THRESHOLDS - June 16, 2023

Agenda

Thresholds, led by Laura Sikstrom

• Thresholds define the boundary between personal space and invasion. Known as a “tipping point,” they act as a narrow balance between spaces and can be easily transgressed.
• There can be signs determining thresholds, whether they are physical (e.g., locked doors, restricted areas) or social rules that guide behaviour.
• Patients in the ED engage in “boundary flirting” at times to test threshold limits and explore the dynamics of their interactions, which can be fun and dangerous for them (e.g., staring at others, sneaking alcohol in).
• These incidents may challenge the existing power structures in place, typically pointing to de-escalation practices to ensure safety rather than overall prevention. This highlights one of the underlying issues at hand – the management of interpersonal relationships between staff and patients in the ED.

Tactical Infrastructure

“It looks like a margin, both spatial and social, but… it acts as the core… [This is the place] where political authority is perpetually remade, legitimacy renewed”
- Threshold: Emergency Responders on the US-Mexico Border by Ieva Jusionyte (2018)

• This quote draws attention to the concept of thresholds surrounding the U.S./Mexico border. It implies that thresholds are critical points where power dynamics are continually renegotiated.
• At the U.S./Mexico border, there has been a strong emphasis on design elements such as fences, roads, bridges, grates, and observation zones that help to maintain control.
• The ED exhibits parallels to the border because it also holds measures to manage control and ensure the safety of both patients and staff. While the goals and context may differ between border control and psychiatric care, the underlying theme of control is present in both.

“Ankle Alley” – U.S.-Mexico Border

• Ankle Alley was the U.S. government’s response to illegal border crossing attempts. Policies were implemented to funnel people through the Sonoran desert, leading to very specific injuries (e.g., mortality due to dehybration, wild animals, heat exhaustion and so on).
• Additional height was also added to the top of the border, which was intended to deter those trying to cross. This resulted in broken ankles when they fell to the other side. These deliberate actions highlight the control-oriented approach of the government.

“The Land of Open Graves” by Jason De Leon (2015)

• De Leon discusses the concept of the “Hybrid Collectif,” suggesting that agency emerges through the interaction of various elements, both human and non-human, such as heat, pathogens, the Sonoran Desert, and wild animals.
• This perspective raises questions about how these non-human actants influence human interpersonal interactions and the nature of those interactions.
• Just as deliberate policies and infrastructure design impact the outcomes and interactions at the U.S.-Mexico border, in the context of the ED, there may be environmental design elements intended to manage and control aggressive and violent behaviour. Exploring the influence of non-human actants in both scenarios can shed light on the dynamics and consequences that arise in interpersonal interactions.

Predictions of Incidents in the ED led by Marta Maslej and Chris Dharma

• Past analyses revealed variations in DASA scores based on race, gender, diagnosis, housing status, and mode of admission (police, self, family)
• This suggests that perceptions of risk are higher for certain patient groups, such as black and middle eastern patients.
• There is high concern about false positives and false negatives recorded during the DASA assessment. False positives are dangerous for patients, while false negatives pose risks for staff, creating a tradeoff.
• Staff safety may influence the tendency to issue false positives to avoid false negatives.
• The main question lies in finding the optimal tradeoff between false positives and false negatives in machine learning. The ideal threshold must be determined.
• Flaws in risk classifications can lead to unfair treatment of patients and disregard for individuals’ circumstances. Some patients are classified as high-risk and are treated unfairly, but don’t have any incidents. There are also patients who do experience incidents but may also face unfair treatment because of their risk classification.

Considerations for Analysis

• Chris and Marta proposed to remove patients with incidents occurring in the ED prior to admission or their first record of DASA score.
• They also proposed adjusting the number of visits in the past year to define new patients (e.g., examining the number of visits from May 2017 to May 2018. After 1 year, the person will be considered a new patient).
• Further factors to be considered in the analysis include removing the most frequent patients known to clinicians. However, this becomes complex when accounting for social stereotypes and defining thresholds for “frequent flyers” – this refers to individuals who visit the ED frequently.
• Chris and Marta hypothesize that DASA assessment duration and patient-staff dynamics may influence risk assessment. As the clinician gets to know the patient better, the likelihood of using demographics and mode of admission as an anchor for risk assessment may decrease.
• Laura and Peter will ask clinicians about this dynamic, but it may be subjective per person.
• If this is the case, it may also make sense to only consider the first-day DASA scores – early stages of assessment are crucial to understanding dynamics.
• Chris and Marta raised the possibility of conducting other types of analyses – specifically exploring the relationship between incidents and the specific day they occur. Marta also suggested creating a covariate to control for clinic type and account for potential variations across different clinics.
• Team is considering race x gender and race x diagnosis analyses for publication, with a focus on race x mode of admission.
• Future analysis: clustering DASA scores together in latent classes to identify which class of aggression is more likely to lead to a false positive.

Environmental Design Literature Review led by Patrycja Szkudlarek and Sitra Adill Mohammed

Environmental Design Perspective

• Much of past psychiatric research aimed at predicting and reducing aggression in environmental design has emphasized examining factors such as patient diagnoses and improvements to staff training.
• These types of approaches have often limited environmental design research in psychiatry to increased security measures (e.g., locks, cameras).
• The literature examined often described a lack of a plausible design theory for effectively reducing overall aggression due to the difficulty of identifying a causal relationship between each specific environmental design feature and aggression/violence.
• Rather than do this, some studies (Harpoth et al., 2022; Ulrich et al., 2018) compared patient records and admissions before and after moving patients to a new, purpose-built psychiatric ward, and compared the overall differences in design features between hospitals.
• Some studies (Jenkins et al., 2015; Rogerson et al., 2021) also clustered design features of the hospitals into dimensions and ran analyses against changes in aggression/violence in patient records on those variables. Studies that did this, had to hypothesize what could be the exact reasoning behind either heightened or reduce aggression in their results.
• Due to limitations such as these, there was a complex mix of findings, making it difficult to pinpoint causality. Despite this, all authors argued that environmental design has an overall impact on aggression, whether their findings were concrete or not.
• The studies each had similarities in design features that they tested. Similarities such as the use of single rooms, private bathrooms, and large rooms were a common focus for reducing aggression.
• In future studies, an emphasis on qualitative measures must be put in place. Without a further understanding of the social interactions between staff and service users, the findings will remain less concrete.

Psychiatric Perspective

• Studies (Sui et al., 2023; Curtis et al., 2013) found that security degenerates into mere technical safety, compromising therapeutic care. Placing too much emphasis on security may result in the removal of design features that could be beneficial to the well-being of patients.
• Curtis et al., 2013, discussed the panopticon theory, developed by Michael Foucault. It theorizes that placing an emphasis on technical safety can lead to a loss of individual autonomy and fosters a sense of constant control.
• Compromising some aspects of technical safety management may be needed to prioritize therapeutic care. A fundamental part of risk is the human and social component.
• The first step in mitigating aggression is identifying its origin. The origin should inform differential prevention and de-escalation strategies (e.g., individual service user factors, social interactions, organizational factors).
• Studies suggested that the quality of social interactions between staff and patients was an important element of space that promoted mental health. When participants described spaces that were supportive of recovery, individuals in those spaces were people who listened and cared.
• The next steps for future studies involve examining staff-patient interactions and identifying design features that facilitate these interactions, as well as investigating hierarchies between staff and patients.

Discussion/Action Items

• Laura and Peter will meet with Juveria to discuss an outline for publication regarding thresholds and managing to improve them through staff and patient interactions.
• Peter will lead the writing, and Patrycja and Sitra will contribute to the background writing.
• Marta to summarize statistical analyses.
• Next meeting scheduled for July 12^th, 10AM – 1PM
• Valeria and Sitra will present their respective projects for the summer along with their progress.
• Igor will present his literature review for the dashboard project.
• Meeting theme: TBD

Word Frequencies

• Chris presented comparisons of word frequencies from records between different groups (i.e., gender differences, racial differences).
• Possible imbalances in sample size were discussed due to the nature of the word frequencies.
• Chris to further refine word frequencies presented to remove incidents not related to patients and link these to demographic data.
• Marta will investigate reviewing weighted word frequencies by incident type and review words through entity recognition to yield further context.

Topic Model

• Marta presented findings on trained model on categorized topics related to objects or spaces in the ED (i.e., client spaces, staff spaces).
• Invasion of client space came up often (specifically in the washroom category), as well as clients invading staff space.
• Laura will further explore sociolinguistics in relation to sharing intimate spaces and discrimination based on the findings of the analysis.

Space and Risk Assessment in Ethnography

• Laura and Peter led the discussion on types of spaces, such as physical, metaphorical, and in-between spaces, and their thresholds at the ED. They shared their observations about risk regarding spaces in relation to the history of psychiatric policies and care.
• There are thresholds in these spaces that define boundaries and where risk takes place. However, thresholds are often transgressed in each.
• In the current ED space, nurses and PAs spend the most time in proximity to risk.
• Laura and Peter will recreate the whiteboard/shadow documents from the ED care station for visualization.
• Team to further investigate gender-responsive de-escalation and racialized care work in interviews.
• Works to be considered: “Emptying Beds: The Work of an Emergency Psychiatric Unit” by Lorna A. Rhodes, NCR: Not Criminally Responsible, directed by John Kastner.

General Planning

• Laura and Sitra to touch base on REB flyers.
• Team to begin working on patient advisory group while awaiting REB approval.
• June meeting theme: Shadow work