Drug Flow Through Clinical Infusion Systems: How Modeling of the Common-volume Helps Explain Clinical Events

Abstract

<jats:title>Abstract</jats:title><jats:p>This review aims to describe analytic models of drug infusion that demonstrate the impact of the infusion system common-volume on drug delivery. The common-volume of a drug infusion system is defined as the volume residing between the point where drug and inert carrier streams meet and the patient’s blood. We describe 3 sets of models. The first is quantitative modeling which includes algebraic mathematical constructs and forward-difference computational simulation. The second set of models is with in vitro benchtop simulation of clinical infusion system architecture. This modeling employs devices including pumps, manifolds, tubing and catheters used in patient care. The final set of models confirms in vitro findings with pharmacodynamic endpoints in living large mammals. Such modeling reveals subtle but important issues inherent in drug infusion therapy that can potentially lead to patient instability and morbidity. The common-volume is an often overlooked reservoir of drugs, especially when infusions flows are slowed or stopped. Even with medications and carriers flowing, some mass of drug always resides within this common-volume. This reservoir of drug can be inadvertently delivered into patients. When infusions are initiated, or when dose rate or carrier flow is altered, there can be a significant lag between intended and actual drug delivery. In the case of vasoactive and inotropic drug infusions, these unappreciated time delays between intended and actual drug delivery can lead to iatrogenic hemodynamic instability. When a drug infusion is discontinued, drug delivery continues until the common-volume is fully cleared of residual drug by the carrier. The findings from all 3 sets of models described in this review indicate that minimizing the common-volume of drug infusion systems may enhance patient safety. The presented models may also be configured into teaching tools and possibly point to technological solutions that might mitigate sources of iatrogenic patient lability.</jats:p>

Inferência dos Perfis de Infusão em Sistemas Intravenosos: Uma Abordagem Empregando Técnicas de Aprendizagem de Máquina

“…Por outro lado, a modelagem para infusões gravitacionais atingiu um resultado um pouco pior, conforme apresentado na Figura 2 parte (a). Isso ocorre devido ao regime de líquido escoado, que caracteriza o comportamento do perfil de liberação do fármaco por infusões gravitacionais [Lovich and Peterfreund 2017].…”

“…Onde na Figura 4, o eixo Y indica os valores da velocidade das taxas de infusão em mL/h, o eixo X mostra a duração das infusões em horas e o escopo da Figura, as caixas de texto informam os valores dos perfis infusões e os volumes nominais de cada infusão intravenosa (rótulo Real) organizados por sequência numérica. Considerando a Figura 4, podemos ver que as inferências das taxas de infusão da Bomba de Infusão tiveram um melhor desempenho com as inferências feitas por infusões gravitacionais; os resultados são esperados, considerando que as bombas de infusão têm como finalidade um melhor desempenho na administração de medicamentos intravenosos [Lovich and Peterfreund 2017].…”

Avaliação do Uso de Aprendizagem de Máquina na Inferência de Perfis de Infusões Intravenosas

“…Isso ocorre devido ao regime de fluidos líquidos, que caracteriza o comportamento do perfil de administração de drogas por infusões gravitacionais. As infusões gravitacionais têm um fluxo semelhante aos modelos de funções polinomiais [13], dificultando uma modelagem mais precisa.…”