Open ArchivePublished:April 08, 2016DOI:
        Dr Gosta B. Pettersson (Cleveland, Ohio). I would like to thank the Association for the honor to discuss this important article related to this epidemic of drug abuse. Thank you and congratulations on a nice presentation and for providing the article in a timely fashion.
        This article addresses 1 important complication of drug abuse: drug abuse presently accounts for more than a quarter of all surgical endocarditis cases at Mass General and Brigham and Women's Hospital in Boston. This is a scary trend, and in their introduction the authors question the utility of aggressive resource-intensive surgical procedures for these patients. This is a moral, ethical, and medical dilemma. Moral and ethical if we look upon drug abuse as a self-inflicted problem and medical if we look at drug abuse as just another comorbidity; for example, comparable to morbid obesity.
        The authors clearly show that drug abuse patients, who are younger and have low surgical risk, return with recurrent infection and they do have a decreased late survival, but survival is still 81% and 72% at 5 and 10 years, respectively, and we certainly treat many patient groups with worse survival without hesitation.
        Although we don't know the exact number, we all believe the majority of the patients return to drug abuse. In this study, 37 patients returned with recurrent endocarditis within 1.5 years, and most of these reinfections were caused by a new organism different from the first, suggesting that they were new injects. We all understand that their long-term outcome is related to our ability to treat their comorbidity more than to our ability to treat their endocarditis—long-term success is all about prevention of continued drug abuse.
        At the Cleveland Clinic we have also recognized the drug abuse epidemic, and we have recently reviewed our experience with drug abuse-related endocarditis for the period 2007 to 2012, corresponding to the second half of the study period of Dr Kim's study. Our experience for this 5-year period includes 41 cases, corresponding to 8% of our total endocarditis patients.
        Although our experience is similar to the Boston experience, there are some notable differences between Cleveland and Boston. Heroin was an even more dominating drug in Cleveland than in Boston; 63% of our patients had tricuspid valve involvement compared with only 25% of the Boston patients. And more of our patients had Staphylococcus aureus infections: 59% of ours compared with 38% in this study.
        We also noticed an interesting thing with regard to the risk of death after surgery. The hazard ratio within the first 90 days was lower for the drug abuse patients, but between 90 and 180 days the hazard ratio was actually 9.8 that of nondrug abusers, and then it stabilized at 1.8, which is very close to your observation.
        My first question is related to the follow-up and the reliability of the Social Security Index to confirm vital status and dates of death. This question is relevant to several studies presented during this meeting. On November 1, 2011, some 40% of the Social Security death master file was deleted from public use by the Congress. According to our understanding, the Social Security Death Index is unlikely to disclose all deaths, and because the data capturing is related to exhausted Social Security benefits rather than death, the Index is more likely to be incomplete for younger patients. Drug abuse patients are younger and notoriously difficult to track.
        Are you still convinced that your follow-up with regard to vital status is really that 100% as you claim or do you agree that we might even have underestimated the number of deaths?
        Dr Joon Bum Kim (Boston, Mass). Thank you, Dr Pettersson. Your comment is very important and has to be addressed appropriately.
        We have been aware of the potential inaccuracy of the Social Security Death Index in the identification of mortality status; however, this was our best research in our database system, and the Social Security Death Index was searched only if the patient showed missing data on vital status. Fortunately, our Partners Center database has some accurate data on mortality status. So most of the patients were covered by our Partners Center registry, and I remember only 30 patients needed to be looked up in the Social Security Death Index. So I think our data is pretty good in terms of mortality identification.
        Dr Pettersson. Do you know what your patients died from? Did they die from drug overdose, endocarditis, or something else?
        Dr Kim. Out of 78 intravenous drug abusers, there were 3 early deaths and 11 late deaths, and we could identify the causes of death in 8 patients. They were reinfection in 6 patients, 1 patient died of lung cancer, and 1 died of intracranial hemorrhaging; 3 were unidentified. Of these, 1 found dead while he was using heroin very actively. So we think that 3 patients died from heroin use, but most of our patients died from reinfection among these patients.
        Dr Pettersson. My third set of questions relates to our observation that 68% of our drug addict patients had hepatitis C, a risk not only to the patient but also to the surgical team. Do you know your prevalence of hepatitis C among your patients? We don't know if some of them also have HIV. Do you test for HIV? What extra precautions do you take in the operating room when handling these patients?
        Dr Kim. Thank you, Dr Pettersson. I have to acknowledge my ignorance when I began researching this issue and we did not include hepatitis C and HIV data for this work, but at the end of manuscript completion I came to realize that hepatitis C and HIV are potentially very important to the outcomes.
        I believe almost all patients received testing for hepatitis C and HIV, because these patients are surgical patients. As far as I remember, although I can't present very precise data, hepatitis C viral infection was more prevalent in these patients—let's say 20% to 30% of patients—but I don't remember many patients who had HIV infection among this population.
        Dr Pettersson. So my next question is most important. How do you address their fatal disease: Drug addiction? Do you have any conditions for operating on them?
        Dr Kim. It is a very challenging question. As far as I believe, the institutional priority of Mass General Hospital is to have some aggressive drug rehabilitation therapy collaborating with other providers of medical care, but I am not sure whether there already are very systematic approaches across the Brigham and Women's and Mass General as well as other individual surgeons' experiences. But I am sure that we need something more than just offering surgical therapy, because we have to eradicate the potential infection sources, let's say, the intravenous drug abusers. So collaborating with the other providers like rehabilitation and psychiatric therapists is very important, I believe.
        Dr Pettersson. Do you ever refuse to operate?
        Dr Kim. I think there are patients who are refused by physicians, but I am not sure, because our data includes surgical data by more than 15 independent surgeons. So I cannot say the uniform strategy related to those patients.
        Dr Pettersson. What do you think about the actual surgical procedure here? What do you do to the tricuspid valve? What do you think about, say, valve replacement in the tricuspid position? What do you think about mechanical valves in these patients? What do you think about pacemaker leads?
        Dr Kim. I think the principle of infective endocarditis surgery is complete removal of infection first. If possible, repair would be the best choice whenever amenable. If you can get rid of infection sources and there is some room for repair, I think repair is probably the best choice, especially for the tricuspid position. But if not, we don't hesitate to replace then. Valvulectomy was not our practice across the Brigham and Women's Hospital and Mass General Hospital.
        I am not sure whether these drug users are compliant on warfarin therapy, although surgeons had very individualized approaches for every patient whether they can really deal with long-term lifelong warfarin therapy.
        Dr Pettersson. Thank you, and once again, congratulations, and congratulations on your excellent defense of your paper.
        Dr Kim. Thank you so much.