Joint Commission between 2004 and 2010 (Smeulers et al. 2014).
Our project assessed three types of communication practices in an ICU: error reporting, providing feedback to authority figures and handovers. We chose these three practices because of their importance to patient safety. Hesitancy to speak up and sub-optimal handovers have both been identified as contributing factors to adverse events. Furthermore, the failure to report errors hinders system redesign and the prevention of future adverse events. Despite their importance, these processes are incompletely implemented across health care, and numerous barriers have been identified.
The barriers to error-reporting identified by our interviews have been well-described in the literature. Like our project, other studies have found that a high percentage of nurses fear disciplinary action and, in turn, decide not to report errors (Ulanimo et al. 2007).
Likewise, the barriers and facilitators we identified to providing feedback to someone in a position of authority have been described in other settings. These include an untrained approach to ‘speaking up’ and a culture of negativity (Okuyama et al. 2014). Feedback is facilitated by a responsibility towards patients, confidence, the perceived safety of ‘speaking up’, administrative support, and teamwork (Churchman & Doherty 2010; Lyndon 2008). Handovers are critical to maintaining the continuity of care, but there is no established standard for how they should be done (Smeulers et al. 2014). Prior studies have described the importance of minimizing interruptions (Kowitlawakul et al. 2015), a common barrier identified in our project.
The identification of these barriers is an important first step in improving the safety culture of our ICU. To encourage the reporting of errors, participants in our project highlighted the importance of simplicity, anonymity and timely feedback, which is consistent with prior reports (Hartnell et al. 2012). To facilitate speaking up, formal training programmes in communication skills have been shown to improve attitudes about speaking up and behaviour in simulated scenarios (Stevens et al. 2012). Finally, some simple interventions in our ICU can optimize handoffs, including the use of a standardized form and a concerted effort to minimize interruptions (Kowitlawakuul et al. 2015).
Adaptive, or cultural, changes are an important part of any quality improvement effort. In this study, we adopted qualitative research techniques to a local quality improvement effort. We used a standardized survey tool to conduct an initial assessment of safety culture, which was followed by in-depth interviews. One-on-one interviews with frontline staff provided rich feedback that may not have been obtainable through standard means. Since our project required few resources and little expertise, it could easily be replicated in other settings. The barriers we identified have been well-described in the literature, which supports the validity of our findings.
Our study did have some limitations. First, even though the interviews were confidential and conducted by nursing students not known to the unit, participants may not have felt comfortable providing honest responses. Second, our findings reflect a single unit at a single teaching hospital and therefore may not be generalizable to other settings. Third, since the response rate to the HSOPS survey was low, the findings may not reflect the general opinion of the ICU staff. However, the low response rate may be a reflection of the staff's level of engagement and the overall ICU culture. Since certain disciplines had limited participation in our interviews, it is possible that unique perspectives were missed.
In conclusion, we used qualitative research techniques to identify barriers to a stronger culture of patient safety within an ICU. The barriers we identified in our rapid assessment have also been well-described in the literature. Developing local strategies to reduce these barriers may promote a safer patient experience at our hospital. Other healthcare settings could model our approach to obtain a rapid yet thorough assessment of how their safety culture can be improved.
Acknowledgements
We would like to thank Alicia Gosselin and Jama Macdonald for conducting the interviews.
Author contributions
Study design: DL, MK, LB, KS, LM, KR, TH, NS
Data collection: KS, LM, KR, AG, JM
Data analysis: MK, DL
Study supervision: NS, DL
Manuscript writing: DL