6.1.1 Patient safety culture is not fully developed in two Finnish state-run forensic psychiatric hospitals (Article I)
The highest percentage score (72%) was for the “teamwork within units”
dimension. Vlayen et al. (2012, 2015) and Hamaideh (2017) have reported similar findings in their studies. Nevertheless, the lowest scores were showed in the dimension “nonpunitive response to errors”, which received regularly lower values in an extensive American study (Sorra & Battes, 2014).
Hospital managers showed more positive perceptions of PSC than frontline staff. This is consistent with previous research, i.e., managers have often scored PSC higher than other members of the organization (Parand et al., 2010; Nordin et
al., 2013; Danielsson 2019; Hao et al., 2020). This should be considered when designing patient safety education because there are clear differences between the views of managerial and frontline staff concerning certain aspects of patient safety and quality improvement (Singer et al., 2008; Price et al., 2007). This may also reflect how the job tasks of frontline workers and managers differ, and may cause certain staff members to misunderstand patient safety problems; for example, nurses may be overworked and suffer from stress due to experiences of patient violence and under-staffing.
Previous studies have shown that a positive PSC requires the adequate
education of health care professionals, including managers (Milligan, 2007; Azimi &
Bahadori, 2012). This was also the case in the research underlying this dissertation, as supplementary education and education seemed to be a
significant factor in transforming PSC. However, only 30% of responses indicated positive scores for management support for patient safety. Ree & Wiig (2020) recently suggested that safety cultures in health care facilities are strengthened by strong management and leadership (Ree & Wiig, 2020). Hence, managers must comprehend their responsibility for fostering PSC. This means that patient safety should be emphasized at the strategic level, and management should implement activities that support the development of PSC (Liukka et al., 2018).
A staff member’s educational background, manager status, and the hospital under study were found to significantly influence PSC scores. This suggests that numerous subcultures exist within organizations, and should be taken into account when planning PSC improvements.
Moreover, the PSC dimensions ‘frequency of events reported’ and ‘non-punitive response to errors’ might be interrelated. The results also revealed that hospitals can differ in the rate at which adverse events are reported and what type of feedback is provided. Feedback should be given to staff without making them feel guilty about their errors. Nevertheless, the managers must find a balance when giving feedback, as each staff member must understand that negligent and unethical behavior will not be tolerated.
6.1.2 Positive changes in patient safety culture following a three-year safety intervention (Article II)
The studied intervention, which was the implementation of a health care incident reporting system, resulted in significantly positive changes in HSOPSC scores, and thus, exerted a positive effect on PSC. These results are consistent with findings that the implementation of CRM (Hefner et al., 2017), TeamSTEPPS (Amiri et al.,
2018), and different patient safety initiatives (Hamdam & Saleem, 2018) positively influence PSC.
During the study period, the dimensions showing the most significant positive changes were “Supervisor/manager expectations & actions regarding patient safety”, “Management support for patient safety” and “Feedback and
communication about errors”. These findings are important because middle managers play a key role in facilitating changes to organizational culture and promoting organizational learning; as such, their commitment to safety can have a significant influence on PSC (Gutberg & Berta, 2017; Reis et al., 2018).
The positive changes in PSC can be attributed to the introduction of a patient safety reporting system, which specifically affected the dimension
“Supervisor/manager expectations & actions regarding patient safety” because it included near-miss situations. The reporting of such incidents helps staff to better understand patient safety risks and learn from each experience, allowing similar situations to be avoided in the future. Moreover, safety incident reports can identify the possible causes of failures in work processes and structures (Reis et al., 2018) and provide feedback about reported mistakes (Boussat et al., 2015;
Hamaideh, 2017). However, this requires substantial effort, and the management must be committed to a culture of safety to set an example for the entire
organization (El-Jardali et al., 2011). The dimension “nonpunitive response to error”
received the lowest scores at the study hospital, which indicates that a just culture is only now emerging despite the positive changes to PSC. This is a commonly observed problem in health care organizations, with a recent systematic review reporting that nearly 70% of previous studies (N=22) observed weak results in this dimension (Reis et al., 2018). For this reason, Burlison et al. (2016) stressed that an organization must foster an inclusive culture in which error reporting helps staff at all levels to learn from errors. Moreover, Okyama et al. (2018) identified three key factors conducive to PSC improvement: feedback following reporting, engaged leadership, and an environment dedicated to learning from errors.
6.1.3 Nurses’ views highlight a need for the systematic development of patient safety culture (Article III)
The responding nurses highlighted six themes which reflect aspects of PSC that could be developed and improved. Issues related to human resources, safety guidelines and management appear to be particularly important to forensic nursing.
Previous studies have shown that an adequate amount of competent staff is associated with patients' violent behavior, i.e., interactions between highly skilled nurses and patients could prevent violent events (Lantta et al., 2016) and the work stress caused by high levels of patient aggression (Pekurinen et al., 2019).
The importance of operational safety guidelines was also highlighed by nurses.
Hospital management and nurse managers are responsible for safety- and etchial guidelines and rules, in which the best available evidence has been considered (Kadivar et al., 2017; Ricciardi, 2021).
The presented research indicates that managers need up-to-date knowledge of their nurses’ working environment to ensure staff competence and remediate any deficiencies through training and education (Davidoff, 2011). Moreover,
management practices strongly influence staff views of patient safety (Kanerva et al., 2013), as satisfaction with leadership was found to be inversely linked with overall exposure to patient aggression (Pekurinen et al., 2019).
Most of the responding nurses felt that the wards generally had an open communication culture, yet the errors were not always discussed in a systematic way. It was previously noted that a dysfunctional communication culture can compromise patient safety (Garon, 2012; Kanerva et al., 2016), and
well-functioning communication was identified as a specialized skill in mental health care (Kanerva et al., 2015). A systematic approach to handling errors was still evolving, with the findings revealing that a blaming culture took precedence over a just culture with transparent decision-making at certain times. Striking a balance between the extremes of punishment and blamelessness is a key goal in the development of a PSC (Dekker, 2008; Boysen, 2013), with organizational justice and collaboration among nurses potentially minimizing patient violence (Pekurinen et al., 2017).
The responding nurses´ views concerning the balance between occupational and patient safety was another crucial factors. According to Yassi et al. (2005), patient safety can only be improved by also focusing on the safety of staff members. Previous research has shown that patient involvement is positively related to the quality and safety of care (Kontio et al., 2014), and can improve health outcomes (Bombard et al., 2018). Furthermore, health care professionals must listen to and respect patients and their family members if they are to foster effective therapeutic relationships (Denhamn et al., 2008; Mazor et al., 2010).
6.1.4 Voluntary safety incident reports dominated by violence (Article IV) Over half (51%) of the reported incidents resulted in no patient harm, with severe/serious harm occurring in only a small minority of incidents (0.5%). These findings are consistent with previous psychiatric studies (Bader et al., 2014, 2015;
Marcus et al., 2018) and assessments of patient safety system databases, in which less than 1% of all incidents were found to result in severe/serious harm or death (Howell et al., 2015; Rauhala et al., 2018).
In the research underlying this thesis, violence represented the most commonly reported patient safety incident, followed by medication errors and accidents.
Papoulias et al. (2014) previously postulated that patient-to-patient interactions could explain up to 25% of the aggression or violence that occurs in psychiatric settings. Meanwhile, Quanbeck (2006) reported that these types of incidents usually result from illegal, and frequently violent, acts (Quanbeck, 2006). According to previously reports, medication errors are quite frequent in psychiatric settings (Marcus et al., 2018; Reilly et al., 2019). An increased prevalance of medication errors relative to other health care settings could be explained by the large amout of medications that the forensic psychiatric patients are administered.
Furthermore, 12.3% of the reports analyzed in article IV were related to accidents, such as falls. Similar proportions of accidents have been observed in acute hospital care (Marcus et al., 2018), with medical-surgical units showing noticeably lower rates of accidents (Abraham, 2016).
Violence against another patient was described in most (67%) of the incident reports, followed by self harm. This can be explained by the fact that many forensic psychiatric patients have been diagnosed with schizophrenia (or
psychosis), and often have histories of violence and substance abuse (Dack et al., 2013; Iozzino et al., 2015). Previous research has reported that a record of violence is a risk factor for future violence (Webster et al., 1997), which can be taken to mean that hospital violence is expected in forensic psychiatric facilities. Notably, numerous publications state that over 50% of forensic psychiatric patients have experienced violence from another patient (Bader et al., 2014, 2015; Novaco &
Taylor, 2015).
According to article IV, a significant share of violent incidents occurred in public spaces with large numbers of patients, such as corridors and day rooms, while incidents of self-harm mainly occurred in the patients’ own rooms. This is consistent with the findings of a literature review by Bowers et al. (2011) and a study by Hamrin et al. (2009).
Insight into where violent events occurred suggests that a lack of personal space, along with certain stimuli (noise levels and close patient-patient
interactions), are aggression-promoting factors that should be considered when designing and monitoring psychiatric wards (Ulrich et al., 2018). The effect of the environment on patients' aggressive behavior has been discussed before (Kontio et al., 2014), with the lack of personal space emphasized by certain researchers (Meehan et al., 2006; Pulsford et al., 2013; Papadopoulos et al., 2014). Other researchers have suggested locked open spaces as a solution (Lantta et al., 2016).
The research presented in this thesis suggests that the daily rhythm and activities in a ward (health care professionals switching shifts and medication or mealtime transitions) could be a better predictor of aggression rates than the time of day.
This conclusion already has empirical support, as prior research has shown violent behavior to be linked to certain clinical activities and nursing staff shift changes (Weizmann-Henelius & Suutala, 2000; Iversen et al., 2016).
When health care managers’ reactions to the incident reports were gauged, the most frequently mentioned solution was that the incident should be discussed or the provider should be notified. This was expected for the psychiatric setting, an environment which has a tradition of assessing human actions and discussing symptoms - often considered the main contributors to violence. This result is consistent with what was reported by Archer et al. (2020), who described two major themes underlying the issues that occur in mental health care settings:
“interaction between patient diagnosis and incidents” and “aftermath of a violent incident-prosecution”. A weakness of only discussing incidents or relaying
information is that this approach focuses more on human behaviors than the initial cause(s) of incidents (Anderson & Kodate, 2015; Mitchell et al., 2016), potentially resulting in single-loop (patient- or ward-specific) learning. This dynamic has previously been noted for patient incident report systems, which provide the optimal results when manged by clinical teams or a specific
department rather than operated at an organization-wide level (Stavropoulou et al., 2015).