Improving Outcomes for People with Injection Drug Related Endocarditis: Are Medications for Opioid Use Disorder Enough?

There is no doubt that medications for opioid use disorder (MOUD), methadone and buprenorphine, reduce mortality from fatal overdose, and improve other clinical and social outcomes (Larochelle et al., 2018; Medications for opioid use disorder save lives, 2019), however, more research is needed to know if MOUD improve outcomes among individuals with injection drug related infective endocarditis (IE), a serious complication of injection drug use. Previous studies show MOUD are rarely initiated in these cases and one study showed it did not improve survival (Rodger et al., 2018; Rosenthal et al., 2016). In this issue of the Journal of Addiction Medicine, Suzuki et al report the results of a retrospective study conducted at an academic, tertiary care hospital in Massachusetts between 2013 and 2015 that suggests initiation of MOUD in individuals with injection drug use related IE did not reduce long-term mortality or valve reinfection. Utilizing MOUD to manage withdrawal in the hospital is beneficial for patients in its own right, but these results are disappointing for those of us advocating and developing systems to integrate care for opioid use disorder (OUD) and infectious diseases (Springer et al., 2018). However, rather than diminishing enthusiasm for integrating MOUD into infectious disease care, we should interpret the study carefully and draw lessons that will drive future interventions.

The findings of this study are not conclusive for several reasons. First, this is a retrospective, observational study with a small sample size of only 26 individuals of which 8 received buprenorphine and 8 methadone, and therefore has little power to detect true differences between patients who received MOUD and those who did not. Second, 60% of the cohort were transferred to the hospital where the study took place, constituting a group of patients with high illness severity, even for those with IE. Furthermore, these transferred patients waited 8 days on average before transfer and subsequent evaluation by an addiction specialist at the study hospital. Third, to enter the cohort, individuals with injection drug related IE had to be evaluated by an addiction consultant, but we do not know if this group is representative of all cases in the hospital. Fourth, patients treated with methadone, one of two exposures in the study, effectively received a taper protocol with average doses of approximately 30 mg daily, less than half the dose shown to improve outcomes. Fifth, the measure of receipt of medication following discharge is crude, defined as any evidence of MOUD receipt, and is unable to differentiate granular data on engagement, retention or re-engagement with MOUD. Finally, individuals treated with buprenorphine were not permitted to receive additional full opioid agonist therapy to treat acute pain from septic emboli or surgery which may have decreased treatment acceptance. Thus, although this study addresses a question of profound significance, its limitations should temper broad conclusions and instead point us in future directions.

We believe that the results of this study may be best put into context as a snapshot amidst a rapidly changing standard of care for people with injection drug associated infections. A case of a patient from the time period when the study took place is exemplary of the challenges and opportunities: A young woman with a history of tricuspid valve replacement for injection drug related endocarditis 18 months prior presented to the hospital with fever. Blood cultures were negative, and an echocardiogram showed no vegetation. She was told the fever was from transient bacteremia due to injection and did not represent a systemic infection. During that admission, she initiated buprenorphine, but when no prescriber could be located near her home, it was tapered and discontinued prior to discharge. She returned to injecting drug use and subsequently represented weeks later, transferred to the intensive care unit from an outside hospital with methicillin sensitive Staphylococcus aureus prosthetic valve infection and a cardiac abscess complicated by mixed cardiogenic and septic shock requiring multiple pressor medications to maintain her hemodynamic status. She experienced opioid withdrawal because she was treated only with low-dose, short acting oxycodone. The consulting cardiac surgeon told the patient, he would operate if she could “promise never to inject drugs again.” She understood the chronic and relapsing nature of the disease of addiction from her own experience and said she could not make such a promise. Her father nodded in agreement, noting she had “burned a lot of bridges” over the years and might not “deserve” the surgery. The patient agreed with her father, a remarkable demonstration of internalized stigma. Despite strong indications, surgery was not offered. She developed heart block from her untreated cardiac abscess and required emergent temporary pacer placement. At this point, a different cardiac surgeon re-evaluated the patient and she was urgently taken for abscess debridement and valve replacement. Following the surgery, the patient was treated with buprenorphine, linked to outpatient care for opioid use disorder, and more than 2 years later still alive without recurrent infection.

As we can learn from this vignette, even if MOUD is offered, patients with injection drug use associated infections face tremendous clinical and structural challenges amidst systems of care that often fail them. In many hospitals, MOUD are not available to patients, but even when these treatments are available, patients may face poorly managed withdrawal and cravings, lack of access to valve replacement surgery (Vlahakes, 2017), post-acute medical care and MOUD upon discharge (Wakeman and Rich, 2017), as well as stigma (Biancarelli et al., 2019), and high rates of homelessness and mental illness. Infectious endocarditis is a feared complication of injection drug use and there is a growing literature reporting unacceptably high mortality rates for this condition (Leahey et al., 2019; Rudasill et al., 2019). The authors rightly call for research into additional strategies to improve care for these patients. Initiation of MOUD in the hospital and linkage to longitudinal care at discharge should be standard of care, however we must acknowledge that patients face multiple structural challenges, and additional interventions are necessary.

Suzuki and others have shown that addiction specialists are an important part of the interdisciplinary teams needed to take care of patients hospitalized with infectious complications of injection drug use (Suzuki, 2016; Trowbridge et al., 2017). Patients need access to MOUD, but they must also access a range of services and supports to address their individual needs. Programs with enhanced case management and peer supports are encouraging, affirming a patient’s value and humanity. However, multiple models that combine MOUD with antibiotics will be necessary: home with peripherally inserted catheters for antibiotics (Fanucchi et al., 2019), admission to post-acute care facilities (Wakeman and Rich, 2017), or transition to inpatient substance use disorder units with integrated care (Englander et al., 2018). Individuals across this spectrum of treatment as well as those who do not wish to or are unable to stop using opioids should receive medical care in conjunction with effective harm reduction services, including syringe services and access to supervised injection sites, though legal barriers will need to be overcome in the United States (Rachlis et al., 2009).

Research into new models of care that integrate addiction and infectious disease care before, during, and following hospitalization is urgently needed to design effective interventions that improve patient outcomes (Springer et al., 2018). Broad adoption of protocols to initiate buprenorphine or methadone with inpatient titration and linkage to outpatient treatment programs and buprenorphine micro-dosing to initiate effective treatment without withdrawal, especially when patients also require full opioids for acute pain are promising. Additionally, long acting injectable formulations of buprenorphine offer a new opportunity and deserve rigorous evaluation.

Further research on clinical outcomes as well as patient preferences and experiences will be needed to ensure new care models are effective, relevant and acceptable to patients. While we build this evidence, MOUD should be initiated in the hospital and efforts to improve linkage and retention to outpatient care prioritized. To achieve these aims, we must take a broad view and work to improve clinical capacity to deliver MOUD while reduce stigma and address social and structural barriers patients face. New models are needed so that patients have meaningful opportunities to survive, heal, and recover.