ABSTRACT
Patients who undergo heart valve replacements with mechanical valves need to take Vitamin K Antagonists (VKA) drugs (Warfarin, Nicoumalone) which has got a very narrow therapeutic range and needs very close monitoring using PT-INR. Accessibility to physicians to titrate drugs doses is a major problem in low-middle income countries (LMIC) like India. Our work was aimed at predicting the maintenance dosage of these drugs, using the de-identified medical data collected from patients attending an INR Clinic in South India. We used artificial intelligence (AI) - machine learning to develop the algorithm. A Support Vector Machine (SVM) regression model was built to predict the maintenance dosage of warfarin, who have stable INR values between 2.0 and 4.0. We developed a simple user friendly android mobile application for patients to use the algorithm to predict the doses. The algorithm generated drug doses in 1100 patients were compared to cardiologist prescribed doses and found to have an excellent correlation.
ABSTRACT
Resumen La razón internacional normalizada (RIN) se utiliza para controlar a los pacientes anticoagulados con antagonistas de la vitamina K. El objetivo de este estudio fue evaluar el desempeño del nuevo dispositivo portátil microINR, que utiliza tromboplastina recombinante, en pacientes anticoagulados con acenocumarol. Grupo 1: los pacientes proporcionaron muestras de sangre venosa y capilar para realizar pruebas paralelas que permitieron comparar microINR con cinco combinaciones diferentes de reactivo/sistema de detección: tromboplastina de cerebro de conejo con detección nefelométrica, foto-óptica y por viscosidad y tromboplastina humana recombinante con detección nefelométrica y fotoóptica. Todas las tromboplastinas tenían un índice de sensibilidad internacional (ISI) específico de equipo informado por el fabricante menor de 1,10. Grupo 2: los resultados de microINR se compararon con CoaguChek como dispositivo preestablecido. La precisión se evaluó utilizando materiales de control líquidos. El coeficiente de variación del material de control en microINR fue de 6,2%. El análisis de regresión de Passing-Bablok y Bland-Altman mostró un coeficiente de correlación superior a 0,92 y un pequeño sesgo cercano a cero para todas las comparaciones de microINR con cada método tradicional realizado con sangre venosa. Ambos dispositivos portátiles tuvieron un muy buen coeficiente de correlación (r=0,98) y un pequeño sesgo de 0,02. Los resultados muestran un acuerdo clínico del 100% con un grado de concordancia Kappa mayor de 0,63 para todos los métodos tradicionales y de 0,82 para la comparación con Coagu- Chek. De acuerdo a los resultados obtenidos, el microINR es adecuado para el control de pacientes anticoagulados con acenocumarol.
Abstract The international normalised ratio (INR) is used to monitor vitamin K antagonist anticoagulated patients. The aim of this study was to evaluate the performance of the microINR portable device in acenocoumarol anticoagulated patients. Group 1: patients provided capillary and venous blood samples for parallel testing comparing microINR with five different pairs of reagent/clot detection systems: brain rabbit thromboplastin with nephelometric, photooptic and viscocity clot detection and recombinant human thromboplastin with nephelometric and photooptical detection. All thromboplastins have an instrument-specific International Sensitivity Index (ISI) lower than 1.10 reported by the manufacturer. Group 2: microINR results were compared with CoaguChek as an established device. Precision was assessed using liquid control materials. The control material coefficient of variation obtained in microINR device was 6.2%. The Passing-Bablok and Bland-Altman regression analysis showed a correlation coefficient greater than 0.92 and a small bias close to zero for all comparisons of microINR with each traditional coagulation method performed on venous blood samples. Both portable devices had a good correlation (r=0.98) and a very low bias of 0.02. The results show clinical agreement of 100% with a Kappa greater than 0.63 for different traditional INR and greater than 0.83 for CoaguChek. The performance microINR is suitable for the anticoagulation control of patients taking acenocumarol.
Resumo A razão normalizada internacional (INR) é utilizada para monitorar pacientes anticoagulados com antagonistas da vitamina K. O objetivo deste estudo foi avaliar o desempenho do dispositivo portátil microINR, um novo dispositivo que utiliza tromboplastina recombinante, em pacientes anticoagulados com acenocumarol. Grupo 1: os pacientes forneceram amostras de sangue venoso e capilar para testes paralelos que permitiram a comparação do microINR com cinco combinações diferentes de reagente/ sistema de detecção: tromboplastina de cérebro de coelho com detecção nefelométrica, foto-óptica e de viscosidade. Tromboplastina humana recombinante com detecção nefelométrica e foto-óptica. Todas as tromboplastinas tinham o Índice de Sensibilida de Internacional (ISI) específico do kit relatado pelo fabricante inferior a 1,10. Grupo 2: os resultados do microINR foram comparados com o CoaguChek como dispositivo padrão. A precisão foi avaliada usando materiais de controle líquidos. O coeficiente de variação do material de controle em microINR foi de 6,2%. A análise de regressão de Passing-Bablok e Bland-Altman mostrou um coeficiente de correlação maior que 0,92 e um pequeno viés próximo a zero para todas as variáveis. sangue. Ambos os wearables tiveram um coeficiente de correlação muito bom (r=0,98) e um pequeno viés de 0,02. Os resultados mostram 100% de concordância clínica com grau de concordância Kappa maior que 0,63 para todos os métodos tradicionais e 0,82 para a comparação com CoaguChek. De acordo com os resultados obtidos, o microINR é adequado para o controle de pacientes anticoagulados com acenocumarol.
ABSTRACT
@#Introduction: The fresh frozen plasma (FFP) is frequently prescribed either for therapeutic or prophylactic transfusion. The international normalised ratio (INR) value of 1.50 and above is frequently reported to be a transfusion trigger for FFP prior to interventional procedure. This study aimed to evaluate the efficacy of prophylactic FFP transfusion in normalising the INR and to determine the post-transfusion outcomes. Methods: A prospective cross-sectional study involved 81 patients who received prophylactic FFP transfusion over a period of three months. All demographic, clinical data and outcomes of FFP transfusion were captured and filled in the research proforma. Results: The proportion of patients achieved posttransfusion INR below 1.51 was 30.30% (n=27). The majority of patients underwent the interventional procedures with posttransfusion INR > 1.50 (n=52) without experiencing any bleeding episodes. Overall, FFP transfusion resulted in significant median INR difference from 1.89 (IQR, 0.53) to 1.60 (IQR, 0.25); p< 0.001. The greater median INR difference was observed in group with pretransfusion INR > 2.00 and who received FFP doses between 10.00 to 20.00 ml kg-1 (p < 0.001). The INR difference showed the significant, positive correlation with pretransfusion INR values (rs= 0.83, p < 0.001) and FFP doses (rs= 0.72, p< 0.001). Conclusions: The interventional procedures were safely carried out despite abnormal posttransfusion INR. The prophylactic FFP transfusions could be avoided in patients with mild coagulopathy (INR 1.50 - 2.00) prior interventional procedures.