Infective endocarditis (IE) is a bacterial or fungal infection of the heart valves and lining that Jean François Fernel first mentioned nearly 500 years ago (1554). William Osler described the disease in detail in a series of lectures in 1885 [ ]. The clinical profile of endocarditis has undergone rapid and dramatic change in recent decades. Both microbiology and risk factors related to IE have evolved with new technology and a changing social milieu. An important risk factor is the intravenous (IV) injection of illicit drugs such as heroin and cocaine, and central to the evolution of this disease is the so-called “opioid epidemic” currently affecting the United States and other countries.
Clinical features and outcomes
IE is an uncommon disease, but recently has rapidly increased in incidence. Recent estimates of data from the early 2000s reported an incidence of 3–10 per 100,000 per year worldwide [ ], but rates of IE have been increasing, and, although regional variability is evident, the United States has reported the highest incidence of IE in the world: 15 cases per 100,000 people per year [ ].
The primary causative agent has transitioned from oral streptococcus to staphylococcus owing to many factors such as the development of effective antibiotics, the increased use of prosthetic implants, invasive IV catheters, immunosuppressants, and, central to our discussion, IV drug abuse [ ]. Endocarditis is diagnosed by a combination of history, physical exam, and echocardiography; it is preferentially treated with antibiotics, but, owing to changes in epidemiology and acuity, IE has increasingly become a surgical disease [ ]. Outcomes remain suboptimal: overall in-hospital mortality rates have been reported to be as high as 20% [ ].
Surgical outcomes for IE are heterogeneous and depend heavily on the valve(s) involved, invasiveness of the disease, and comorbidities. Overall perioperative mortality is less than 10% [ ]. Perioperative mortality for isolated native aortic and mitral valve infections confined to the leaflet are similar, at 7% and 6%, respectively, but outcomes for invasive (infection extending beyond the leaflets into adjacent structures) and multivalvular IE are worse. Overall survival is 66% at 5 years, with better survival for noninvasive aortic valve endocarditis and worse survival for mitral valve disease, combined valve disease, and invasive disease. IV drug use is an important comorbidity affecting long-term survival; relapse into drug use and reinfection can lead to worse outcomes long-term [ ].
The use of prescription pain relievers has drastically increased in the United States over the last several decades—an estimated 4.3 million people participate in nonmedical use of prescription pain medications, and these patients are 40 times more likely than the general population to use IV drugs [ ]. About 80% of new heroin users have previously misused prescription pain killers [ ].
The incidence of intravenous drug use-associated infective endocarditis (IVDU-IE) has doubled in the United States in the 15 years between 2002 and 2016 [ ], and just in the 5 years between 2010 and 2015, has doubled from 15.3% to 29.1% of all IE cases [ ]. While IVDU patients with IE have better short-term survival than non-IVDU patients, probably a function of their younger age, their rates of readmission and drug use are higher [ ]. Midterm outcomes are not as promising—at three–four months, the hazard of death or reoperation for IVDU-IE patients are more than 10 times the hazard for IE patients who do not use IV drugs [ ]. In addition, long-term outcomes are worse for patients with IVDU-IE compared with patients who do not use IV drugs: 5 and 10 year survivals are 46.7% and 41.1% versus 71.1% and 52.0%, respectively [ ].
Major ethical issues
Ethical dilemmas are common in the treatment of IVDU-IE, and may substantially affect treatment and outcomes. Common issues often stem from prejudicial attitudes toward drug use, such as the merits of treating patients with apparent self-harming behavior, allocation of scarce resources in the cost-conscious environment of modern care delivery, and doubts about long-term outcomes of treatments owing to possible continued drug use. Less common issues may be seen in some cases, but may nonetheless present serious dilemmas, for example, in the treatment of IVDU-IE in the context of a viable pregnancy.
Caring for the patient who self-harms
Caring for the IV drug user who contracts IE can be challenging on several levels, but many difficulties arise not from questions of “how to treat the problem?” but from questions of “should we treat the problem?” Clinicians find fulfillment in a job well done—knowing that the right patient received the right treatment at the right time, and will consequently do well. Conversely, satisfaction might wane when clinical judgment tells us that a patient will do poorly no matter what we do. Some may feel that whatever treatment they provide to IV drug users, their patients will simply return to IV drug use soon after they are discharged, incurring a high risk of recurrent IE. In addition, IV drug users are commonly denigrated; even though opioid use disorder (OUD) is a recognized diagnosis and disease process, drug addiction is often viewed as a “moral failing” [ ]. Many studies have indicated that negative attitudes and perceptions toward IV drug users are widespread [ ]. The stigma attached to IV drug use has substantial implications for the therapeutic relationship, often causing discontent for both doctor and patient, and possibly damaging treatment outcomes.
Limiting or changing treatment based on patients’ lifestyle or behavior
The presence of “self-harming” behavior and its impact on disease affects the attitude of many care providers. Most surgeons will operate for an IV drug user with first-time endocarditis; many may draw the line there, however, with a “one and done” policy toward valve replacement. IV drug use is clearly the independent action of the patient, but addiction is complex and many external influences affect addicts’ behavior. A claim that patients have no responsibility for their health would be specious, undermining the principle of autonomy, but other factors may play an important role; for example, opioid addiction often originates from prescribed pain medications, and lack of access to adequate addiction treatment may inhibit or prevent recovery. The presence of social issues or psychological conditions may make stopping the use of drugs more difficult. OUD is highly associated with other mental health disorders that require concurrent treatment in order to optimize results [ ]. To simply refuse to treat recurrent IE due to the presence of drug use that has not been adequately treated would be irresponsible, denying the patient needed care [ ]. OUD is recognized as a medical disease, and the available evidence indicates widespread undertreatment of OUD in association with IE and other IVDU-related infections [ ].
IVDU-IE is not the only disease that stems from apparently self-inflicted harm. It would seem inconsistent for a cardiothoracic surgeon to refuse operating on IV drug users with recurrent endocarditis but to have no hesitation in operating on smokers with recurring lung cancer or obese uncontrolled diabetics with coronary artery disease who have already received stents.
Baldassarri et al. make the interesting comparison of IVDU-IE to other diseases or injuries incurred by more positively viewed risky lifestyles [ ]. They present examples of the avid hiker who presents with recurrent tick-borne illnesses due to outdoor tick exposure and the cyclist or motorcyclist who presents with recurrent fractures owing to accidents while riding. These ailments are incurred from participation in activities that are part of active lifestyles that are viewed positively by society. Even though the patient knows the risk of participating in these activities, and continues to do so despite previous adverse events, no physician would refuse treatment. The authors assert that perhaps the IV drug user, whose neurochemistry is pathologically affected by chemical dependency, is less to blame for developing complications than the hiker or cyclist. One could dispute the relative risks of participating in various activities, but this observation of inconsistencies in how physicians view risky behaviors is thought-provoking.
A specific example of changing treatment based on a patient’s lifestyle is the choice of valve type for young patients with IVDU-IE. We know of no formal studies addressing this topic, but anecdotally it is relatively common for young patients with IVDU-IE to receive a bioprosthetic valve. Based on age alone, they might achieve better long-term results with a mechanical valve, but such patients are often assumed to be “unreliable” regarding compliance with anticoagulants. When this assumption is correct, the patient may be spared from valve thrombosis, stroke, and early death, but those who could manage long-term anticoagulation are unfortunately consigned to suffer the sequelae of structural valve degeneration and probable reinterventions in the future.
A single-center study of surgical endocarditis outcomes noted that more than 95% of IVDU-IE patients received a bioprosthetic valve, despite an average age of 43 years in this cohort; by comparison only 73.7% of non-IVDU-IE patients, with an average age of 48 years, received a bioprosthetic valve [ ]. Despite the younger age of the IVDU-IE patients, the median postoperative survival was only 3 years, and there was no increased risk of reoperation in those patients. The authors concluded that a bioprosthetic valve is reasonable for IV drug users with endocarditis, regardless of age. Of note, none of the small number of IV drug users who received mechanical valves had valve-related complications.
Drug use contracts—are they useful?
The Merriam-Webster dictionary defines a contract as “a binding agreement between two or more persons or parties, especially, one legally enforceable” [ ]. Opioid contracts, or pain contracts, are formal, written agreements between prescriber and patient to outline ground rules for appropriate use. They are intended to discourage opioid abuse, but unlike traditional contracts, they are not binding or legally enforceable. For this reason, or owing to discomfort with the word “contract,” some physicians have taken to using the terms opioid or pain “agreement.” Agreements are drafted by a physician with “superior bargaining power,” however, and this power differential abrogates any semblance of collaboration or balanced consensus [ ]. Opioid contracts contain rules and stipulations regulating the prescription, filling, and consumption of opioids, and the consequences of breaking these rules, but little or nothing about goals of care—the execution of an opioid contract speaks to lack of trust or trustworthiness. While lack of trust of the patient by the physician may be warranted, a patient’s detection of such a lack may damage the therapeutic relationship.
Success has been reported with opioid contracts. A retrospective study of patients on long-term opioid contracts in an internal medicine practice found that 63% remained adherent to the terms of the contract during the study period. Of 37% who were not adherent, 20% stopped therapeutic medication voluntarily, and in 17%, a physician voided the contract for “breach of contract” [ ]. This study population consisted of patients with chronic pain who were not previously taking opioid pain medications. Drug addiction history is not reported, and it is not reported if any of the study participants were previous IV drug users. Unfortunately, whether this strategy prevents progression from chronic oral opioid to IV opioid use is unclear—random urine drug screens in this population are expected to be positive for opiates. For these reasons, the results of this study are probably not generalizable to the IVDU-IE population.
Opioid-use contracts may be applied to surgical therapies as well. The effectiveness of contracts to prevent recurrent drug use after valve replacement has not been established. In the previously mentioned population of chronic pain patients who had no known history of drug addiction, nearly 20% of patients were found to have breached an opioid contract [ ]. One might expect that the rate of breached contracts in a population of IVDU patients would be substantially higher. As Wurcel et al. editorialize, “Substance use disorders have a complex pathophysiology, influenced heavily by comorbid psychiatric illness and socioeconomic factors … It is unrealistic to expect that a signature on a piece of paper, in a time of medical extremis, will insulate the patient against recurrent drug use” [ ]. A response to that editorial argues that OUD is a fatal disease without treatment, and it is not unreasonable for surgeons to insist that the patient “agree to make a good faith effort to follow the treatment plan after the operation” [ ]. OUD is an important comorbidity to IVDU-IE, and must be treated to ensure the best possible long-term results. If a patient understands, to the best of their capacity, that IV drug use is the reason they have a life-threating valve infection, then it seems logical that they would want to stop using drugs. Arranging for adequate addiction therapy seems more likely to be effective than a contract to stop drug use.
The creation of a written agreement for postoperative addiction treatment does not seem inherently harmful or unethical. If used, the purpose of such a document should be to clarify the patient’s and physician’s shared goals, and focus on expected pitfalls and available resources. It should not outline punitive measures to be taken in the event of drug use, as this may make the patient less likely to seek help in the event of relapse. In addition, the physician should always acknowledge that any agreement between the patient regarding their drug use or addiction treatment is not legally binding. It is not ethically or legally defensible to deny a patient care based on such a document. As DiMaio et al. remind us, physicians’ obligation is to make medical decisions for their patients, not to pass moral judgment on them [ ].
Care of a “second life”—maternal versus fetal rights in cases involving pregnancy
Endocarditis during pregnancy is rare, appearing in only 1 of 8000 pregnancies [ ], but considerably complicates management owing to potential conflicts between maternal and fetal needs. Fetal mortality after cardiac surgery ranges from 16.7 [ ] to 30% [ ] and fetal mortality is much higher earlier in pregnancy [ ]. In comparison, maternal mortality is 2%–3%. Fetal organogenesis occurring in the first trimester suggests that surgery should be deferred until the second trimester [ ]. Waiting until the 28th week has also been suggested [ ], and operating late in the third trimester increases the risk of premature labor [ ].
The development of IE in the presence of IVDU during pregnancy opens the door to substantial conflicts between maternal and fetal interests. Firstly the future mother’s use of drugs places the fetus at risk, and vigorous attempts must be made to dissuade the patient from further drug use and to arrange optimal addiction treatment. Secondly what is beneficial to the pregnant woman is not necessarily beneficial to the fetus. If a pregnant woman develops IE during pregnancy, surgery should ideally be delayed until after delivery. Urgently required surgery during pregnancy imposes substantial risk on the fetus. A fetus delivered in the first trimester and early second trimester will not be viable and delaying urgent surgery will most likely result in the death of both woman and fetus, so surgery should be pursued with the understanding that the risk of spontaneous termination of the pregnancy is high in this setting. Surgery in later term pregnancies carry less risk of fetal death, so caesarean delivery of a viable neonate is a reasonable alternative, although preterm birth carries a considerable risk of mortality or long-term neurological problems [ ]. Ultimately a decision to operate must be individualized based on the age of the fetus and the acuity of the pregnant woman’s disease.
A pregnant woman may refuse to undergo an indicated procedure such as caesarean section despite the urging of her medical team. Fortunately, future mothers almost always act in the interest of their fetus [ ], but when they do not, they generate a conflict between their own interests and those of their fetus. Unless the patient lacks capacity to make decisions because of mental illness, she has a right to autonomy and to determine what actions may be taken on her body. The rights of the fetus, however, are less clear, and depend on local culture. In the United States, the rights of the woman supersede the rights of the fetus [ ], and consensus indicates it is almost never justifiable to force a procedure on a pregnant woman in the interest of the fetus [ ]. The best way to proceed in cases where maternal and fetal interests appear at odds may be to focus on fetal survival and health as a common goal for all parties involved [ ].
Specific considerations in patients with arrhythmia devices
Pacemakers and implantable cardioverter-defibrillators (ICDs) are medical devices that may be life-saving. They are also intravascular foreign materials that are susceptible to infection, especially in the context of IVDU-associated bacteremia. Literature on ethical issues encountered in the management of pacemaker and ICD infections in IV drug users is scarce. One report broaches the topic of ethics but draws only one conclusion: at the time of informed consent discussions, IV drug users should be told of the higher risk of device infection due to their IV injection [ ]. The report notes that management of an infected pacemaker or ICD is more straightforward than an infected prosthetic valve because removal of pacemaker leads or an ICD is simpler than reoperative valve replacement. The decision to reimplant leads may be more difficult. Patients who are pacemaker-dependent or have a history of sudden cardiac death certainly must undergo reimplantation. Infection may be less likely in IV drug users if epicardial or subcutaneous leads are implanted. In patients with less compelling indications for implantation—for example, a patient whose pacemaker was inserted for symptomatic bradycardia but no longer requires pacing—it is reasonable not to reinsert a device; such patients may be closely monitored and referred for addiction treatment.
Caring for “difficult” patients
Among IE patients, those with IVDU are among the most difficult to care for. As Buchman and Lynch state, “the hospital setting is often not a friendly one for people who use drugs” [ ]. Providers commonly view IVDU drug use negatively, and comorbid psychiatric disorders may lead to conflicts between patient and staff. Additionally, the treatment of IE with IV antibiotics often requires prolonged hospitalization, allowing time for frustrations to grow on both sides. Among the population with IE, IV drug users are significantly more likely to leave against medical advice (AMA) than nondrug users [ ]. Successful and complete treatment of IVDU-IE requires interventions focused on managing both patients and staff.
Negative attitudes toward IV drug users are pervasive among health-care workers [ ]. A report on patients refusing needed medical treatment notes, “In some cases patients who were viewed as ‘undesirable’ were permitted to refuse care in the hope that they would thereby be discharged more rapidly” [ ]. Patients with IVDU-IE fall into the “undesirable” category for many physicians, nurses, and ancillary staff, who may rather see the patient leave prematurely rather than continue to care for them. Such feelings may derive from unpleasant interactions between patient and staff, or from views that IVDU-IE is a self-inflicted disease. Staff may react to negative feelings toward such patients by acting out, delaying administration of pain medications, removing access to wanted items or devices such as television, and rapidly tapering methadone [ ]. Appropriate staff interventions may include supportive listening of frustrated personnel and holding regular staff meetings, especially with psychological counselors, as they can help staff understand the patient’s behavior and provide for improved communication.
Patients wanting to leave against medical advice
IV drug users are at high risk for leaving AMA during treatment for IE. A study of the disposition of all endocarditis patients over a six-year period reported that eight of nine IVDU-IE patients left AMA compared with none of the 17 nondrug-using patients [ ]. Some patients may be involuntarily admitted for the good of themselves or others; civil commitment is regulated by state law, including the allowable duration of commitment. In general, patients may be involuntarily held in the hospital if they represent a danger to themselves or others, and if they actively suffer from a mental illness, learning disability, or intoxicant that warps their interpretation of reality [ ].
It is generally in the best interest of the IV drug user with endocarditis to stay in the hospital and continue to receive treatment. The high rate of AMA discharges among this population may be caused by conflicts in the hospital or by a desire to return to drug use. If IV drug users wish to leave AMA, however, a conflict arises between the ethical principles of beneficence and autonomy: what is best for such patients is keeping them in the hospital, but doing so breaches their right to make their own decisions.
Because the process of civil commitment infringes individual liberties, the physician must consider legal ramifications of committing a patient. IV drug users have a high incidence of comorbid psychiatric disease, but laws regulating civil commitment would support involuntarily committing a patient only for the previously mentioned criteria [ ]. The presence of OUD in itself is not enough, and while it may be frustrating, the IVDU-IE patients who wish to leave the hospital AMA should not be stopped if they are of sound mind. Physicians should nevertheless thoroughly discuss the consequences of leaving AMA with the patient, and should question whether repairing some unmet but remediable need or underlying problem could persuade the patient to remain and complete treatment. A single-center study found that medications such as methadone, buprenorphine, and naltrexone were associated with decreased AMA discharge rates, but the difference in this underpowered study was not statistically significant [ ].
Special circumstances may allow commitment for substance use alone. The use of drugs during pregnancy is considered child abuse in 23 states in the United States, and is considered grounds for civil commitment in Minnesota, South Dakota, and Wisconsin [ ]. Therefore, in select circumstances, a pregnant woman may be involuntarily committed in order to ensure that she does not continue drug use, because this behavior places the life of the fetus at risk as well as her own.
The role of the physician in assessing cost and resource allocation
In this age of cost-containment, what is the role of the physician in allocating scarce or expensive resources? Often the demands of society and the needs of the individual patient diverge, and the physician may be caught in the middle. Moreover, health care organizations may pressure physicians to make decisions based on financial rather than medical considerations. IV drug users are often uninsured and not profitable for hospitals; preconceptions of their social worth cast doubt on their future productivity, and therefore the utility of offering them expensive medical and surgical therapy. Cost control in medical decision making is sometimes easy, as in changing brands of suture material, but deciding who deserves to undergo life-saving surgery is fraught with ethical dilemmas.
Is it ethical to deny surgery to an IV drug user because it is a limited resource?—Responsibilities of the physician to “society” versus the individual patient
IE is expensive to treat. IV drug users with endocarditis are often considered “undesirable,” owing to both the social aspects of their drug use and the likelihood of having no insurance. The Centers for Disease Control and Prevention recently reported that an average hospitalization for IVDU-IE patients costs about $50,000, and that 42% of these patients are either uninsured or are insured under Medicaid [ ]. This large percentage of unfunded or publicly funded patients shifts the economic burden of treating IE onto taxpayers and local health systems. Hospitalizations for IVDU-IE had increased 12-fold and hospital costs had increased 18-fold over their study period—only 5 years. The economic cost of illicit drug use in the United States was $193 billion annually, as reported in 2011, and is certainly much higher now [ ]. This figure includes health care costs and direct and indirect costs of criminal activity and lost productivity. The annual societal cost of heroin abuse specifically has been estimated recently to be $51.2 billion—about 20% of this amount is from loss of productivity [ ].
Given these facts, should patients who use IV drugs receive expensive medical care such as heart surgery? IV drug users who contract native valve IE and are otherwise operative candidates will almost always receive a valve replacement, but considerable controversy surrounds the question of performing repeat valve operations if drug use continues.
A decision not to offer a repeat operation may be made on grounds of societal and economic interests, or may be based on personal biases. Managing IVDU-IE is expensive, but failure to offer surgery to patients with IVDU-IE on the basis of economic costs makes little sense in the context of the economic cost of legal substances used in the United States. Excessive alcohol consumption in the United States is associated with an annual cost of $249 billion [ ]. Cigarette smoking is similarly costly—it is responsible for costs of more than $320 billion annually: $170 billion in health care costs [ ] and $150 billion in lost productivity [ ]. Both behaviors are self-harming and avoidable. Despite these high costs, no outcry is heard against treating the complications of drinking and smoking. The distinction between “acceptable” and “unacceptable” self-destructive behavior is difficult to make [ ], as is quantification of how much of the distinction rests on biases against social worth or on adequacy of insurance coverage and of the ability to pay for care.
A historical perspective of the economics of treating IE is instructive [ ]. Before the Medicare and Medicaid programs were established in 1965, indigent patients were treated on a charitable basis in both public and private hospitals. After 1965 denying a patient needed medical care based on their inability to pay has been considered to be unethical. Legal precedent has determined that the physician is the agent of the patient rather than of society in the context of delivering expensive medical care [ ].
Successfully operating for IE in IV drug users returns them to their community, where they may resume drug use, only to return with new complications and to incur further associated costs. If a patient is not treated and dies, future health care costs are prevented, but loss of the patient’s future productivity is also ensured. Decisions about resource allocation must be made at the social policy level, and not at the bedside [ ]. Physicians should not view their responsibility to society as including an imperative to weed out burdensome patients in order to save public funds. After treating the acute heart disease, physicians should strive to help these patients to obtain treatment for their second disease, drug addiction. In doing this, physicians provide the best care for their patients, satisfy the ethical imperative of medical professionalism, and return to society a person who has the best chance of returning to a productive life.
An important barrier to successfully treating IVDU-IE is the preconception that treatment is futile. Futility in the context of treatment of this disease can be interpreted two ways. Medical futility indicates that no medically obtainable goal is possible, so withholding treatment is ethically acceptable. Applying the concept of medical futility to IVDU-IE is complicated. Intervening on IVDU-IE could be considered to be ultimately medically futile because of poor long-term survival, but the immediate reality is antibiotics and valve replacement surgery are highly effective treatments for infective valve lesions, and short-term outcomes are often relatively good. “Psychosocial futility” has been applied to addiction and maladaptive behaviors that lead to a high likelihood of returning to drug use, and therefore life-limiting recurrent disease [ ]. Cardiac surgeons routinely assess operative risk, so analysis of short-term and long-term mortality and morbidity risks should be reasonably straightforward. Much more difficult is assessment of whether treatment is psychosocially futile, as IV drug users may be past the point where they can overcome their addiction.
Long-term survival of IV drug users with endocarditis—how does it stack up?
Short-term outcomes are often better in IVDU-IE patients than in non-IVDU-IE patients because of their younger age and lack of important comorbidities [ ], but long-term results suffer from high rates of reinfection in repeat drug users, which has been reported to be the leading cause of death in this population [ ]. Reinfection is not the only life-limiting complication of drug use, however, as drug users may die from violence or from drug overdose [ ].
A single-center series found that midterm mortality (6 months to 5 years) after operation for endocarditis was significantly worse in IV drug users than in nonusers: 53% and 31%, respectively, despite non-IVDU-IE patients being significantly older (60 vs. 38 years, mean) [ ]. Another single-center study found the median survival of IVDU-IE patients to be 3 years after operation, and was significantly lower than their non-IVDU-IE counterparts at both 5 years (46.7 vs. 71.1%) and 10 years (41.1% vs. 52%) [ ]. A multicenter study found long-term outcomes to be similar between IVDU-IE and non-IVDU-IE patients: 5 year (78.9 vs. 76.1%) and 10 year (69.5 vs. 68.7%) survival, respectively [ ]. That study had a greater age difference between groups than the preceding single-center study (23 and 5 years, respectively), which may account for the divergent findings in long-term survival. All three studies found that IVDU increased risk of reinfection, but only in the multicenter study was IVDU associated with an increased incidence of reoperation. Perhaps this indicates differing attitudes toward reoperation at each institution, which may affect long-term survival of these patients. Straw et al. found 5-year survival to be <50% in IVDU-IE patients [ ], and Shrestha et al. found 5-year survival to be 50%–60% [ ]. A meta-analysis including 19 studies also found that IVDU-IE was associated with worse long-term survival than non-IVDU-IE, reporting 5- and 10-year survivals of 62 vs. 70% and 56 vs. 63%, respectively [ ].
Studies of long-term survival of IV drug users who have undergone valve operations have found that, despite the young age of these patients, survival is not good—as high as 78% at 5 years [ ], but more commonly 40%–60% [ , , ]. Is this survival rate so poor that surgery for the IV drug user with endocarditis be considered futile? Consider the outcomes of surgical therapies for heart and lung failure. Left ventricular assist device (LVAD) therapy and lung transplantation both are associated with a 54% 5-year survival, yet these are considered neither futile interventions nor a waste of resources [ , ]. Both lung transplantation and LVAD therapy improve survival compared to medical management; no high-quality data exist, however, on whether valve replacement surgery improves the long-term outcomes of IVDU-IE. The reported mortality rate of surgically treated IVDU-IE is higher than of those treated medically, but these are retrospective data and probably are reflective of worse disease in the surgical group [ ]. The study that reported the highest rate of reoperation for reinfection had the best long-term outcomes [ ], suggesting that long-term survival may be improved with a more aggressive attitude toward reoperation for recurrent IE.
Difficulty determining which patients will stop using IV drugs
Perhaps the most challenging aspect of treating IVDU-IE is predicting whether an individual will refrain from further IV drug use. Research into medication-assisted therapy (MAT) has removed some of the guesswork by showing significantly lower relapse rates compared with other therapy [ ], yet using alternative opioids or opioid antagonists seems not to have been widely embraced by the surgical community. The data may not be widely known, although methadone has been effectively used to treat heroin addiction since the 1960s [ ]; more likely, the stigma surrounding OUD has led to reluctance to prescribe these medications, and access to addiction specialists may be limited. Still, relapse is the most important driver of reinfection and mortality, and IE is closely associated with OUD, not an independent process. For these reasons, combining the surgical treatment of endocarditis with the medical and psychiatric management of addiction is mandatory for ensuring good long-term outcomes.
Treatment of dual disease—valve disease and addiction
The need to treat the IV drug user with IE with dual-disease therapy is being increasingly recognized. The authors of the 2016 American Association of Thoracic Surgery consensus guidelines for the surgical treatment of IE make the Level 1 recommendation that “patients with a history of injection drug use should be treated for their addiction” [ ]. OUD is now recognized as a medical disease, and therefore the treating surgeon should consider addiction a comorbidity rather than a character flaw. Cardiac surgeons recognize the importance of treating comorbid hyperlipidemia and diabetes in patients undergoing coronary artery bypass grafting; in the same way, MAT gives patients with IVDU-IE the best available treatment for OUD to optimize long-term outcomes.
Rates of referral for addiction treatment
Unfortunately, addiction is undertreated in the IVDU-IE population. Inpatient addiction services are a powerful tool, and initiation of inpatient MAT increases rates of compliance with outpatient MAT and decreases rates of illicit opioid use [ ]. Only 24% of patients with IVDU-IE admitted to one medical center received inpatient addiction or a psychiatry consult, and even worse, only 7.8% were discharged with a plan for MAT, and none was discharged with medication for the treatment of addiction [ ].
Abstinence only versus medication-assisted therapy
Current addiction therapies comprise a spectrum of medical and psychiatric therapies. Many medications have proven effective in curbing drug cravings and helping prevent relapse. Additionally, cognitive-behavioral therapy and other psychiatric methods help patients to use drugs rationally and build coping skills.
MAT is the use of alternative opioids or opioid antagonists such as methadone, buprenorphine, and naltrexone to prevent relapse into high-risk IV drug use. This contrasts with the “abstinence only” approach consisting of either no treatment, in the hope that the patient will stop drug use on their own, or psychiatric methods, such as cognitive-behavioral therapy. Abstinence only remains a prevalent approach in the United States, even though it does not adequately address OUD as a medical disorder. Additionally, it has been shown to have higher rates of recurrent OUD than MAT.
The history of MAT starts with Dole and Nyswander’s landmark 1965 study, in which they reported the successful treatment of 22 patients with heroin addiction using methadone [ ]. All patients reported decreased “narcotic hunger,” and were able to refrain from further IV drug use and to return to working, productive lives. Alternative drugs are also efficacious. In a randomized controlled trial of buprenorphine versus placebo, patients in both arms of the trial underwent cognitive-behavioral therapy and weekly counseling sessions, and the outcome difference at 1 year was stunning—75% of patients receiving buprenorphine remained in treatment compared with none in the placebo group [ ]. In another trial, extended-release naltrexone (XR-NTX) compared with placebo showed significantly less opioid use in the XR-NTX group (99% of days free of opioid use vs. 60% in placebo group), as well as longer retention in treatment (168 vs. 96 days, mean) [ ]. Another study of XR-NTX, specifically in US criminal justice offenders, found lower rates of relapse in patients on XR-NTX (43% vs. 64%), and lower rates of overdose (0% vs. 4.5%) [ ].
MAT is clearly superior to abstinence-only approaches for treating OUD. Nevertheless, MAT has not been widely used to treat addiction: only 34% of patients receive MAT for OUD [ ]. Barriers to widespread MAT use exist at the individual, infrastructure, and policy levels. The stigma attached to drug use is widespread, and many individual providers view MAT as merely substituting one addiction with another. Both patients and providers commonly have inadequate knowledge of OUD and MAT [ ]. Lack of qualified addiction specialists may limit access to medications, as generalist practitioners may not feel comfortable managing these drugs. An extensive survey to determine barriers to adoption found that access to MAT varies widely on a state-by-state basis [ ]. States with good infrastructure, specifically a suitably sized workforce of addiction specialists and adequate funding, had higher rates of MAT adoption. States that had a large investment in 12-step programs and other “social detoxification” programs were highly resistant to MAT adoption. Many of those states do not allow the use of methadone for OUD outpatients. Moreover, policy-makers and legislators may harbor biases regarding MAT, which can substantially hamper adoption in those states.
Harm reduction strategies
Harm reduction is a strategy aimed at reducing the negative consequences of drug use in order to make it safer, rather than preventing or discouraging drug use [ ]. Contemporary examples include needle exchange programs (NEPs) and safe injection teaching. While many would view the ongoing use of drugs in any capacity distasteful, this nonjudgmental approach to drug use may lead to gradual progress to abstinence. The patient is integrated into the health care system, where a therapeutic relationship can be built. Conversely, attempting to force a patient into an abstinence or MAT program before they are ready risks losing trust with the patient and perhaps losing contact until some catastrophic consequence of drug use has occurred. Much data suggest that harm reduction is effective. MAT is technically a form of harm reduction, but for the purposes of this section, we will focus on injection hygiene teaching, safe injection sites (also known as supervised injection sites), and NEPs.
A randomized controlled intervention to teach needle hygiene and safe injection practices found large reductions in unsafe practices and a lower risk of bacterial skin infections in those who underwent the intervention [ ]. NEPs allow return of used needles in exchange for new needles. In theory this reduces the risk of infectious complications such as human immunodeficiency virus, hepatitis C virus, endocarditis, and other bacterial infections. Safe injection sites allow patients to safely and hygienically inject in the presence of a medical professional, instead of in public or in other sites designated for IV drug use, both of which increase risk of infection [ ]. Several studies have documented the safety and effectiveness of NEPs and safe injection sites, documenting decreased rates of overdose, reuse, and sharing of needles, and public use or discarding of needles [ ].
NEPs are controversial in the United States, but have become more commonplace. As of 2018, 39 states plus Washington D.C. and Puerto Rico have passed laws allowing needle or syringe exchange programs [ ]. To be successful, the concept of needle exchange must be generally accepted and law enforcement must be cooperative. If officers of the law stake out needle exchange sites and use them as an opportunity to easily arrest drug users, the NEP cannot serve its intended purpose. Safe injection sites are fundamentally different than NEPs, in that they are not just providing equipment, but actually a location for illicit drug use to occur. Safe injection sites have been used successfully in Europe and Canada for years, but they are currently illegal in the United States [ ].
Ethically, harm reduction programs are consistent with the principles of autonomy and nonmaleficence. The choice of the user to continue drug use is respected, and the risk of harm reduced. The principle of beneficence may be violated in the sense that continued IVDU is a self-harming behavior, or to the contrary harm reduction therapies may lead to participation in MAT or abstinence therapy, which would be of great benefit; on balance harm reduction programs seem mostly beneficial. Legally, justice is violated in condoning and facilitating the use of illegal drug: drug users are still being pursued and prosecuted in many jurisdictions. Furthermore, harm reduction programs may sustain contact with health professionals and build rapport, which may lead to cessation of illegal drug use. Ethically these programs are consistent with the principle of justice because they exist not for the sake of the drug use, but for the safety of the user, who is respected equally with all those who seek to improve their health. No major ethical objections to harm reduction are evident.
Peripherally inserted central catheters in harm reduction
The outpatient use of peripherally inserted central catheters (PICC lines) in patients who use IV drugs is controversial. PICC lines are a standard tool in the treatment of severe bacterial infections, and they allow for outpatient treatment with IV antibiotics, freeing up inpatient beds and resources and reducing known risks associated with hospitalization. A major concern about placing a PICC line in an IV drug user is that such lines provide a convenient site for injecting drugs, so the prevailing practice is not to discharge IV drug users from the hospital with PICC lines for prolonged IV antibiotics administration. A study of high-risk drug users using PICC lines reported a relapse rate of 40%, even though they were not discharged home, but rather relapsed into IV drug use in either the hospital or nursing home, underscoring the severity of their addiction [ ]. The use of tamper-deterrent PICCs in outpatient IV drug users resulted in successful completion of outpatient antibiotics with no evidence of PICC tampering for drug use in 96% of patients [ ]. The criteria for inclusion in this study were relatively stringent, including a “no tolerance” policy toward PICC tampering.
Outpatient PICCs can be used as a harm reduction strategy; such lines can help facilitate hospital discharge, sparing valuable hospital resources [ ]. Addiction is recognized as a “bumpy road” in which a patient may have periods of compliance and relapse. The patient should be taught safe injection practices so that relapse would be associated with lower risk of complications.
Recurrent endocarditis and repeat operations
Perhaps the most controversial and well-documented topic in the care of IVDU-IE patients is whether to offer repeat valve surgery to the patient who has undergone valve replacement, starts again to use IV drugs, and returns with a prosthetic valve infection. Setting a limit on the number of repeat valve replacements is a relatively common practice [ ]. Some surgeons offer one valve replacement and refuse additional treatment if the patient uses IV drugs again and returns with an infected valve, while others will reoperate as many times as a procedure is surgically feasible; many surgeons use a more nuanced approach. Consideration of what is the right thing to do in such cases requires maintaining the proper context:
aiming for an objective evaluation and decision,
avoiding emotional influences and moral judgments,
considering the therapeutic needs of two separate but related diseases, OUD and IE, and
incorporating the values and the desires of the patient.
Yeo et al. argue there is no ethical duty to reoperate in the noncompliant continuing drug user who develops recurrent IVDU-IE [ ]. They invoke first the argument of autonomy, claiming that patients are ultimately responsible for their actions and health, and secondly of justice, stating that patients had prior opportunities to improve their health, and that others deserve access to limited resources as well. Their focus is on population health more than individual health. They find the treatment of recalcitrant drug users to be contrary to optimizing population health because these patients do not take agency for their own health, overutilize scarce resources, increase crime rates, and are often unfunded, creating a further drain on resources. While they acknowledge that medical decisions cannot be based on ideas of social worth, they assert that these patients have broken the “social contract.” They support the idea of health care rationing at a national and societal level, but discourage physicians from bedside rationing, stating “Physicians must always remain the patient’s advocate.”
Kirkpatrick points out that the primary payors for care are either government agencies or insurance companies, and neither has decided to ration care based on behavior or lifestyle [ ]. Like Yeo et al. he believes that it is not the place of the doctor to decide on their behalf. The duty of the doctor is simply patient care.
Buchman and Lynch argue that relapsed users should receive a second valve replacement if they express a desire to stop using drugs [ ]. They frame their approach to drug relapse in the mindset of harm reduction, describing addiction as a chronic illness characterized by periods of relapse, claiming it is unrealistic to expect IV drug users to abstain indefinitely, and recommending that teaching safe injection practices be included as part of a harm reduction strategy.
Hull and Jadbabaie take a middle road, attempting to balance the duties of the physician to the patient and to society [ ]. The authors state that physicians should not make medical decisions on the patient’s perceived social worth or moral failings, at the same time being “good stewards” of medical resources so as not to limit access to care. The crux of decision-making lies in assessing the likelihood of success or failure of surgery and addiction treatment in individual situations. They agree with others that physicians must make arrangements for postoperative substance rehabilitation. Ultimately, they offer the solution of giving the patient “three strikes,” which is more than the “one and done” approach, but still sets a limit, presuming that a patient who has suffered through two valve surgeries and still continues IVDU and returns with another infection cannot kick the habit. Choosing this approach requires committing great amounts of resources to ensure that these patients get optimal addiction treatment.
Miljeteig et al. present an interesting ethical analysis that provides a systematic and unbiased approach to the recurrent IV drug user, and includes evaluating the patient’s knowledge base, parties involved, preexisting regulations and recommendations, benefits, conflicts of interest, and ethical principles involved [ ]. Their review of the data found that repeat valve surgery for recurrent IVDU-IE adds an average of 1–2 years of life if drug use is continued, and possibly much more if drug use is discontinued. They perform surgery for other diseases in which they expect 1–2 years of survival, and conclude that they should offer valve surgery to recurrent users. It may be ethical to refuse treatment, in their view, if conditions exist that would prevent treatment from having benefit. Previous lifestyles should not preclude or change treatment choices, but predictable future lifestyles may. While the authors advocate for repeat valve surgery, they clearly allow for the possibility that some cases of drug addiction may be so severe as to make surgery futile.
The various arguments for and against a second operation are comprehensively captured in a debate between DiMaio and Salerno [ ]. DiMaio argues that surgeons may refuse a second operation, citing the high mortality of drug users, the risk of infectious disease to the surgical team, resource allocation, and the professional obligation of surgeons to not perform treatments they believe to be futile. Salerno presents opposing viewpoints, calling on the perceptions of several other professionals, including a psychiatrist and lawyer. The psychiatric view highlights the medical implications of the OUD diagnosis and emphasizes the necessity of addiction treatment and a multidisciplinary approach. From a legal standpoint, treatment contracts are not legally enforceable and are not a justifiable reason to deny treatment. Salerno asserts that surgeons’ duty is to treat their patients, not to judge them.
Deciding whether to reoperate
Repeat users with IE present clinical and ethical challenges. Many are gravely ill and have advanced disease requiring urgent surgery and extensive reconstruction. After the first valve replacement for IVDU-IE, they have nearly always been instructed to stop drug use and now have failed to do so. Additionally, they may be emotionally challenging patients whose personalities clash with those of the staff.
The most important first step for a surgeon in deciding whether to reoperate on these patients is to understand and acknowledge how much a part their own personal biases play into the decision. While reoperative surgery is higher risk and more technically complex than the first operation, many of these patients are young and have a life expectancy of many years, even with continued drug use. In the absence of IV drug use, most surgeons would not hesitate to operate. Former behavior should not dictate treatment, especially if the patient is receptive to undergoing addiction treatment aimed at stopping drug use. Prior treatment contracts to refrain from drug use should be considered neither binding nor a reason to withhold treatment, as that would not be ethically or legally defensible. Surgeons who will not operate on IVDU-IE patients whom others might consider operable, should not turn them away without treatment, but rather should offer to transfer the patient to another surgeon who would independently consider operating.
A neurochemical pattern usually underlies drug-seeking behavior—recognizing and accepting this can relieve frustration in caring for addicted patients. Ultimately they need help with a problem, and they are present because they sought help. Personal biases must be put aside. The medical aspects of OUD must be considered, including the adequacy of previous treatment of IVDU-IE. The primary obligation of physicians is to treat the patient in front of them, not to arbitrate expenditures for insurance companies or government agencies. Prosthetic valves are expensive, but they are not a limited resource in the same sense as transplantable organs. Some cases of recurrent IVDU-IE are futile, owing to the patient’s rejection of addiction treatment, medical futility, or prohibitive risk of operation. In such cases, it is ethically defensible to refuse treatment. But for patients who have a reasonable probability of surviving operation and wish to undergo treatment of OUD, it is always defensible to operate, regardless of past behavior.
The treatment of IVDU-IE is fraught with ethical dilemmas, which mostly arise from three major sources: the self-destructive nature of IV drug addiction, social and professional biases toward IV drug use, and psychiatric problems associated with addiction.
IVDU-IE is not the only disease caused by self-harming behavior, but the complications of IV drug use are severe, and the life expectancy of IV drug users is substantially lower than others; nevertheless, long-term survival is possible. The primary driver of early mortality is relapse into drug use, but determining which patients are likely to abstain from further drug use and which will relapse is difficult, so assessing treatment options is challenging.
Physician attitudes toward IV drug users are often negative, and many find it difficult to care for patients with IVDU-IE. In addition, this population has a large burden of comorbid psychiatric issues that can make their care even more frustrating. Their associated psychiatric conditions are often undertreated. Patients with IVDU-IE have a high risk of leaving AMA, owing to their desire to resume drug use, psychiatric factors, or conflicts with staff; while the best interest of some such patients may lie in involuntary commitment to hospitalization, this extreme measure is available in most jurisdictions only if the patient is psychotic, delirious, suicidal, or homicidal.
Beyond medical and surgical issues, economic and legal aspects of IE are important and complex. The high cost of care for IE and appropriateness of resource allocation are often considered in the context of a general impression that these patients are of low social worth. Physicians find themselves in a difficult position between patients and society, sustaining external pressures to control costs while understanding that their primary responsibility is to care for their patients. In our view, the most ethically well-grounded position for physicians to take is to remain an advocate for their patients, leaving cost considerations to others, because the physician’s trustworthiness and the patient’s trust in the physician and in the medical profession are critically important to the healing relationship.
From a legal standpoint, the use of treatment contracts is not advisable because they detract from the rapport of the patient–physician relationship, are unenforceable, and are not a legally or ethically defensible reason to withhold appropriate treatment.
The overall care of the patient with IVDU-IE is best viewed in the context of their OUD. Physicians are human and as such have both intrinsic and extrinsic biases, but however they view various lifestyles, their privilege and obligation is to treat medical disorders. Moving from a view of addiction as a moral failing to that of OUD as a complex medical disorder helps to remove stigma and facilitate an objective view of these patients and their problems. IVDU-IE is a consequence of addiction in these patients, and operating on the valve disease while ignoring or undertreating the underlying OUD is a serious error—treating IVDU-IE requires aggressive therapy of two distinct but related pathologies. MAT has been proven effective in reducing drug use, and for patients who are not ready to abandon their addiction, harm reduction strategies may be helpful in reducing complications of drug use and the resultant medical costs.
Patients who present with recurrent IVDU-IE after initial valve surgery are usually not beyond help. Referral rates for addiction treatment in these patients are low—many have not received adequate addiction treatment after their first valve replacement. Evaluation of these patients should focus on long-term life-expectancy and willingness to accept treatment for their addiction. The decision whether or not to reoperate for recurrent IVDU-IE should be based on the technical feasibility of the surgery and the expected benefit of the treatment rather than on past or current drug use, especially if the patient has not received adequate addiction treatment in the past, because such patients may have better success with high-quality follow-up and multidisciplinary treatment.
Acceptance of OUD as a medical disorder and recognition of the psychiatric comorbidities of OUD are growing, so we can hope to progress to a world where IVDU is no longer stigmatized. Substantial barriers currently exist to delivering the best care to patients with IVDU-IE, but many of them are cultural and those can be overcome with greater insight and understanding.