Background

People who inject drugs (PWID) are at increased risk for severe injection-related infections (SIRIs), including endocarditis, bacteremia, osteomyelitis, and skin and soft tissue infections (SSTIs) [1,2,3]. SIRIs are the most common reason for hospitalization among PWID [3, 4] and have had increasing incidence in the United States [5,6,7,8]. These hospitalizations can require complex surgical and medical interventions and are associated with high healthcare expenditures [4, 5, 9]. PWID also often have traumatizing hospital experiences, receiving inadequate treatment for their symptoms, especially pain and withdrawal [10,11,12]. There are many barriers to care for PWID, including difficulty navigating complex and fragmented systems of care and clinicians that are ill-equipped to treat—or do not prioritize treating substance use disorders (SUDs) [13,14,15]. About 1 in 6 PWID hospitalized with a SIRI leave the hospital early under “patient-direct discharge”, also known as “against medical advice” (AMA) [3, 14, 16,17,18], leading to incomplete antibiotic treatments and readmission. There is a critical need for novel approaches to help this vulnerable population [19].

A multidisciplinary approach that incorporates infectious disease (ID) and addiction care in the hospital and after discharge has the potential to mitigate many of the barriers to successful individual and health system outcomes for patients experiencing SIRIs [20]. Some existing models of integrated ID/SUD teams include ID specialists, addiction medicine specialists, psychiatrists, and surgeons and may incorporate pharmacotherapy, behavioral treatments, harm reduction and post discharge follow-up for patients [21,22,23,24,25]. Previous research has demonstrated that integration of evidence-based addiction treatment—such as medications for opioid use disorder (MOUD)—is associated with fewer patient-directed discharges [26, 27], lower readmission rates [21, 28], higher rates of antimicrobial therapy completion [26, 29], higher post-discharge SUD treatment engagement [30], and reduced substance use at 30 days [31]. Additionally, when care teams integrate extensive outpatient support and follow-up, SIRI patients who leave under patient-directed discharge on oral antibiotics can achieve equivalent 90-day readmission rates to SIRI patients receiving inpatient IV antibiotics [21]. Despite these established evidence-based practices, there is a need to develop programs designed to implement such practices with fidelity in diverse care environments.

In 2017, Miami-Dade County experienced an estimated 1100 hospitalizations related to complications of injection drug use with approximately 400 of those hospitalizations in our public safety net hospital, Jackson Memorial Hospital (JMH) [18, 32]. Using administrative data and diagnostic codes, the 90-day readmission rate for IDU-associated conditions was estimated to be nearly 50% [32]. Based on the local burden of SIRIs and lack of inpatient addiction medicine consult services, we developed an integrated infectious disease/SUD treatment intervention in 2020 called the Jackson SIRI team. Using Englander and colleagues’ taxonomy of hospital-based addiction care, the SIRI team is a hospital-based opioid treatment (HBOT) program, but additionally includes substantial post-hospital care [25]. The SIRI team intervention provides integrated infectious disease and SUD treatment across the healthcare continuum, starting from the inpatient setting and continuing for 90-days post-hospital discharge. The team uses a harm reduction approach, provides intensive care coordination, focuses on low-barrier access MOUD, and utilizes a variety of infectious disease treatment approaches to suit each patient, such as oral antibiotics and long-acting lipoglycopeptide antibiotics. In the hospital, the team serves as a medical consult service and provides additional services focused on securing appropriate discharge plans and coordinating complex medical, surgical, and socio-behavioral obstacles to care. Based on dual expertise in ID and addiction medicine, the SIRI team is well suited to guide the treatment plan for patients’ infection, especially regarding questions surrounding outpatient parenteral antimicrobial therapy. After discharge, the team maintains frequent contact with patients and continues to provide infectious disease/SUD medical care and case management. The team consists of three physicians with expertise in infectious disease and addiction medicine, and an ID nurse practitioner. The team also works closely with a pain/SUD pharmacist and the affiliated SSP’s team of peer counselors and social workers. Details of the development and team function have been published previously [32].

This qualitative study aims to examine the perceived barriers and facilitators to implementation and sustainability of the SIRI team intervention in one healthcare setting, looking at both patient and clinician-level perspectives in the first 8 months of the team’s clinical services. The goal of this work is to use the tools of implementation science to document the implementation context of the SIRI team intervention to improve its function, discover core components, and understand potential adaptations for other healthcare systems.

Methods

Study design and procedures

We conducted key informant semi-structured interviews with inpatient clinicians and patients who had been hospitalized at JMH with a SIRI to provide contextual knowledge regarding the barriers and facilitators impacting the implementation and sustainment of the SIRI team intervention. The study was approved by the University of Miami Institutional Review Board (#20200962).

Study setting and participants

This study was conducted at an academic medical center that includes JMH and the nearby affiliated IDEA Miami syringe services program (SSP) housed within the University of Miami Miller School of Medicine, Division of Infectious Diseases. A purposive sampling method was used to recruit clinicians with experience working with PWID and with the SIRI team, as well as a mix of PWID hospitalized for SIRIs both pre- and post-SIRI team implementation. We included patients without exposure to the SIRI team to gain insights into challenges faced by PWID with SIRIs that may have been avoided by SIRI team intervention. Participants were identified, contacted by the study PI through email, telephone, or in-person and asked to participate in an interview. Targets for clinician interviews were hospital administrators, physicians, social workers, and nurses based on proximity to SIRI team implementation. Patients were recruited either at the IDEA Miami SSP or post-discharge from JMH. Interviews were conducted within the first 8 months after the SIRI team began providing services. All clinician participants in this study had exposure to the SIRI team during this early pilot phase. All participants received compensation of 50 USD for their participation.

Semi-structured interviews

A semi-structured interview guide was created using the Consolidated Framework for Implementation Research (CFIR) (Appendix). This validated, conceptual framework can be used to explore the determinants of how evidence-based interventions can be implemented into real-world systems [33]. The CFIR includes 5 domains (intervention characteristics, outer setting, inner setting, characteristics of individuals, and process) with 39 constructs [33]. The research team assessed all domains and constructs to determine which were most salient for our implementation evaluation effort. We focused on four of the five CFIR domains most relevant to this study. A description of the four domains and constructs operationalized is provided in Table 1. The interview guide was created using open-ended questions in which perceived barriers and facilitators to implementation and sustainability of the SIRI team intervention were explored.

Table 1 Domains and constructs of the Consolidated Framework for Implementation Research operationalized in current study
Full size table

Data collection and analysis

Verbal informed consent was obtained from all clinicians and patients before participating in a 40-min semi-structured interview. A research team member (B.H.) conducted the interviews face-to-face at the IDEA Miami SSP, through videoconferencing software, or by telephone. The interviews were audio-recorded and transcribed by a third-party transcription service. Transcribed interviews were analyzed using both deductive and inductive methods [34]. An a priori codebook was created using the CFIR constructs adapted to the implementation of the SIRI team. Additional codes for patient interviews were created using a general inductive approach, allowing findings to emerge from the most frequent and dominant codes and themes encountered throughout the analysis [35]. Additional patient codes were listed under the CFIR Patient Needs and Resources construct. The authors met regularly to discuss emerging themes and categorization using CFIR. Four study team members coded transcripts using Dedoose (Version 8.2.14, 2020). All four members coded a subset of interviews to ensure reliability of code application. Interviews were double coded, and the study principal investigator (D.P.S.) reconciled any discrepancies between coders to create the final set of coded transcripts for content analysis. Themes were categorized and reported based on CFIR constructs.

Results

Fifteen interviews were completed with eight clinicians (“clinician participants”) and seven patients (“patient participants”). Three of the patient participants had been patients of the SIRI team, while the other four had been hospitalized for a SIRI at JMH before the SIRI team’s implementation. Patient and clinician demographics and descriptive statistics are presented in Table 2. Barriers and facilitators to the implementation of the SIRI team intervention were categorized using CFIR constructs and are described below. Table 3 summarizes themes that emerged across interviews with patients and clinicians and include additional representative quotations.

Table 2 Demographics of Qualitative Interview Participants
Full size table
Table 3 Summary of themes from Clinician and Patient Qualitative Interviews about the SIRI team
Full size table

CFIR domain 1: intervention characteristics

Facilitators

Patients and clinicians acknowledged that the SIRI team’s provision of MOUD is a core component of the intervention’s effectiveness and crucial in treating infectious complications of SUD (Evidence Strength and Quality). Patients noted that relative to the prior standard of care, having the same team provide guidance on infectious disease management, withdrawal management, and assist with pain control, is a significant improvement in care quality (Relative Advantage). The comprehensiveness of the SIRI team—addressing infection, SUD, pain, cravings, anxiety, and social barriers to care—is advantageous compared to the standard fragmented care. One clinician described this advantage stating:

“I think the ability of this team to bring in other providers together so that we have a coherent treatment plan, not piecemeal by different providers and different consultants, I think is also important.”

Clinicians highlighted the SIRI team’s expertise in navigating often complex decisions about intravenous versus oral antibiotics and helping decide the best location (hospital, shelter, etc.) to complete antibiotic therapy. Participants also noted that the SIRI team demonstrated a unique level of compassion and advocacy for an often-mistreated community (Relative Advantage). A patient shared:

“When I was having a hard time—[SIRI team physician] always told me I could call him. I did, and he talked to me when I was having a bad day 1 day and stuff. He followed through with what he said he would do for me. He’s always let me know about things coming up that could help me- like with the apartments, the grant things for living situations and stuff.”

Patients also appreciated the SIRI team’s lack of judgment regarding their substance use:

“[The SIRI team] wanted to make sure I had a safe place to go when I left and that if I did go back to using that [they] wouldn’t frown upon that, to make sure I came and got clean needles and everything but to hang in there. [They] kept me on methadone and everything.” (Patient Needs & Resources).

Barriers

One barrier to successful implementation and sustainment of the SIRI team that emerged was the complexity of caring for PWID with SIRIs as well as the complexity of the intervention itself. Clinicians stated that having an additional treatment team beyond infectious disease and psychiatry could lead to confusion about clinician responsibilities (Complexity), even though they acknowledged the importance of having this specialized team. A clinician noted a case where the SIRI team recommendations contradicted the recommendations from the general ID consult team: “Sometimes it can feel like there's too many cooks in the kitchen.” Others expressed concerns that there are already many delays in getting patients needed care and that the wait time for a SIRI team consult could prolong the length of stay. Clinician participants predicted administrators would be resistant to funding another clinical service and that the team’s interventions could lead to longer lengths of stay and increased hospital costs (Cost).

CFIR domain 2: outer setting

Facilitators

In the outer setting, patients and clinicians highlighted the SIRI team’s efforts to help patients meet needs that are often unmet in healthcare, such as obtaining stable housing, securing health insurance and financial support for MOUD, and arranging follow-up medical care (Patient Needs & Resources, Cosmopolitanism). A clinician highlighted the importance of this facilitator:

“That’s gonna be important to a mental health team, individual therapy, group therapy, psychiatry, and the social worker that can link you to all these resources specifically to you so that you can achieve the goals that you wanna achieve whatever they are at the hands of a team that is willing to provide you the tools to do so.”

Clinicians felt the SIRI team had created strong relationships outside of the institution to advocate for patients and leverage local resources that had previously not been utilized by the hospital (Cosmopolitanism). A clinician noted:

“It just seems to be that patients that are seen by that team get into the right places more often, I think. I think it’s because they’re knowledgeable about community resources in ways that I think a lot of the other providers aren’t, because this is their specialty.”

Interviews with patients revealed that patients appreciated the SIRI team’s commitment to maintaining communication post-discharge, particularly given the stigma and judgment often experienced by patients seeking care for SUD and SIRIs (Patient Needs & Resources). One patient described meeting a member of the SIRI team, stating that:

“When [the SIRI team physician] introduced himself to me, I thought he was just like any other doctor. He was very nice. He got me on the Suboxone. I didn’t know I was getting into a program to where [they] would check up on me after I leave the hospital. When I was in rehab and [they] contacted me, I was very, very surprised. What a gift.”

Barriers

Clinician participants felt that the lack of resources available for patients with SIRIs could limit the effectiveness of the SIRI team intervention. Despite the SIRI team’s efforts to identify resources for their patients, the intervention’s impact is limited by the availability of substantive resources, such as the number of beds in housing facilities for people experiencing homelessness and SUD or funds available to financially support patients (Patient Needs & Resources). Clinician participants also cited a lack of desire for change from patients as a barrier, “…it just depends on the stage of readiness for the patient… if they’re ready to get off drugs or not.” (Patient Needs & Resources). Furthermore, without overarching policies designed to increase access to health insurance, safety net resources, behavioral health care, and financial assistance, the SIRI team may face multiple barriers to providing for patients (External Policy & Incentives). Clinician participants saw the policies and financial forces that guide healthcare as having a negative impact on the SIRI team’s sustainability. The focus on reducing the length of stay in the short term might work against the team’s securing a stable discharge plan that would reduce longer-term healthcare costs.

CFIR domain 3: inner setting

Facilitators

Clinician participants described the hospital as an institution that has historically been innovative, open-minded, and willing to make longer-term investments for the good of underserved populations. A clinician participant cited prior investment in an “early discharge program” that had been successful in allowing uninsured persons experiencing homelessness to receive IV antibiotics in a medical respite facility. Regarding the SIRI team itself, patients and clinicians were impressed with the ease of communication with the team and highlighted this low barrier access to the team as a major strength (Networks & Communications). A nurse who worked with the SIRI team shared:

“You’re trying to find the team, you cannot find the team, and that takes the whole day, and that’s an extra day for the patient inside the system. I’m always lucky when I see [SIRI team] is on the case because I know I have [their] phone number. I can text [them], call [them]. [They] will pick up.”

The SIRI team’s clinical documentation was also praised as being highly effective, educational, and serving to coordinate care and improve efficiency. Additionally, clinicians thought the SIRI team helped improve culture by modelling appropriate language, respect, and patient-centeredness for PWID.

Barriers

Institutional barriers to SIRI team implementation included the perception that there is a lack of investment in helping PWID. Clinicians discussed how the healthcare system, such as scheduling financial assessments, are difficult for persons experiencing homelessness. A clinician explained:

“it seems like we have resources for other things when we need them, but this [PWID, patients experiencing homelessness] isn't a priority.” (Implementation Climate).

Clinicians described frustration that focus on length of stay might adversely affect SIRI team patients who need more time to secure optimal discharge plans.

“[Hospitals are] like, ‘We can’t keep somebody for two days just waiting for rehab. Send them back on to the streets. Call them in 2 days when their bed is ready.’ That kind of stuff doesn’t work.”

The other main barrier identified was lack of clinician awareness of the SIRI team’s implementation due to lack of effective advertisement to stakeholders (Networks & Communications) which limited the reach of the SIRI team intervention. Clinicians also suggested there was not sufficient education to stakeholders in the hospital about what the SIRI team intervention is and when to consult this service (Access to Knowledge & Information; Networks & Communications). A clinician described this barrier stating that:

“If you do not put together a good education plan and training plan, and the staff does not understand the why behind what it is you’re asking them to do, then that could be a huge barrier, and you won’t get their support.”

Another identified barrier was the emotional toll on clinicians when caring for patients whose needs are not traditionally met in hospitals (Patient Needs & Resources). One clinician explained:

“They require a lot of time from nurses. They require a lot of services from doctors at a time when people are very overburdened with work... That might also be a factor with them going without a—leaving AMA, is they feel that their needs are not being met.”

CFIR domain 4: characteristics of individuals

We focused on the CFIR construct of “knowledge and beliefs about the intervention”, which was operationalized as participants’ recommendations on how to improve the SIRI team’s function and reach. One key facilitator was to further establish the SIRI team as the de facto SUD/Addiction Medicine team:

“I think that it would be great if we had a team that is really, truly dedicated to addiction, and I'm saying that because we right now don't have any available dedicated team”

Participants also felt that more formally incorporating psychiatrists, psychologists, and social workers to the SIRI team would help ensure the holistic care. One patient participant shared:

“I think that a social worker should be more involved… [hospitals] would just give me a list of shelters to go to and kick me out… Maybe they didn’t have time, or they don’t have the resources, but I think havin’ a social worker more involved in somebody’s release is important.”

A clinician further underscored the need for holistic care:

“You are dealing with [a] patient. Yes, they have this infection. They have those addiction issue, but they have [mental health] issues in the background that [are] preventing this patient from fully profiting from the care we are providing to this patient.”

Clinicians also suggested that the SIRI team could train other physicians, social workers, and nurses on best practices when caring for patients with SUD and SIRIs. Providing education to other clinicians outside the SIRI team could extend the team’s reach and help mitigate the stigma and judgment that patients with SUD often experience when seeking care.

Lastly, clinicians thought it was important for the SIRI team to facilitate rapid and ongoing feedback about patient clinical course and outcomes with others involved in a patient’s care. In the instances where the SIRI team contacted hospitalists to give an update on post-discharge successes and failures, the clinicians felt this helped affirm the quality of care provided by the SIRI team.

Discussion

In this study, we used patient and clinician perspectives to evaluate contextual determinants of the continued implementation of an integrated infectious disease and SUD treatment intervention for PWID hospitalized with injection drug use-associated infections. The CFIR was operationalized to guide interview questions and frame responses to maximize actionable results of our study. The primary facilitators of SIRI team success were (1) the team’s holistic, patient-centered, and non-judgmental approach, (2) the effective low barrier communication with patients and other clinicians, (3) the provision of close post-hospital follow-up, and (4) the team’s ascent as the local authority on providing MOUD to hospitalized patients. Barriers to the SIRI team’s success included (1) entrenched stigma in the healthcare system against PWID, (2) lack of availability of critical resources like insurance, housing, and financial support, (3) ineffective communication about the team’s existence, and (4) detrimental effects of focus on maximizing patient volume and reducing costs. The results of this study will be used to adapt and optimize the SIRI team intervention to be studied in a randomized controlled efficacy trial.

All participants who had interacted with the SIRI team, either as a patient or as a colleague, highlighted the non-judgmental, compassionate approach of the team as a central component of its success. Numerous studies have highlighted the adversity faced by PWID with infections when interacting with the healthcare system. PWID presenting for healthcare report stigma from providers, lack of belief and attention to their chief concerns, and in some cases, abusive and cruel behavior [36]. This mistreatment leads PWID to delay presenting for care—potentially exacerbating infections—or attempting self-treatment of infections [37, 38]. While hospitalized, PWID with infections report maltreatment by staff, ignored pain, and a generally carceral and punitive approach to in-hospital substance use [10, 39, 40]. These qualitative results corroborate quantitative results of improved patient trust when exposed to harm reduction-focused SUD treatment teams [41].

The SIRI team represents one of several different emerging models of integrated infectious disease and SUD care for patients with SIRIs. A common theme across interventions is the involvement of addiction medicine experts in infectious disease and surgical care to educate, reduce stigma, ensure patient-centered care, and navigate controversial clinical decisions, like cardiac surgery or use of peripherally inserted central catheters [42]. Multidisciplinary care meeting approaches such as the DUET, MEET, and OPTIONS-DC programs integrate an array of SUD professionals, sometimes including persons with lived experience, to inform treatment plans [23, 24, 42]. Other programs have focused on the integration of low-barrier post-hospital MOUD and harm reduction with ongoing infectious disease management [22, 43] and are currently being tested in randomized controlled trials (RCTs).

Our interviews highlighted the benefits of the close post-hospital follow up and the continued low barrier, harm reduction-centered approach to care for PWID experiencing SIRIs. The program described by Lewis and colleagues [21] similarly employed repeated contact by counselors with patients during and after a hospital stay, with increased engagement with the team associated with fewer hospital readmission [21]. Although models differ, RCTs of patient navigation for hospitalized persons with SUD have had mixed results [44,45,46]. In an intervention with similar aims as the SIRI team but without infectious disease treatment, Gryczynski and colleagues [46] showed that proactive case management, advocacy, service linkage, and motivational support reduced hospital readmission and increased SUD treatment engagement compared to treatment as usual [46]. In contrast to many linkage-to-care interventions, rather than “link” to outpatient SUD or infectious disease treatment, the SIRI team continues to directly provide the needed care seamlessly between inpatient and outpatient settings. Patients reported pleasant surprise when experiencing the team's continuing to call, prescribe buprenorphine, and ensure infection resolution after discharge. We believe this continuity and familiarity with patients improved efficiency, patient experience, and outcomes.

The results of this study will be used to improve the implementation of the SIRI team, such as deploying context-specific strategies (i.e. hospital workforce training on harm reduction, increasing institutional knowledge of the SIRI team intervention, improving cross-department communication streams). The team will be systematizing a feedback protocol to ensure discharging hospitalists are updated on the outcomes of their SIRI team patients to increase clinician engagement. SIRI team members’ education of hospitalists, house staff, and nursing staff is ongoing. Beyond improving care for individuals with SIRI, the SIRI team aims to show improvement in important health system level outcomes, such as length of stay, readmission rate, and patient-directed discharge. Finally, a multicenter RCT evaluating the SIRI team intervention versus treatment as usual is under development in order to test efficacy in increasing readmission-free survival [47].

Our study has several limitations. Due to the small sample size, few patients who had been cared for by the SIRI team, and lack of implementation outcomes measured quantitatively, we were limited in our ability to evaluate the Process domain of CFIR. Additionally, the small sample size increases the risk of missed themes and misleading conclusions. We had intended to interview hospital administrators but were unable to arrange interviews in a timely manner. Thus, the administrator perspective, an important perspective in the outer context, is missing from these data. Another perspective missing from these data is that of the SIRI team members themselves. Due to the overlap between the researchers implementing the SIRI team intervention and evaluating the implementation efforts, we were unable to ascertain the implementer perspective for this study. All clinician participants interviewed in this study cared for patients with SIRI in the inpatient setting and may not be as familiar with post-discharge care experiences of PWID. Finally, interviews pertained to a specific SIRI team model in one hospital and results may not be externally applicable to other health systems. Studying SIRI team implementation in multiple sites will help elucidate how local contextual factors impact the efficacy of the intervention in diverse settings.

Conclusions

Using an implementation science framework, we conducted an implementation evaluation of an integrated infectious disease/SUD intervention for persons experiencing IDU-associated infections. Patient and clinician participants highlighted the myriad barriers to care for PWID both within—and external to—the hospital and identified how the SIRI team mitigated these obstacles while contributing to culture-change and reducing stigma toward PWID. Ongoing research will further evaluate the clinical effectiveness of the team on infection, substance use, and healthcare utilization-related outcomes as well as examining implementation strategies that improve our implementation outcomes. Testing of the intervention in a hybrid RCT is necessary to evaluate efficacy and guide implementation considerations of SIRI teams across health systems heavily impacted by the infectious disease/SUD syndemic.