Management of anticoagulant therapy using a portable point-of-care international normalized ratio device and social networking service in a patient with a left ventricular assist device

Abstract

Device infection and stroke are still frequently reported as complications of left ventricular assist devices, and strict management of anticoagulation therapy is sometimes difficult at the time of infection status. We report the case of a 55-year-old man with a HeartMate II (Abbott, Inc., Abbott Park, IL, USA) as a bridge to cardiac transplantation. The patient measured his prothrombin time-international normalized ratio (PT-INR) by himself using a point-of-care device at home and reported the result promptly on a social networking service (SNS). Physicians instructed the patient on how to adjust his dose of warfarin based on the result and suggested the next time of measurement on the SNS. Until cardiac transplantation, we adjusted the dose of warfarin 106 times using the SNS because of unexpected PT-INR fluctuations caused by antibiotics. The time in the therapeutic range was maintained at 83.2% without complications, including major bleeding, stroke, or pump replacement; however, there was transient intra-pump thrombosis triggered by severe dehydration due to hyperthyroidism.

<Learning objective: The strict control of the prothrombin time-international normalized ratio (PT-INR) is crucial to avoid complications in patients with left ventricular assist devices. However, this is sometimes difficult during infections. Although the self-measurement of PT-INR using a point-of-care device is recommended by the guideline, it is not clear who should adjust the dose of warfarin based on the results. A social networking service may help physicians to decide on the warfarin dose without delay, and this should complement monthly hospital visits.>

Introduction

Continuous-flow left ventricular assist devices (CF-LVAD) are increasingly used to support patients with advanced heart failure, both as a bridge to transplantation and destination therapy [1].

However, the incidence of serious complications such as device infection and stroke is still high. Anticoagulation therapy is sometimes difficult at the time of infections, so strict control of the prothrombin time-international normalized ratio (PT-INR) is crucial to avoid complications associated with the destabilization of anticoagulant therapy such as pump thrombus, major bleeding, and stroke. Although the self-measurement of PT-INR using a point-of-care (POC) device by patients at home is recommended by the guideline [2], it is not clear who should adjust the dose of warfarin based on the results of this measurement. Therefore, we evaluated the usefulness of a social networking service (SNS) on PT-INR control to complement face-to-face monthly hospital visits.

Case report

A 55-year-old man (height, 159.9 cm; weight, 57.4 kg) was implanted with an LVAD (HeartMate II, Abbott, Inc., Abbott Park, IL, USA) for advanced heart failure due to ischemic cardiomyopathy as a bridge to cardiac transplantation. Although he had cerebral infarction with neuropsychological disorder after the LVAD implantation, he was able to manage the LVAD by himself.

After 199 days of the LVAD implantation, he developed an infection at the driveline penetration site, 280 days after implantation he had methicillin-resistant Staphylococcus aureus bacteremia, and abscess around the outflow graft to the ascending aorta. We sequentially administered a drip infusion of vancomycin, linezolid, and daptomycin. As a result, the infection was stabilized, and pump replacement was avoided. After de-escalation of the drip infusion of antibiotics, he received oral sulfamethoxazole-trimethoprim and minocycline daily. Additionally, he needed to wash the driveline penetration site with sterilized water three times a week to maintain the remission. Although temporary PT-INR fluctuation was observed after the start of daily antibiotic use, thereafter, the in-hospital management was stabilized, but the PT-INR increased again to 4.0 at the time of first trial home management. Because of the fluctuation of PT-INR probably due to the antibiotics, we instructed the patient to self-measure his PT-INR using a POC device (Coagucheck® XS personal, Roche Diagnostics Corporation, Mannheim, Germany) (Fig. 1). After obtaining patient consent, we used a medical SNS (Medical Care STATION®, Embrace Co., Ltd., Tokyo, Japan) to share the results of all measurements with his family, attending physicians, hospital medical staff, visiting nurses, and collaborative hospital staff. We started the POC device and SNS support at the first discharge, 465 days after LVAD implantation.

Fig. 1.

The point-of-care device (Coagucheck® XS personal). The patient posted the prothrombin time-international normalized ratio values on the social networking service after measurement using photos and sentences.

The patient posted his PT-INR values on the SNS immediately after measurement using photos and sentences, and physicians selected the warfarin dosage and the date of the next measurement according to the PT-INR values and the patient's condition. Even if the physician found it difficult to respond, a substitute physician could reply and respond as a team. The physicians determined the target PT-INR according to the International Society of Heart and Lung Transplantation (ISHLT) guideline [3] and calibrated the POC device by measuring the PT-INR simultaneously in the hospital laboratory at every monthly hospital visit. The SNS conforms to the security standard recommended by the Japanese Ministry of Health, Labor, and Welfare guidelines for implementation of telemedicine.

During the waiting period for cardiac transplantation, there were some periods of prominent fluctuation of the PT-INR. The first period of fluctuation was 756 days after the LVAD implantation. Pump thrombosis was suspected because of an increase in serum lactate dehydrogenase (LDH) levels. Therefore, we changed the target range of PT-INR from 2.5–3.5 to 3.0–4.0 (Fig. 2, arrow 1), according to the ISHLT guideline [3]. We administered a drip infusion of serine in our hospital (Fig. 2, arrow 2), and the LDH levels improved within a few days. The time in the therapeutic range (TTR), used as a measure of anticoagulation control by calculating the percentage of days where the INR values are within a target range (Rosendaal method) [4], was 28.3% to 36.3% (Fig. 2) for 1 month before the suspected thrombotic event. After that event, the patient needed oral administration of thyroxin for hypothyroidism without an obvious trigger. Two months later, we discontinued hormone replacement with normalization of thyroid function. However, the patient became dehydrated with frequent diarrhea due to thyrotoxicosis associated with subacute thyroiditis, the PT-INR was prolonged, and LDH levels increased again. Pump thrombosis was strongly suspected, and the patient was admitted to our hospital (Fig. 2, arrow 3). We started a continuous drip infusion of unfractionated heparin, increased the LVAD pump speed, optimized the thyroxin dose, and administered a continuous drip infusion of dobutamine and a phosphodiesterase-3 inhibitor for suspected right heart failure. The result of these interventions was a decrease in LDH levels without pump replacement. The TTR during the second event was 91.7% (Fig. 2).

Fig. 2.

The changes in prothrombin time-international normalized ratio (PT-INR) and time in the therapeutic range (TTR) over time. Fig. 2 shows the PT-INR values and target ranges from the left ventricular assist device implantation to heart transplantation. The TTR in each outpatient period is shown in the upper part of the graph. The TTR was 28.3% for 1 month prior to changing the target range due to suspected thrombus, and 36.3% for 1 month before hospitalization due to suspected thrombosis. The pre-hospital TTR before the second thrombus was 91.7%. Arrow 1: Changed PT-INR target range from 2.5–3.5 to 3.0–4.0. Arrow 2: Admission, suspected pomp thrombus. Arrow 3: Admission, suspected pomp thrombus. DAP, daptomycin; SMX/TMP, sulfamethoxazole-trimethoprim; VCM, vancomycin; MINO, minocycline.

The physicians changed the warfarin dosage 106 times according to the 242 patient-reported PT-INR on the SNS until cardiac transplantation. Despite prominent fluctuation of the PT-INR, the TTR during the entire waiting period for cardiac transplantation was 83.2 %. Successful cardiac transplantation was performed three years after LVAD implantation.

Discussion

Although the incidence of pump thrombosis is decreasing due to the use of the recently approved centrifugal LVAD, such as the HeartMate 3, device-associated infections have not been eradicated [5]. Once infection occurs, patients must take antibiotics for a long period, and this makes it difficult to keep the PT-INR stable. The ISHLT guideline [3] recommends that the optimal range of PT-INR for the HeartMate II is 2.0-3.0. If pump thrombosis is suspected, it should be increased by 0.5, sequentially. Halder et al. [6] reported that the median values of PT-INR before major hemorrhage and thrombosis were 2.7 and 2.2, respectively, and the range of PT-INR that could suppress thrombotic events was narrow. In their report, the overall TTR in 51 patients was 52 %. Although an association between TTR and clinical outcomes in patients with an LVAD on chronic warfarin therapy is not well understood, Lea et al. investigated 30 thrombotic events in 25 patients [7]. In their study, the TTR (target INR = 2–3) was 11.4% lower 1 month before thrombus than the comparable month in the control period, and they concluded that the risk of thrombosis with lower TTR in the months leading up to thrombus was increased compared to a thrombus-free period. The optimal TTR in the management of patients with an LVAD has not been elucidated. Martinez et al. reported a weighted average TTR of 46.6 % in LVAD patients based on a meta-analysis (95% CI: 36.0–57.3%, I² = 94%) [8].

The most common reason for lowering the TTR was large individual differences in the pharmacokinetics of warfarin. It has been reported that the half-life of warfarin is highly variable from 55 to 133 h. In addition, warfarin interacts with various medicines, and its metabolism is affected by patient conditions such as infection, dietary intake, hyperthyroidism, hypothyroidism, and so on. In the present case, the TTR declined due to prominent PT-INR fluctuation caused by infection and the administration of antibiotics during the first suspected pump thrombotic event. However, we could not prevent the second thrombotic event, even though the TTR was maintained at 91% during the month prior to the event. Various factors other than the TTR, such as von Willebrand factor, were suspected of causing pump thrombosis. Some studies have shown an improvement in oral anticoagulation management in patients with an LVAD. Bishop et al. reported that when a pharmacist managed the PT-INR in a patient who self-tested using a point-of-care device, the TTR increased [9]. Lee et al. conducted a systematic review and meta-analysis of 3 studies to assess the benefits and harm of telehealth interventions for oral anticoagulation management [10]. In their report, the telehealth interventions mainly consisted of telephone contact by clinicians, pharmacists, and specialists; there were significant improvements in the telehealth group for major thromboembolic events compared to the usual care group, but there was no significant difference in major bleeding events. For our patient, we provided a POC device for PT-INR self-measurement and utilized the SNS instead of telephone contacts.

Telephone contact is performed individually between the patient and a member of the medical team and requires time for each person to respond to the call. In contrast, SNS monitoring allows patients and some members of the medical team to check medical information, including photo and condition of the patient, at their convenience regardless of time and place and this can be discussed within the medical team.

In this case, medical staff could complement each other's work by using their time efficiently and sharing information with the medical team. It is expected that the use of SNS will function as a work-sharing tool for the increasing number of LVAD patients.

In the home-management of patients with LVAD, collaboration with related hospitals and visiting nurses as well as in-hospital multidisciplinary teams is important. The SNS allowed the patient and the heart failure management team to share information that included the PT-INR, the patient’s condition, and warfarin dosage. Therefore, even an inexperienced staff could have specific knowledge, and SNS was beneficial as a tool for staff education.

We suspect that face-to-face hospital visits were complemented by the oral anticoagulation management system utilizing the SNS-assisted POC device, and this improved the TTR.

Conclusion

An SNS-assisted POC device for oral anticoagulation management in a patient with an LVAD may be useful to complement face-to-face hospital visits and improve the TTR. Because there is insufficient evidence, further investigations in this area are needed.

Declaration of Competing Interest

Dr. Machiko Asaka and Dr. Norihiko Kotooka have an endowed chair at Asahi Kasei Co., Ltd.