Abstract
Dengue hemorrhagic fever (DHF) is one of the critical public health problems in Thailand. During the disease progression, DHF patients will enter the critical phase in which hypovolemic shock potentially occurs and might lead to death. The shock in dengue is caused by substantial plasma leakage from blood circulation. Plasma leakage potentially leads to hemoconcentration, insufficiency of blood plasma, and ultimately hypovolemic shock. Thus, current guidelines on management of dengue fever/ dengue hemorrhagic fever recommend monitoring patient’s hematocrit levels by performing invasive blood collections every 2-4 hours during the critical phase. hematocrit levels are then used to adjust intravenous fluid replacement in DHF patients to prevent hypovolemic shock. This project aims to address two main research questions, including 1. the feasibility of using a noninvasive hemoglobin monitoring device (Masimo), already approved for medical used, for monitoring hemoconcentration instead of collecting patient’s blood through conventional methods and 2. development of a non-invasive prototype measuring hemoconcentration with a comparable or higher level of efficiency compared to the commercialized noninvasive device (Masimo). The prototype has been designed to be affordable for public hospitals in Thailand and to potentially replace expensive commercialized devices (Masimo). For the first research question, hemoglobin levels were measured in 154 patients with dengue fever (from 336 patients in this study) using the commercialized devices. Data were analyzed for the relationship between hemoglobin from noninvasive device and hematocrit from blood collection. The analysis revealed that hematocrit levels obtained from venipuncture or fingertip prick and hemoglobin levels from the Masimo device tended to change in the same direction. However, the incidence rate of dengue infection was very low over the course of the data collections. In addition, the majority of dengue patients in this study had dengue fever (DF) that does not have a critical phase seen in DHF. Thus, the overall variation of hematocrit levels observed in patients in this study was small due to the lack of data from the critical stage. Hence, our proposed model for the relationship between hemoglobin and hematocrit was not sufficiently tested with patients during the critical stage. In order to apply our model to routine hematocrit monitoring, an interface extension from Masimo was built to receive signals from the Masimo device, analyze the signals using our proposed mode, and convert them to hematocrit. In addition, we also proposed a new guideline for monitoring patients with dengue infection before, and after, reaching the critical stage using the developed interface and Masimo device. For the second research question, we adopted AFE4490 Integrated Analog circuit and Nellcor ds100a probe for non-invasive hemoconcentration measurement prototype. This prototype used Arduino DUE as a processor. We have designed the real-time algorithm and auto alert system calibration by weight average technique using signals from red and infrared sensors. The data were collected by a computer connecting to the Arduino processor. The prototype was calibrated with FLUKE Index 2 pulse Oximeter simulator. We found that the signals read were accurate. A comparison of data from the prototype and the Masimo device suggests that the correlation between REDDC signals from the prototype and SpHb values from the Masimo is very strong although accuracy of measurement from the prototype is not as high as the Masimo. Initially, the prototype was also used to monitor hemoconcentration of 73 patients with thalassemia (with about 70% of accuracy) after they received blood transfusion. The prototype will be further improved to increase its accuracy in measurement of hemoconcentration through laboratory tests and model adjustment.
