Introduction

Approximately 15.6 million people inject drugs globally and are at elevated risk of bloodborne infection, chronic morbidity, and premature mortality [1, 2]. Robust global evidence identifies needle and syringe programs (NSPs) as a key intervention for reducing these negative outcomes [3, 4]. NSPs provide people who inject drugs (PWID) with access to sterile drug consumption supplies and other supports, and have been implemented in over 80 countries and within various contexts (e.g., fixed and mobile sites, pharmacies, dispensing machines, and prisons) [5]. Meta-analyses and systematic reviews indicate that PWID who regularly attend an NSP have lower rates of injection-related risk behaviours and HIV seroconversion [6, 7]. NSPs also promote communication and trust between PWID and health care providers and facilitate uptake into health services [8, 9]. These benefits appear most impactful when an NSP is easily accessible and located close to where injection drug use occurs [3, 9, 10].

Despite abundant evidence and widespread implementation in community and outpatient settings, NSPs and other harm reduction interventions are not routinely available in hospitals [11, 12]. This is problematic because PWID experience high rates of hospitalization [11, 13] and cohort and survey data suggest that 30%-50% of hospitalized PWID continue to consume drugs while admitted [14–18]. In-hospital drug use is driven by several factors. Most physicians receive inadequate training on managing substance use disorders in their primary specialty, and access to inpatient addiction medicine specialists remains limited [19]. This is further compounded by limited medication options for people experiencing problematic stimulant use. Moreover, PWID frequently report insufficient pain and withdrawal management [20, 21], which can lead to discomfort and distress and a need to self-manage symptoms through drug use. Even when specialized care is available, a subset of PWID may continue to inject. Qualitative studies with PWID suggest that ongoing drug use may result from anxiety, stress, boredom, and stigma experienced during acute care episodes [20, 21].

For people who continue to inject drugs while hospitalized, hospitals can become a high-risk environment where they encounter more difficulty accessing NSP than in community settings [21]. Formal and informal bans on injecting in hospital result in constrained access to sterile injection supplies, which increases the risk of syringe sharing or reuse, and infection [15, 20, 22, 23]. Such restrictions on drug use can exacerbate well-documented mutual mistrust and poor communication between PWID and hospital staff [19, 24] and increase the risk of patient-initiated hospital discharge, involuntary discharge, unplanned readmission, and death [15, 17, 25–27]. Similar to impacts seen in community settings, inpatient NSPs may improve patient outcomes by facilitating a reduction of injection-related infections, supporting patients to complete treatment, and engendering communication and trust between patients and staff [9, 10].

In the United States, empirical research describes an NSP located adjacent to a hospital [28, 29] and the distribution of supplies to hospitalized patients upon discharge [30]. In Canada, the distribution of supplies has been noted to occur in one acute care facility [31], within specialized sub-acute care facilities [32, 33], and sporadically within emergency departments [34]. However, data on the perspectives of patients who accepted NSP supplies are lacking. To address this gap, we undertook a mixed-method study to assess the implementation of an NSP integrated within an urban, 850-bed hospital in Western Canada.

Methods

We adopted a focused ethnographic design, which is often applied in research within healthcare settings to answer specific, time-limited research questions in response to a particular problem [37, 38]. In data collection, we used ethnographic interview techniques to probe context and focus on the timely understanding of emic perspectives of participants [39]. Compared to traditional ethnography, focused ethnography can limit or omit participant observation to generate rapid data [37, 38]. A member of our research team (HB) conducted semi-structured interviews with predetermined open-ended questions and probes to rapidly elicit detailed information [39]. The study protocol was developed in consultation with our research team’s community advisory group of people with lived experience of drug use and hospitalization, over half of whom identified as Indigenous; Indigenous Peoples are overrepresented amongst PWID in the study setting.

Eligible participants included admitted inpatients seen by the AMCT who reported recent injection drug use. AMCT staff approached eligible participants and asked about potential participation in the research project. HB approached those interested, obtained written informed consent, and conducted audio-recorded interviews in a location of the participants’ choosing. HB ensured participants’ names and other potentially identifiable information were kept private from AMCT staff. Interviews were conducted between April 20, 2017, and March 7, 2018, and averaged 51 minutes. Participants received $20 CAD for their time and expertise. The audio recordings were professionally transcribed and the transcripts deidentified by HB prior to analysis. We used ATLAS.TI 8 to manage the transcript files and facilitate analysis.

We used latent content analysis to identify and organize the meaning and significance of the transcripts [40]. HB identified clusters of words that related to the same focal meaning and then highlighted, labeled, and sorted those clusters into categories. Data collection and analysis took place iteratively, allowing for initial interpretations and then refinement as interviews continued. This process continued until all transcripts had been analyzed [41]. To ensure rigour, a second member of the research team coded 20% of the transcripts to ensure concordance of interpretation [42]. Additionally, the research team sought feedback from the community advisory group on the validity of the analysis, main findings, and recommendations.

Results

We completed 25 interviews with patients eligible to access the NSP. Of the 25 participants, 21 accepted supplies prior to their interview. The four participants who did not accept supplies reported that they were not actively injecting in hospital. Participants’ motivations to inject were primarily to manage pain and withdrawal symptoms or due to habit or cravings. For details on the characteristics of the study participants, see Table 1.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Participant information.

https://doi.org/10.1371/journal.pone.0297584.t001

Discussion

Empirical research conducted in North America indicates hospital settings may increase risks of harm for PWID, who often continue to use drugs and may encounter stigma, experience poor pain and withdrawal management, and are at increased risk of negative outcomes [16, 17, 20, 26, 43]. Recent scholarship proposes the implementation of hospital policies and interventions that accommodate active drug use, as opposed to explicit or implicit prohibition, to improve outcomes for hospitalized PWID [12, 44–46]. Results from this study support this recommendation. Participants who accessed NSP supplies described the NSP as a trusted source of safer drug-use education and related support, reported a more consistent use of sterile supplies, and expressed greater ease receiving treatment and remaining hospitalized.

However, social forces commonly found within acute care hospitals in North America (e.g., stigma) [47] and risk management policies that influence a healthcare provider’s decision to acknowledge or accommodate active drug use [48] may inhibit or constrain patient-centred harm reduction interventions such as an NSP. These forces were evident in this study. Despite a health authority-wide policy endorsing a harm reduction approach in the care of patients who use substances [49], anticipated stigma or punishment from staff impacted participants’ willingness to access the NSP and accept supplies.

The primary factor that impacted participants’ decisions to access the NSP was trust in AMCT staff. This suggests that all staff should make a concerted effort to build trust with patients by integrating trauma-informed and culturally-safe practices into patient care [50, 51]. One practice hospital staff should adopt is to trust patients’ perceptions of pain and withdrawal symptoms, and ensure immediate access to effective medical management [21, 24]. Hospital administrators should also provide guidance on how staff can recognize the social forces that produce stigma and how they can modify their actions when interacting with PWID (e.g., use non-stigmatizing methods to screen and monitor for drug use) [52, 53]. Hiring staff with lived experience of drug use would facilitate this process [54].

Further modifications to hospital policy may be necessary to facilitate NSP uptake. After accepting supplies, participants described feeling uneasy because they were uncertain of the subsequent ramifications of accepting supplies. To quell patients’ apprehension, hospital administrators may need to establish consistent hospital-wide policies delineating the rights, roles and responsibilities for patients who accept NSP supplies, and of staff caring for those patients. This should include a clear description of any rules or polices governing where (e.g., on or off hospital property) and when (e.g., how soon prior or post medication dispensation) patients should inject if they continue to do so. Hospital policies should also consider providing guidance on medication administration designed to minimize the risks of medications being taken in a non-prescribed way (e.g., the use of liquid oral formulations as is the current practice of the AMCT). Jurisdictions in Canada currently vary in their approach to prescribing oral opioids intended for injection use; however, this type of prescribing was not employed by the AMCT or used in this jurisdiction during the study. Staff may also consider providing instructions regarding how patients may store their drugs [55]. Staff should regularly follow up with patients to ensure they understand the policies and that they are satisfied with their hospital experience.

Additionally, hospitals may choose to establish an inpatient supervised consumption service [12, 56, 57], which several participants described as potentially beneficial. This would address the inconsistencies apparent in the NSP model described within this study where patients were provided with sterile supplies but told they were not allowed to use drugs on site. Supervised consumption services provide a safe and welcoming space where people can use drugs with sterile supplies under the supervision of trained staff without fear of legal prosecution [58, 59]. In part due to the challenges described by participants within this study, the AMCT established a supervised consumption service within the hospital in 2018 [56]. Patients who accept NSP supplies are now encouraged to attend the SCS to consume drugs; drug consumption is still not permitted anywhere else on hospital property. A recent evaluation conducted by some of the authors of this paper found that supervised consumption service participants reported benefits and perceived the space as safer than where they normally would use in hospital [60]. They nevertheless also reported concerns regarding perceived negative social forces within the broader hospital environment and potential impacts on their stay (e.g., changes to their medications). Some refrained from utilizing the service due to these concerns [60]. Further strategies are needed to change hospital culture and support PWID.

Integrating harm reduction interventions such as NSP or supervised consumption services into hospitals where wider social and political forces such as stigma and drug criminalization still exist may not sufficiently reduce harms for hospitalized PWID. Continual efforts must be made to counteract stigmatizing behaviours amongst staff, promote trust and improved communication pathways, and provide patients access to social support and harm reduction services both in-hospital and post-discharge. However, broader societal changes, such as drug decriminalization, may ultimately be necessary.

Conclusions

In this study, we explored the integration of an NSP into a large, tertiary acute care hospital. Patients described access to sterile injection supplies while hospitalized as beneficial but described occasional negative experiences. Hospitals aiming to improve care for PWID should consider developing hospital policies and interventions that acknowledge the reality of ongoing drug use for some patients and take steps to reduce associated harms. More research is needed on ways to modify the hospital environment to facilitate NSP access and to identify additional acute care interventions that can reduce the health and social inequities experienced by PWID.

Acknowledgments

We thank the study participants for sharing their time and experiences with us. Thank you to Professor Tania Bubela for providing guidance to the authors during the study and feedback on early versions of this analysis and to Brynn Kosteniuk for coding a portion of the transcripts. We would also like to acknowledge members of our team’s community advisory group for sharing their valuable insights and advice related to this research project.