Discussion

Infective endocarditis developing after device-mediated patent ductus arteriosus closure is rare. However, an early report on the natural course of patent ductus arteriosus emphasised that infective endocarditis/endarteritis is an important cause of mortality.^{Reference Campbell3} Thus, many centres prefer to close even small patent ductus arteriosus. Currently, percutaneous device-mediated closure is the first-line treatment. Although the operation is minimally invasive, complications including infective endocarditis, although rare, can develop. The infection may be acquired during device placement, and infection of the soft tissue surrounding the device can occur after placement.^{Reference Edelstein and Yahalom4} Our most remarkable finding was that the device was clearly visible on echocardiogram when the patient visited the emergency room, which was probably attributable to inflammatory oedema in the surrounding tissue enhancing the ultrasound image. This information should be included in the text of the Duke criteria for diagnosing infective endocarditis; currently, endocarditis is often missed when diagnosis is carried out according to these criteria. In our case, although the clinical and laboratory findings indicated endocarditis, the Duke criteria indicated that this was only a possibility. Some reports have suggested that the Duke criteria are less sensitive in cases of prosthetic valve endocarditis or cardiac device infections.^{Reference Nishimura, Otto and Bonow5,Reference Pérez-Vázquez, Fariñas and García-Palomo6} In our opinion, the Duke criteria should be revised to cover patients with indwelling prosthetic materials.

Three case reports have described infective endocarditis developing during follow-up in patients with residual shunts after patent ductus arteriosus device implantation.^{Reference Latson, McManus, Doer, Kilzer and Cheatham1,Reference Campbell3,Reference Edelstein and Yahalom4} In contrast, our case exhibited no residual shunt on echocardiography. Device-related infections are usually resistant to antibiotherapy, unlike native endocarditis.^{Reference Edelstein and Yahalom4,Reference Habib, Lancellotti and Antunes7} The therapeutic options for device-related endocarditis have received scant attention^{Reference Saint-Andre, Iriart, Ntsinjana and Thambo8} and most patients heal after surgery. Nevertheless, the device may serve as the source of infection, so surgical removal is essential. Peng et al^{Reference Peng, Yang and Liu9} suggested that inappropriate device implantation, associated with turbulent blood flow via a residual shunt, may injure the vascular endothelium. However, infective endocarditis does not develop in most patients with residual shunts.

We found no literature pertaining to the prevention of infective endocarditis associated with transcatheter device-mediated closure. However, the general guidelines for asepsis and antisepsis must be rigorously followed during interventional procedures. A 2015 guideline on intracardiac electronic device placement recommended that the air exchange rate in cardiac catheterisation rooms should be equal to that recommended for operating theatres, that is, 25 changes per hour rather than the currently recommended 15 changes per hour.^{Reference Sandoe, Barlow and Chambers10}

In conclusion, infective endocarditis is a rare but very serious complication of percutaneous device-mediated closure of heart defects. The risk of endocarditis or endarteritis associated with patent ductus arteriosus closure devices remains unknown. Patients with early or late high fever should be carefully evaluated in terms of infective endocarditis.