Endocarditis in people who inject drugs

A thirty-two-year-old woman is on the twelfth day of hospitalization for tricuspid valve endocarditis caused by Staphylococcus aureus. Blood cultures have turned negative, the fever has subsided and the follow-up echocardiogram is scheduled. She is on intravenous antibiotics and, since admission, a dose of methadone that seems insufficient: she spends the afternoons restless, in pain and with chills, and tells the nursing staff she is leaving today. The team has a form for discharge at the patient's request ready and four weeks of antibiotics ahead. This text uses endocarditis in people who inject drugs as a setting for clinical reasoning: what changes in management, why treating the dependence is part of treating the infection, when partial oral antibiotics become an option and how to decide in the face of discharge at the patient's request without abandoning the person. The scene and the examples throughout the text are fictional and were written for teaching.

NEXO is a daily medical diagnosis game: one clinical case a day, up to six progressive clues, and ICD-10 proximity feedback across five levels.

The text also avoids being an antibiotic protocol. Regimens, doses, durations and surgical criteria follow the endocarditis guidelines and each service's protocols, with the infectious diseases and cardiology teams. The focus here is the way of thinking: which questions to ask when the infection and the dependence meet in the same hospitalization and how to keep the person in care. Recognition of endocarditis and the role of stigma in diagnosis were already discussed in the post about stigma and language. If you play the case of the day on NEXO, you will recognize the move: the clue that changes management is not always the most striking one.

Key points

• In endocarditis in people who inject drugs, treating the dependence is part of treating the infection, and poorly controlled withdrawal is a frequent cause of discharge at the patient's request.
• Starting buprenorphine or methadone during hospitalization increases engagement in treatment after discharge, according to a randomized trial.
• Partial oral antibiotics, studied in the POET trial in left-sided endocarditis and in people who inject drugs in observational studies, widen the options when a long hospitalization is not feasible.
• Discharge at the patient's request calls for a harm reduction decision: antibiotics where possible, naloxone, an opioid agonist, a scheduled appointment and an open door to come back.
• Recurrence is common, and each hospitalization is also an opportunity to start or adjust treatment for the dependence.

Why endocarditis changes the whole hospitalization

Infective endocarditis in people who inject drugs is a serious and increasingly frequent disease. Alysse Wurcel and colleagues described a rise in hospitalizations for endocarditis among young people who inject drugs in the United States (Wurcel et al., 2016). In Canada, Rodger and colleagues showed that, among first episodes of endocarditis at three hospitals, more than half occurred in people who inject drugs, and they highlighted the potential role of treating the dependence in outcomes (Rodger et al., 2018).

Treatment usually requires weeks of antibiotics, often intravenous, a cardiology assessment and, in some people, surgery. The 2023 European Society of Cardiology guideline and the American Heart Association scientific statement reinforce management by a multidisciplinary endocarditis team (Delgado et al., 2023; Baddour et al., 2015). For people who inject drugs, that team gains a member who is not always in the room: someone who takes care of the dependence.

The surgical decision

The indication for valve surgery follows guideline criteria, such as heart failure, uncontrolled infection and risk of embolism. In people who inject drugs, the discussion usually gains one more component: the fear that the new valve will also become infected. Rodger and colleagues pointed out the need to better study the selection criteria for surgery in this population (Rodger and colleagues, 2018). For the reasoning, it helps to separate two questions. The first is clinical: is surgery indicated by the usual criteria? The second is about care: what can the team do so that addiction treatment accompanies the surgery and reduces the risk of recurrence? Mixing the two questions favors decisions based on moral judgment, and the multidisciplinary team helps keep them separate.

Infection and addiction in the same bed

During the hospital stay, the person faces withdrawal, pain, anxiety, hospital rules and, often, distrust from the team. If withdrawal is not treated, the discomfort grows every day, and leaving early starts to seem the only way to find relief. Clinical reasoning needs to include the question: what is making this hospital stay unbearable, and what can we change today?

Treating addiction is treating the infection

The trial by Jane Liebschutz and colleagues randomized people hospitalized for medical causes, with opioid dependence and not actively seeking treatment, to detoxification or to starting buprenorphine with referral. Starting buprenorphine increased engagement in treatment and reduced illicit opioid use six months after discharge (Liebschutz and colleagues, 2014). The message for the reasoning is direct: the hospital stay is an opportunity to start treatment, and waiting for discharge to "refer" loses that window.

Laura Marks and colleagues analyzed people hospitalized with severe infections related to opioid use and showed that consultation with the addiction medicine team was associated with fewer readmissions (Marks and colleagues, 2019). Where this service exists, calling it early usually changes the course of the hospital stay. Where it is lacking, the clinical team can take on the basics: treat withdrawal, offer an agonist and manage pain.

Choosing the agonist during the hospital stay

The choice between methadone and buprenorphine follows the same principles as in any context, with some particularities of the hospital stay: interactions with antibiotics, need for procedures, acute pain and what will be available after discharge. The post about buprenorphine or methadone organizes this shared decision. If the choice is buprenorphine in someone who uses potent opioids, the post about precipitated withdrawal when starting buprenorphine helps plan the induction.

Thinking about the exit from the first day helps with the choice. An agonist started in the hospital only protects if the person can continue it afterward. So, before the final choice, it helps to ask where the person lives, which treatment service is nearby, whether they have been in treatment before and what worked or got in the way. These answers usually weigh as much as the pharmacology. When the hospital stay ends earlier than expected, this preparation makes a difference: the bridge to the first appointment is already planned, and the person leaves knowing where to get the next dose.

Pain and withdrawal are clues, not manipulation

A person with endocarditis may have real pain from emboli, septic arthritis, abscesses or procedures. At the same time, opioid tolerance changes the dose needed for relief. Treating pain properly, including with opioids when indicated, and maintaining the baseline agonist is usually safer than reducing everything for fear of "feeding the addiction". Some questions help organize the reasoning at the bedside:

• Is withdrawal being measured and treated every day?
• Has the agonist dose been adjusted since admission?
• Does the pain have a new clinical cause that needs investigation?
• Is there a safe way for the person to leave the room, get some air or smoke?
• Has someone explained what to expect in the coming weeks?

Partial oral antibiotics: what changed

For a long time, endocarditis meant weeks of intravenous antibiotics in hospital. The POET trial, by Kasper Iversen and colleagues, randomized people with left-sided endocarditis in stable condition to either continue intravenous antibiotics or switch to oral antibiotics. Switching to the oral route was non-inferior to continuing the intravenous route (Iversen and colleagues, 2019). The 2023 ESC guideline incorporated partial oral treatment as an option in selected stable people (Delgado and colleagues, 2023).

The OVIVA trial, by Ho Kwong Li and colleagues, studied bone and joint infections, not endocarditis. It showed that oral antibiotics were non-inferior to intravenous antibiotics in the first six weeks of treatment of complex orthopedic infections (Li and colleagues, 2019). For reasoning, OVIVA reinforces a broader idea: the oral route can be adequate in severe infections when the antibiotic and the patient are well chosen.

Were people who inject drugs studied?

Few people who inject drugs took part in these trials. Observational studies fill part of the gap. Marks and colleagues assessed partial oral antibiotics in people who inject drugs admitted with invasive infections and concluded that offering this option to those who cannot complete an inpatient intravenous course was beneficial (Marks and colleagues, 2020). In a retrospective single-center cohort, Wildenthal and colleagues studied S. aureus bacteremia complicated in this population: discharge without oral antibiotics was associated with a lot of morbidity, and the transition to the oral route with outpatient support, after at least ten days of effective intravenous treatment following negative cultures, emerged as an alternative for those who refuse the intravenous standard (Wildenthal and colleagues, 2023). It is a cohort finding, not a validated criterion. POET did not include right-sided endocarditis, which is the one in the scene; for it, the basis is the observational studies in people who inject drugs (Marks and colleagues, 2020; Wildenthal and colleagues, 2023).

Oral antibiotics, therefore, enter as one more option in the repertoire, to be decided with the infectious diseases team and with the person. They are especially useful when the real alternative is leaving without any treatment.

And what about the intravenous catheter at home?

Peripherally inserted central catheters for home intravenous antibiotics are often avoided in people who inject drugs, for fear of use of the access. Rodger and colleagues described a high risk of endocarditis recurrence in this population, especially when there was misuse of the catheter (Rodger and colleagues, 2019). The decision is individual. In some services, structured programs allow outpatient intravenous antibiotics with follow-up; in others, the oral route is the safest option for discharge.

Discharge on request: the size of the problem

People who use drugs leave the hospital against medical advice more often than other patients. In endocarditis, Kimmel and colleagues showed, in a national cohort in the United States, that discharges on request increased and were more common and more sharply rising among people who inject drugs (Kimmel and colleagues, 2021). Appa and colleagues studied people who use drugs admitted with invasive S. aureus infections: the chance of readmission for infection was almost four times higher after discharge on request, and overdose was a common cause of death in this group (Appa and colleagues, 2022).

The systematic review by Lianping Ti and Lianlian Ti on hospital discharge against medical advice among people who use drugs called for attention to the individual, social and structural characteristics that lead to discharge, and not only to the person (Ti and Ti, 2015). Poorly treated withdrawal, pain, perceived stigma, rigid rules, lack of space to smoke and problems at home appear among the reasons described in the literature.

Why the expression matters

Many services have started using "person-directed discharge" instead of "discharge against medical advice", to remove the tone of blame and keep the focus on continuity of care. Alfandre described that discharge at the person's request is often treated as the end of medical responsibility, when the person still needs care (Alfandre, 2009). The word chosen changes what the team does next, as the post about stigma and language shows.

When the person says they are leaving

The moment the person announces they are leaving is often tense. The automatic reaction is to call the attending physician, hand over the form and record the refusal. Clinical reasoning calls for one step before that: understanding the reason. A short conversation, unhurried and without threats, often reveals a cause that can be treated right away.

Some questions help to understand what is going on:

1. "What is hardest for you here right now?"
2. "Are you feeling withdrawal or pain at this moment?"
3. "Is there something out there that needs your attention today?"
4. "What would make you stay a few more days?"
5. "If you decide to go, what can we arrange so you leave safer?"

The post about motivational interviewing offers techniques for conversations in which the person is divided, such as reflecting ambivalence without pressing.

Capacity and choice. A person with decision-making capacity can choose to leave. Assessing capacity is part of the reasoning, especially when there is intoxication, severe withdrawal, delirium or a brain embolism. When capacity is preserved, the task changes: the team now decides how to reduce the risk of the departure. Recording the capacity assessment, the options offered and the person's reasons protects continuity of care and helps the next professional pick up the conversation.

Negotiating time. Not every departure has to happen right away. It is often possible to agree on a few hours to prepare the prescription, talk with infectious diseases about the oral route, secure the next dose of the agonist and hand over the naloxone kit. This interval turns an abrupt departure into a planned transition. When the person sees that the team is helping, and not trying to hold them back, the conversation usually gets easier, and sometimes they decide to stay one more day.

Risk reduction at discharge

Discharge at the person's request can be a moment of care, not of closure. The table organizes what usually goes into the decision, always adjusted to the case and to what the service can offer.

Step	What to consider	Question that helps
Antibiotic	Oral route possible, with infectious diseases, even if the ideal regimen is different	What is the best option the person will be able to use?
Opioid agonist	Prescription or bridge until the next dose, according to local rules	Does the person leave with guaranteed access to methadone or buprenorphine?
Naloxone	Kit and guidance for the person and for those who live with them	Has tolerance dropped during the stay?
Follow-up	Scheduled appointment, contact with a specialized service or outpatient clinic	Who will see the person in the coming days?
Red flags	Fever, dyspnea, chest pain, neurological deficit	Does the person know when to come back and that they will be welcomed?
Record	Reasons for leaving, options offered and the person's decision	Will the next professional understand what was agreed?

The open door

Stating explicitly that the person can come back at any time is part of the approach. Many people avoid returning out of shame or fear of being treated as "the one who ran away". The sentence "if you change your mind, you can come back today and we will continue where we left off" is simple and keeps the bond.

Harm reduction in use

After a hospital stay, opioid tolerance is often lower, and the risk of overdose increases. Guidance on using less, avoiding using alone, having naloxone nearby and using sterile supplies is part of the care. The post about harm reduction beyond all or nothing develops this logic, and the post about opioid overdose and the naloxone decision helps to think through reversal.

Biases that get in the way of the approach

Pat Croskerry describes cognitive shortcuts that affect clinical decisions (Croskerry, 2003). In endocarditis in people who inject drugs, some of them appear often over the course of the stay:

• Affective bias. Irritation with requests for analgesia or with difficult behavior changes decisions without being noticed.
• All-or-nothing thinking. If the person does not accept the ideal regimen, the team stops offering what is possible.
• Moral attribution of recurrence. The second endocarditis is read as the person's failure, not as a sign that the addiction treatment needs adjustment.
• Closure in infection. The team treats the valve and forgets the addiction, which is the cause of the next admission.

The post about affective bias explains how the professional's emotions enter the reasoning. Recurrence deserves special attention: Rodger and colleagues showed a high risk of a new episode in this population (Rodger and colleagues, 2019), which makes each admission an opportunity to start or adjust treatment. Reading the second admission as information, not as failure, changes the team's questions: what was missing in the addiction treatment the first time, was the person able to stay on the agonist, was there access to sterile material, did anyone follow up in the first days after discharge? Each answer points to a possible adjustment for this admission.

How NEXO helps train this reasoning

In NEXO, the case of the day is fictional, with a real clinical presentation and up to six progressive clues. You propose a diagnosis by searching for the ICD-10 code and receive feedback on proximity in five levels, which shows whether the guess landed on the exact code, the same category, the same group, the same chapter or a different chapter. This format speaks to the theme of this text: the person's context is a clinical clue, and the feedback shows whether you gave it the right weight.

To use the case of the day as training, try three habits. Before each guess, ask which context factors change the probability of the hypotheses. When the answer seems obvious, write down the second most likely hypothesis. After the result, note which clue you set aside.

Three exercises with cases

List the two problems. In a case of infection in someone who uses drugs, write the management for the infection and the management for the addiction, side by side.

Plan the exit. Imagine the person decides to leave today and write down what they would take with them.

Rewrite the record. Take a chart sentence such as "left against advice" and rewrite it with person-centered language.

The post about case-based learning shows why training with cases fixes this type of decision.

The limits of using the game are described on the NEXO economy page. If you research teaching of clinical reasoning, the page for researchers explains how to get in touch. Professionals who write or review cases find information on curation and on creators, and institutions can learn about the proposal on the partnerships page.

Frequently asked questions

Can people who inject drugs receive oral antibiotics for endocarditis?

They can, in selected situations and with infectious disease input. The POET trial studied stable people with left-sided endocarditis, and observational studies assessed partial oral therapy in people who inject drugs (Iversen and colleagues, 2019; Marks and colleagues, 2020).

Is it worth starting buprenorphine during the admission?

It is. The trial by Liebschutz and colleagues showed more engagement in treatment and less use of illicit opioids after discharge (Liebschutz and colleagues, 2014).

What to do when the person wants to leave before the antibiotic course ends?

Understand the reason, treat what is possible at the time, assess decision-making capacity and, if the departure is maintained, offer the best possible regimen, opioid agonist, naloxone, follow-up and an open door to return.

Does discharge at the person's request end the team's responsibility?

The responsibility continues. The person still needs care, and the departure is a moment to reduce risks and keep the connection (Alfandre, 2009).

Why does tolerance matter at discharge?

During the admission, opioid tolerance can fall, and returning to use at the previous dose increases the risk of overdose. That is why naloxone and guidance are part of the conversation.

Final checklist

• I treated withdrawal and adjusted the opioid agonist from admission.
• I called the addiction medicine team when available.
• I assessed pain with the same attention given to the infection.
• I discussed with infectious disease whether partial oral antibiotics are an option.
• I asked about the reason before accepting the discharge at the patient's request.
• I offered a possible antibiotic, an agonist, naloxone and follow-up at discharge.
• I made it clear that the person can come back at any time.

Want to practice your reasoning with progressive clues? Play the case of the day.

References

1. Alfandre DJ. "I'm going home": discharges against medical advice. Mayo Clin Proc. 2009. PMID 19252113.
2. Appa A, Adamo M, Le S, et al. Patient-Directed Discharges Among Persons Who Use Drugs Hospitalized with Invasive Staphylococcus aureus Infections: Opportunities for Improvement. Am J Med. 2022. PMID 34508704.
3. Baddour LM, Wilson WR, Bayer AS, et al. Infective Endocarditis in Adults: Diagnosis, Antimicrobial Therapy, and Management of Complications: A Scientific Statement for Healthcare Professionals From the American Heart Association. Circulation. 2015. PMID 26373316.
4. Croskerry P. The importance of cognitive errors in diagnosis and strategies to minimize them. Acad Med. 2003. PMID 12915363.
5. Delgado V, Ajmone Marsan N, de Waha S, et al. 2023 ESC Guidelines for the management of endocarditis. Eur Heart J. 2023. PMID 37622656.
6. Iversen K, Ihlemann N, Gill SU, et al. Partial Oral versus Intravenous Antibiotic Treatment of Endocarditis. N Engl J Med. 2019. PMID 30152252.
7. Kimmel SD, Kim JH, Kalesan B, et al. Against Medical Advice Discharges in Injection and Non-injection Drug Use-associated Infective Endocarditis: A Nationwide Cohort Study. Clin Infect Dis. 2021. PMID 32756935.
8. Li HK, Rombach I, Zambellas R, et al. Oral versus Intravenous Antibiotics for Bone and Joint Infection. N Engl J Med. 2019. PMID 30699315.
9. Liebschutz JM, Crooks D, Herman D, et al. Buprenorphine treatment for hospitalized, opioid-dependent patients: a randomized clinical trial. JAMA Intern Med. 2014. PMID 25090173.
10. Marks LR, Munigala S, Warren DK, et al. Addiction Medicine Consultations Reduce Readmission Rates for Patients With Serious Infections From Opioid Use Disorder. Clin Infect Dis. 2019. PMID 30357363.
11. Marks LR, Liang SY, Muthulingam D, et al. Evaluation of Partial Oral Antibiotic Treatment for Persons Who Inject Drugs and Are Hospitalized With Invasive Infections. Clin Infect Dis. 2020. PMID 32239136.
12. Rodger L, Shah M, Shojaei E, et al. Recurrent Endocarditis in Persons Who Inject Drugs. Open Forum Infect Dis. 2019. PMID 31660358.
13. Rodger L, Glockler-Lauf SD, Shojaei E, et al. Clinical Characteristics and Factors Associated With Mortality in First-Episode Infective Endocarditis Among Persons Who Inject Drugs. JAMA Netw Open. 2018. PMID 30646383.
14. Ti L, Ti L. Leaving the Hospital Against Medical Advice Among People Who Use Illicit Drugs: A Systematic Review. Am J Public Health. 2015. PMID 26469651.
15. Wildenthal JA, Atkinson A, Lewis S, et al. Outcomes of Partial Oral Antibiotic Treatment for Complicated Staphylococcus aureus Bacteremia in People Who Inject Drugs. Clin Infect Dis. 2023. PMID 36052413.
16. Wurcel AG, Anderson JE, Chui KK, et al. Increasing Infectious Endocarditis Admissions Among Young People Who Inject Drugs. Open Forum Infect Dis. 2016. PMID 27800528.