Patient Safety Strategies in Psychiatry and How They Construct the Notion of Preventable Harm: A Scoping Review.

OBJECTIVES: The literature on patient safety in psychiatry has not been explored systematically in terms of what interventions are used, how they are used, and what type of (preventable) harm is targeted. The aims of this scoping review are to explore patient safety strategies used in psychiatry and determine how they construct the notion of preventable harm.
METHOD: A scoping review of literature on patient safety in psychiatry published in English between 2000 and 2019 was conducted using Scopus, MEDLINE, PsycInfo, and CINAHL. Keywords of patient safety strategies and possible outcomes were coded from the results, discussion, or conclusion. Patient safety strategies were inductively categorized into themes according to the focus of the strategy.
RESULTS: The review introduces 7 focus areas of patient safety strategies identified within the psychiatric literature: "risk management," "healthcare practitioners," "patient observation," "patient involvement," "computerized methods," "admission and discharge," and "security." The result shows that patient safety strategies mainly aim to reduce suicide, self-harm, violence, and falls and present a large diversity of measures, often aimed at reducing variability while increasing standardization.
CONCLUSIONS: The strategies that are supported in the literature to achieve safer psychiatry mainly arise from linear cause-effect models and rely on staff performance, competence, and compliance. Contemporary safety science acknowledges the performance variability of everyday normal work and sees risk as the dynamic migration of these daily activities. The field of psychiatry has not yet included this view of safety in the strategic actions to reduce preventable harm.

Patient safety is defined by the World Health Organization as the absence of preventable harm.[1] In the medical literature, there are at least 7 different definitions of preventable harm; the most common definition is the presence of an identifiable, modifiable cause of harm.[2] By this definition, if patients suffer from harm in health care and if that harm is identifiable as a consequence of a preventable cause, patient safety was compromised. This argument raises questions about how health care can ensure patient safety. In this scoping review, patient safety strategies in psychiatry are explored. Psychiatry is faced with unique patient safety challenges through the risk of patients being violent, harming themselves, or committing suicide. Efforts to strengthen patient safety are generally made through quality improvements such as best practices and standardization, which implicitly addresses patient safety.[3] However, patient safety strategies traditionally tend to address errors, and quality improvements focus on standards and processes to reduce unwanted outcomes. Both share the underlying assumptions that performance variability is a problem (a quality problem or a risk) and that errors emerge from deviations from the one best way to perform a task.[4] Consequently, when adverse events are seen as the products of “error” or a failure to comply with accepted practices, they are seen as preventable.[5] This theory of preventable harm and causality has been challenged by other disciplines of safety research.

Rather than seeing safety as the absence of preventable harm, contemporary safety science introduces the view that failure and success originate from the same kinds of processes of performance variability in complex systems[6,7] and the idea that symmetrical cause-and-effect relationships rooted in a linear and reductionist ontology fall short of explaining harm in complex and adaptive systems such as health care.[8] Wiig and colleagues[9] suggest that health care investigations of preventable harm should experiment with a variety of approaches to strengthen patient safety and should not be limited by root-cause analysis. It could be more useful to understand why preventable harm does not occur despite the variability in everyday work.

Although patient safety research and (to a lesser extent) discourse has to some degree been informed by contemporary safety science, the theory of preventable harm has produced different strategies in efforts to strengthen patient safety. The literature on patient safety in psychiatry has not been explored systematically in terms of what interventions are used, how they are used, and what type of (preventable) harm is targeted. Hence, the aims of this scoping review are to explore patient safety strategies used in psychiatry and determine how they construct the notion of preventable harm.

METHODS

A scoping review was chosen as the methodological approach.[10] This is suitable when investigating various research areas, identifying gaps within the studied field, and exploring the literature.[11-13] The scoping review followed the methodological framework of Arksey and O’Malley,[14] which includes synthesizing and interpreting the data by coding the selected studies according to keywords and themes.

To explore the literature on patient safety in psychiatry and to identify potential research gaps, the following research questions were formed: How does the literature define patient safety in psychiatry, and what strategies does it advocate in order to achieve safer psychiatry? How does this literature relate to contemporary safety science? To identify relevant keywords, a basic search using “patient safety” and “psychiatry” was conducted in the Scopus database. By snowballing search hits, 17 additional keywords were introduced: “accident,” mental health,” “error,” adverse event,” “deviations,” “harm,” “hazards,” “incident,” “injuries,” “failure,” “risk,” “near miss,” “safety,” “safety strategy,” “safety culture,” “safety management,” and “resilience.” It became clear that patient safety can be implied in keywords, meaning that a certain treatment or action, such as how to decrease the days stayed in the hospital as an inpatient, can be interpreted as a patient safety strategy because it reduces the likelihood of an adverse event. Because of the wide association of “patient safety,” finding articles within psychiatry that describe strategies for reducing inpatient suicide without mentioning patient safety could be worthwhile. However, this expanded concept means that other common areas often interpreted as reflecting patient safety, such as “structured communication” or “hygiene routines,” should also be included. To reduce the number of specific areas within patient safety, those not directly linked to the research questions were excluded. The research questions refer to strategies from an organizational perspective, so keywords such as “quality” and “prevention” were included. Proximity operators and hyphens were used to include plurals and spelling variants. The Scopus “suggested databases” option for the search string was used. Database selection was narrowed to Scopus, MEDLINE, PsycInfo, and CINAHL because of their multidisciplinary coverage of the psychiatric field. Searches in the Web of Science showed large overlaps in hits with Scopus and were excluded. The search code showed an increase in the number of articles after the millennium, so the small number published before the year 2000 was excluded. The search was limited to the English language; then, after removal of duplicates, 2355 documents remained (Fig. 1).

FIGURE 1

Sketch of literature search.

A data extraction form was developed using the Scopus Comma Separated Values export to Excel, which generated a chart of author(s), journal, year of publication, and title. Six additional categories were added to the chart: aim, methodology, essence message, comments, strategy focus, and discussed strategy outcome. The title and abstract were screened for inclusion; 2184 documents were excluded because they did not explicitly focus on patient safety in psychiatry. The identification process was repeatedly discussed with colleagues to solve any discrepancies about excluded articles; ultimately, 171 documents were included for full review. A gray literature search was added and included a snowballed reference list, advice from researchers within the field, and symposia articles from the following Web sites: the United Kingdom National Patient Safety Agency, the Canadian Patient Safety Institute, and the Resilience Engineering Association. Overall, 79 documents were excluded, mainly because the discussion of risk in these documents did not render a strategy to reduce that risk. In total, 92 documents were included in the final review.

The methodology of a scoping review aims to explore the literature with a thematic construction without judging the quality per se.[15] Keywords of patient safety strategies and possible outcomes were coded from the results, discussion, or conclusion. Patient safety strategies were inductively categorized into the following themes according to the focus of the strategy: “risk management,” “healthcare practitioners,” “observation,” “patient involvement,” “computerized methods,” “admission and discharge,” and “security.” The perceived or discussed outcome was linked to the strategy in a chart to highlight which outcome was the most common (Table 1). The themes were reviewed by colleagues and discussed until consensus was obtained.

TABLE 1

Focus Areas of Patient Safety Strategies, Divided by Argued Strategy and Potential Outcome

RESULTS

The reviewed articles were published between the year 2000 and the first half of 2019. Authors originated mainly from the United States (n = 35) and the United Kingdom (n = 22). The articles were dominated by primary research (n = 78) followed by reviews (n = 13) and one book chapter. Although the articles could consist of multiple subject areas, medicine (n = 54) and nursing (n = 41) were the most common and included a diversity in approaches. The included 92 articles describe 64 different patient safety strategies and 33 possible outcomes (Table 1). The Results section of this scoping review introduces the 7 focus areas of patient safety strategies identified within the psychiatric literature.

Risk Management

The studies within the risk management theme (Table 1) predominantly concentrated on how to efficiently calculate the risk of, or enhance conditions for assessments of, suicide, self-harm, violence, or falls.[14,17,20-22,26,28,29,32,34-36] These studies highlighted the importance of carefully choosing the right instrument to measure risk and showed that a poorly conducted risk assessment process can impact practice, policies, and management strategies in an undesirable way.[22,24,25,27] It is even argued that interventions to reduce risk, in some cases, can give rise to and legitimize ineffective and unethical practices within psychiatric inpatient care.[30] Other studies noted that adequate assessment of a patient’s risk of adverse events or suicide required enhancing the ability of the staff to engage in an empathic dialogue with the patient.[32,36] Studies also suggested that the ability to encourage a dialogue could be more important than the use of a specific protocol.[32]

There are arguments that general risk does not predict suicide for an individual in the short term, even though the standard of care requires the use of suicide risk assessments[37]—that is, a simple denial of suicidal intent, with no other clinical information, serves no meaning. Likewise, risk categorization of individual patients seems unable to prevent suicide of psychiatric inpatients.[18] Risk assessment could instead benefit from a combination of judgment-based and actuarial-based approaches.[26]

Health Care Practitioners

The studies in this theme highlight the importance of the work environment and clinical skill to enhance patient safety. Patient safety strategies aimed at health care personnel are associated with more generally described outcomes, sometimes without a clear definition of patient safety (Table 1). A health care practitioner’s skill, adequate training, and education are required prerequisites for patient safety.[38,46,51,67] Several areas of competence development are introduced, including how to initiate a therapeutic relationship to reduce suicide,[55] good leadership,[42,45] and team performance.[62,65] Measures such as staff education and sufficient resources,[56] as well as patient centeredness and a deinstitutionalized environment,[48] are also suggested to improve safety. The staff-to-patient ratio is identified as a factor in the frequency of violence, self-harm, and absconding,[63] and optimizing that ratio could contribute to a reduction in seclusion and restraints.[47] However, management with too strong of a focus on policy compliance and other regulatory processes could instead reduce quality engagement with patients.[41] To increase staff performance and to more or less implicitly enhance patient safety, strategies such as mindfulness to reduce stress,[66] emotional skill training,[60] improved morale,[44] and e-learning for staff to reduce patient risks of falls[59] have been tried with varying results. More often, strategies develop into standardized procedures as measures to increase patient safety, including structural communication to reduce suicide,[68] structured handover tools,[54] standardized handoffs,[39] and standardized admission booklets to improve clerking.[58]

Patient Observation

Continuous observation of patients is identified as having a preventive effect on suicide and self-harm,[70,71,75] but observation requires planning and reflection because it is associated with dissatisfaction from both clients and practitioners.[72] In addition, the complexity of formal observation of patients as a patient safety measure is sometimes problematized, and studies question the empirical evidence supporting its use.[73] Observation entails an intrusive and costly intervention but could be modified by engaging the patient in individual activities instead of observing as a passive bystander.[76]

Patient Involvement

Several studies highlight the importance of involving patients to enhance patient safety. To reconnect with the previous theme, continuous observation may be improved by shared decision making, and the measure better supports overall recovery when it occurs in collaboration with patients.[88] Helping patients develop coping and problem-solving skills in a collaborative manner could reduce or prevent self-harm.[82] Safety plans, which are formed together with the patient, directly involve patients in describing symptoms and identifying ways to counter them. Safety plans have proven effective in outpatient care in decreasing hospitalizations and reduction of suicide attempts.[89]

Computerized Methods

There have been attempts to reduce preventable harm by means of technology; examples include electronic health record alerts to remind therapists to structure safety plans for patients at risk[90] and notifications in the patient file system that flag violent patients.[91] The medication process has technological systems to reduce the likelihood of medication errors and to reduce unintended injuries from medications. For example, a barcode system can verify the patient’s identity, although this measure has met with criticism from patients.[92] Computerized drug surveillance programs and programs that search for potential adverse drug reactions are other methods that may improve medication safety.[93]

Admission and Discharge

This theme includes strategies that focus on admission or discharge from inpatient care, times primarily associated with an increased risk of suicide or self-harm. Literature reviews support discharge planning as an effective way to reduce rehospitalization and note that patient-controlled admission can enhance patient autonomy and reduce the total time spent hospitalized.[94,98] Unplanned admissions could be reduced through service-level cooperation to prevent urgent and involuntary admissions. Systematized collaboration with other health care units or organizations may have a positive effect on care transition for patients with multiple health care needs.[96] Irregular discharge increases the risk of suicide, but a planned follow-up visit to an outpatient unit is identified as a strategy to decrease this risk.[99]

Security

The theme of security revolves mainly around the protection from violence in psychiatric care. Staff with verbal de-escalation skills have influence,[105] but policies on ward security more often describe how unsafe items entering the unit could be reduced.[100] A London-wide survey to describe current safety and security measures used on acute psychiatric wards showed a large variation in practice.[102] Banned items, restrictions placed on inpatients, searching of patients, use of security guards, and unique alarm systems varied in the studied wards. Items that were banned in one ward were not banned in others. Even measures such as locking the ward door seemed to reduce absconding to a certain extent. Temporary staff and conflict on wards were identified as factors that could increase absconding[103]; a locked door could have negative effects on patient well-being by inducing frustration and despair, whereas staff anticipate fewer adverse effects from such a security barrier.[104]

DISCUSSION

The findings in this scoping review identify several key messages about patient safety strategies in psychiatry:

Individual adjustments and the skill of the staff to engage in a therapeutic dialogue could be more vital than the use of standardized risk assessment protocols.

Well-trained, educated, and sufficiently appointed staff are fundamental for patient safety. Safety strategies should enhance staff capabilities and guide performance with standards and recommendations.

Continuous observation requires planning to reduce intrusiveness and become more subtle and better understood by both patients and staff.

Mutual agreements between patients and health care practitioners for planned care could reduce adverse events and have recuperative benefits.

Reminders and notification alerts within the electronic patient file system could help reduce adverse drug events.

Patient-controlled admissions and discharge planning with follow-up visits in outpatient units are suggested as a strategy to reduce rehospitalization.

Violence reduction is mainly connected to de-escalation skills from staff. Policies on restrictive measures and banned items vary widely and show no universal success to reduce violence.

In the broader safety science literature, health care is described as a complex system in which functioning relies on the continuous use of adaptive capacities to cope with great system variability.[107] This system is a framework in which safety is seen as the emergent property of interactions and relations between system actors and elements—that is, what the system does, rather than the absence of risk, deviation, noncompliance, and so on (as if the system could be inherently safe to begin with)—not as something the system has.[108] This interpretation challenges the view of preventable harm as a result of simple and linear cause-effect relationships. However, the studied literature most often focuses on identifying risk factors and their potential consequences. The risks are typically and implicitly described in static terms, as something the system/clinic/measure has. This description contrasts with the complexity theory–rooted safety sciences, which see risk instead as dynamic migrations of activities toward the boundary of acceptable performance and measure how these migrations are affected by a number of goal conflicts (e.g., cost efficiency and workload).[109] Risk assessment in psychiatry is a valuable tool for understanding general risk for a patient within psychiatric care, and the execution of the assessment is crucial for a reliable outcome. The findings in this scoping review display some conflicting evidence in the effectiveness of using risk assessments; these assessments not only can support probability judgments but also can become instrumental ends in themselves.

In the theme of healthcare practitioners (Table 1), several articles point toward general improved patient safety or patient safety culture. The concept of preventable harm (adverse event as a product of an “error”) is left out, as the strategy is suggested as a comprehensive measure. However, this review shows that patient safety strategies mainly aim to reduce suicide, self-harm, violence, and falls. The strategy focuses shown in Table 1 include a large diversity of measures, often aimed at reducing variability while increasing standardization. An alternate approach would be to observe the challenges faced by frontline staff or monitor organizational pressure over time (resources, schedule, and patients visits).[110] Furthermore, variability could be perceived as an asset, a prerequisite to tackle unpredictable events, where the adaptive capacities used by frontline staff are reinforced. The included articles support the use of their suggested measure by listing previous incidents or failures. Accident investigations can be influenced by manuals or guidelines that portray a norm of why accidents happen.[111] An assumption of linear causality can lead to a preoccupation with parts and a lack of attention on the system as a whole. There is a risk that patient safety strategies in psychiatry focus on failure and single, regulated measures to reduce unwanted outcomes. Safety measures could instead originate from successful outcomes, in which resilient capabilities are presented. Safety investigators then need to ask; How is work usually done? And what need to be in place (resources, time, competence) in order for the system to work?[9] There are several investigation methods that support analysis of system dynamics under “normal” working conditions, such as Systems Theoretic Accident Modelling and Processes,[112] Functional Resonance Analysis Method,[113] or Resilience Analysis Grid.[114]

Resilient health care is often explained as the ability to adjust to current conditions and anticipate risk before errors and harm occur.[8] Patient safety strategies in psychiatry often overlook performance variability (or resilience) as a factor for understanding preventable harm in health care. If harm in psychiatry is defined as preventable and uses a theory of linear causality, then measures suggested to reduce that harm seem more regulatory and controlling. Patient safety strategies that promote resilience see risk as a changing property within the system and should focus on measures that create foresight[115]—that is, the ability to anticipate an increased risk of preventable harm and the flexibility to answer accordingly. Health care practitioners can create safety under resource and performance pressure because they constantly adapt their activities in response to increased risk.[116] A patient safety strategy could take advantage of this ability.

The psychiatric field is widening the concept of patient safety to include self-harming actions from patients, arguing that these could be prevented if health care had acted differently. The notion of patient safety is connected to preventability,[1] but the term is also used as a general concept for health care improvements. On the other hand, if an adverse event is classified as preventable, the literature shows that arguments of preventability can justify a variety of improvements for the same type of adverse event (Table 1). If adaptive capabilities are incorporated, the improvements would ensure adequate feedback so that adjustments could be made before incidents happen.[107] The improvement efforts should therefore support practitioners in their everyday clinical work but also be adjusted to their variable dynamic conditions. When an adverse event is examined for preventability, the investigation must understand why a certain action made sense, given the circumstances at the specific time for the incident, thus challenging assumptions of deviations and noncompliance.[117] However, detection of increased risk over time is difficult, and patterns, traditions, and norms should be included in the patient safety investigation to better understand dynamic risk migrations.[110] The circumstances for an incident are then challenged, and the perspective of risk could more easily shift from one based on judgment to one based at a system level.

Implementations for Clinical Practice

This scoping review was undertaken to explore patient safety strategies used in psychiatry and how they construct the notion of preventable harm, uncovering the diversity of patient safety strategies in the context of safety theory. To provide safer care and increase patient safety, it is essential to understand that measures come with unintended and potential undesirable consequences that affect other parts of the health care system. This means that the frontline staff needs to be closely supported when new measures are introduced. Therefore, patient safety strategies in psychiatry should incorporate a feedback system so that continuous adjustments can be made from real-time conditions. A measure should have the ability to change over time in accordance with the dynamic properties of the risk in health care.

Limitations and Future Research

The methodology provides an overview of research areas, identifying gaps within the studied field and exploring the literature. Shortcomings of this method involve the quality level of the included articles. However, this methodology was considered the best way to answer the research question, and the sample size of the included articles was large enough to conduct the analysis and detect significant differences. The coded safety strategies are simplified from the included articles, but coding keywords and themes helps organize disparate safety strategies from many studies into categories for study. This scoping review offers an overview of the studied field but does not claim to provide a complete representation of patient safety strategies within the psychiatric field. Additional questions remain about how to better integrate contemporary safety theory of complex systems in the practice of psychiatry.

CONCLUSIONS

This scoping review explores a variety of patient safety strategies to reduce unwanted outcomes in psychiatry. The strategies mainly arise from linear models with measures to prevent previously seen accidents, but there is no universal recipe for success. The strategies that are supported in the literature to achieve safer psychiatry often rely on staff performance, competence, and compliance. Contemporary safety science acknowledges the performance variability of everyday normal work and sees risk as the dynamic migration of these daily activities. The field of psychiatry has not yet included this view of safety in the strategic actions to reduce preventable harm.