Promoting Safety and Equity in Chaos: A Feasibility Study of Peer Support Workers to Support People Who Use Drugs in Emergency Departments
试验速览
- 阶段
- 不适用
- 状态
- 招募中
- 入组人数
- 3,181
- 试验地点
- 1
- 主要终点
- Distribution of randomized Peer shifts by day of week
研究概览
简要总结
Many people with high-risk opioid use visit emergency departments (EDs), facing a high risk of death if they leave before completing care. Peer support workers (Peer) - people with lived experience of substance use - may improve patient comfort, reduce early departures, and potentially lower mortality rates. This study aims to evaluate how a hospital-implemented ED Peer program can enhance patient support, reduce barriers to care, and provide harm-reduction resources in a person-centred and trauma-informed manner. The study will also validate the outcome and obtain preliminary estimates of the benefits of decreasing patients who leave the ED before completing care.
详细描述
The main objectives of the study are to 1) to describe, and evaluate the feasibility and acceptability of a novel ED Peer program implemented by the Vancouver Coastal Health at the Vancouver General Hospital (VGH) ED, 2) to evaluate the feasibility of creating a randomized shift schedule in support of a future definitive randomized controlled trial, 3) to validate outcomes related to patients leaving before completing care using administrative data from the National Ambulatory Care Reporting System (NACRS), and 4) to obtain preliminary estimates of incidence of patients who leave the ED before completing care, to inform sample size calculations for a definitive randomized controlled trial.
The study will take place at Vancouver General Hospital (VGH), a large ED with approximately 97,000 visits per year, serving a diverse urban population. Investigators will evaluate our randomized shift schedule generation, and will describe our intervention and patient population through data within our Peer interaction logs and linkages to regional data. To understand our intervention and patient and provider experiences, investigators will use a mixed methods design to comprehensively collect, analyze, and interpret both qualitative and quantitative data. This approach will provide insights from multiple perspectives, offering a robust understanding of the relationships and potential contradictions within the evidence, thereby enriching the data and deepening our understanding of the findings.
Investigators will evaluate feasibility and acceptability of the Peer program at engaging patients with specific demographic, health history, and risk profiles. Investigators will assess the degree to which patients of different backgrounds access and engage with Peer services, and what characteristics affect patient acceptance of Peer services offered. Investigators will then work with the ED leadership team to target and tailor the program to best meet the needs of patients at highest risk.
This feasibility study will set the foundation for a definitive randomized controlled trial, which will be the first to compare patients with substance use who leave EDs before completing care during shifts with Peer availability versus without, addressing a critical knowledge gap. The definitive trial will engage multiple sites across Canada to rigorously evaluate ED Peer effectiveness, with the potential to scale up an intervention that could improve patient outcomes on a national scale.
研究设计
- 研究类型
- Interventional
- 分配方式
- Randomized
- 干预模型
- Parallel
- 主要目的
- Supportive Care
- 盲法
- None
入排标准
- 年龄范围
- 18 Years 至 —（Adult, Older Adult）
- 性别
- All
- 接受健康志愿者
- 否
入选标准
- •Emergency Department (ED) patient at Vancouver General Hospital (VGH)
- •Age 18 years or older
- •Identified as using substances
- •Able to receive or decline peer support services during the ED visit
排除标准
- •Under 18 years of age
- •Does not report substance use
研究组 & 干预措施
Peer support service
Patients presenting with substance use related complaints during days and times where a Peer is available to provide services, as per an a priori, randomized schedule
干预措施: Peer support service (Behavioral)
No Peer intervention
Patients with substance use presentations who present to EDs on days and times when a Peer is not available, as per an a priori, randomized schedule
结局指标
主要结局
Distribution of randomized Peer shifts by day of week
时间窗: Over the 48-week randomized schedule period approved to meet hospital staffing policy
Assessed by examining the distribution of scheduled Peer shifts across each day of the week over the 48-week randomized schedule period. The schedule was developed in collaboration with human resources and ED operational leadership and approved as meeting hospital employee policy. The randomization approach used block-based scheduling with built-in constraints (e.g., mandatory weekends off). This outcome evaluates whether the resulting schedule achieved an even weekday/weekend distribution. Unit of Measure: Number and percentage of Peer shifts scheduled per day of the week
Proportion of Peer interactions with complete standardized documentation
时间窗: Up to 15 months post-study initiation
Assessed by the proportion of Peer interactions that were fully recorded using the standardized Peer interaction log, with completion defined by filling all required fields. Unit of Measure: Percentage of interactions with complete logs
Identification of substance use-related ED visits for Peer and No Peer shift comparison
时间窗: Up to 15 months post-study initiation
Investigators will obtain regional ED data with all visit information on all ED visits made during our study period. Investigators will use randomized schedule to identify dates and times of Peer shifts and control shifts. Investigators will then assess proportion of patients who left before medically advised among those presenting with substance use-related complaints during times with ED Peers available, compared to times without Peers. Investigators will analyze patients with presenting complaints that denote common complications of substance use (localized swelling/redness, substance withdrawal, and substance misuse/intoxication). This cohort can be identified using administrative data alone, allowing investigators to identify a reliable comparator group when ED Peers are not present. .
Characterization of Patients Receiving the Peer Intervention
时间窗: Baseline. Up to 15 months post-study initiation.
Investigators will describe characteristics of patients who received the Peer intervention using information from interaction logs (including sex and gender) and linked administrative data (e.g., preceding substance use-related encounters; databases described below).
Patients leaving the ED prior to completing care
时间窗: Baseline. Data will be linked and assessed at the conclusion of the feasibility study, up to 15 months post-study initiation.
Investigators will use a linear regression model to compare the shift-level percentage of visits resulting in left before medically advised(primary outcome) during intervention (Peer) and control (no Peer) shifts, among all visits made by patients with our presenting complaints of interest. Patients presenting in the Peer shifts will be included in the analysis irrespective of whether they received the Peer intervention, i.e., an intent-to-treat approach. The model will adjust for shift-level covariates including: day of week, shift type (morning/afternoon/night), and ED volume.
"Treatment effect in the treated" analysis
时间窗: Up to 15 months post-study initiation
Investigators note that the shift-level analysis includes many patients who are not targets for the intervention. This is unavoidable because the absence of screening in the control arm renders it impossible to identify who would be eligible for the intervention, but including non-target patients introduces noise into the analysis and dilutes the observed effect. Hence, investigators will use a mixed-effect logistic regression at the patient level to compare the probability of left before medical advice in patients who presented with a specific complaint (e.g., substance misuse) and received the intervention, to patients in the control arm who presented with the same complaint. Investigators will use medical record numbers collected by Peers as part of the interaction logs to identify people who interacted with a Peer. The model will adjust for age, sex, time of day, day of week, month, ambulance arrival, triage acuity, and presenting complaint.
Distribution of randomized Peer shifts by day of week
时间窗: Over the 48-week randomized schedule period approved to meet hospital staffing policy
Assessed by examining the distribution of scheduled Peer shifts across each day of the week over the 48-week randomized schedule period. The schedule was developed in collaboration with human resources and ED operational leadership and approved as meeting hospital employee policy. The randomization approach used block-based scheduling with built-in constraints (e.g., mandatory weekends off). This outcome evaluates whether the resulting schedule achieved an even weekday/weekend distribution. Unit of Measure: Number and percentage of Peer shifts scheduled per day of the week
Proportion of Peer interactions with complete standardized documentation
时间窗: Up to 15 months post-study initiation
Assessed by the proportion of Peer interactions that were fully recorded using the standardized Peer interaction log, with completion defined by filling all required fields. Unit of Measure: Percentage of interactions with complete logs
Identification of substance use-related ED visits for Peer and No Peer shift comparison
时间窗: Up to 15 months post-study initiation
Investigators will obtain regional ED data with all visit information on all ED visits made during our study period. Investigators will use randomized schedule to identify dates and times of Peer shifts and control shifts. Investigators will then assess proportion of patients who left before medically advised among those presenting with substance use-related complaints during times with ED Peers available, compared to times without Peers. Investigators will analyze patients with presenting complaints that denote common complications of substance use (localized swelling/redness, substance withdrawal, and substance misuse/intoxication). This cohort can be identified using administrative data alone, allowing investigators to identify a reliable comparator group when ED Peers are not present. .
Characterization of Patients Receiving the Peer Intervention
时间窗: Baseline. Up to 15 months post-study initiation.
Investigators will describe characteristics of patients who received the Peer intervention using information from interaction logs (including sex and gender) and linked administrative data (e.g., preceding substance use-related encounters; databases described below).
Patients leaving the ED prior to completing care
时间窗: Baseline. Data will be linked and assessed at the conclusion of the feasibility study, up to 15 months post-study initiation.
Investigators will use a linear regression model to compare the shift-level percentage of visits resulting in left before medically advised(primary outcome) during intervention (Peer) and control (no Peer) shifts, among all visits made by patients with our presenting complaints of interest. Patients presenting in the Peer shifts will be included in the analysis irrespective of whether they received the Peer intervention, i.e., an intent-to-treat approach. The model will adjust for shift-level covariates including: day of week, shift type (morning/afternoon/night), and ED volume.
"Treatment effect in the treated" analysis
时间窗: Up to 15 months post-study initiation
Investigators note that the shift-level analysis includes many patients who are not targets for the intervention. This is unavoidable because the absence of screening in the control arm renders it impossible to identify who would be eligible for the intervention, but including non-target patients introduces noise into the analysis and dilutes the observed effect. Hence, investigators will use a mixed-effect logistic regression at the patient level to compare the probability of left before medical advice in patients who presented with a specific complaint (e.g., substance misuse) and received the intervention, to patients in the control arm who presented with the same complaint. Investigators will use medical record numbers collected by Peers as part of the interaction logs to identify people who interacted with a Peer. The model will adjust for age, sex, time of day, day of week, month, ambulance arrival, triage acuity, and presenting complaint.
次要结局
- Patient-reported acceptability of the Peer program using the Peer patient survey(Up to 15 months post-study initiation)
- Thematic summary of Indigenous patients' experiences with the Peer program based on storytelling and sharing circles(Within 6 months after the patient's ED visit involving Peer interaction)
- Thematic summary of healthcare providers' perceptions of the Peer program assessed through focus groups(From 3 to 15 months after program implementation)
研究者
Jessica Moe
Assistant Professor UBC Department of Emergency Medicine
University of British Columbia