P1630: WEARABLES IN TRANSFUSION MEDICINE: CHALLENGES AND FUTURE PERSPECTIVE

Abstract

Background: The digital era has granted us new techniques like photoplethysmogram or ECG-based wearable devices, that can remotely monitor physiological parameters of patients, including heart rate and activity measures continuously. These developments yield many opportunities but also great challenges. For instance, in the Myelodysplastic syndrome (MDS), most patients will become RBC transfusion dependent during the course of their disease. It has been well established that Hb-level alone is an insufficient marker to guide in RBC transfusions. Nonetheless, we base transfusion regiments on Hb-thresholds combined with a subjective interpretations of the patient’s quality of life. This leads to great variability in transfusion strategies causing patients to often be over- or undertransfused. We therefore need objective parameters to evaluate and guide our transfusion strategies. Aims: We intend to share our practical knowledge of the encountered challenges. Methods: Following pilots with the Withings Steel HR and VitalPatch (Table 1.), we have started the REMOTE 2 trial (NTR: NL9289) and are collaborating internationally to initiate larger trials. Here we have gathered the experiences we deem important for researchers new to the field. Results: When choosing the device, first establish the desired outcomes. Is a true continuous heart rate collection necessary, or will a semi-continuous measurement suffice? Within an elderly MDS population a more manageable device with a long battery-life can help reduce data missingness. On the other hand, with only daily averages of parameters, a deeper understanding of interaction of the collected parameters is challenging. Also, product placement will impact the outcome, as wrist-worn devices are more susceptible to noise than chest-worn devices. Importantly, inter-device variability has been reported to be high. Therefore, it is strongly preferable to use one device only in a study. Many wearable biosensors are not medically validated (yet). We believe, however, this should not discourage the investigator to use such a device. Unvalidated continuous heart rate measurements have been shown to produce very accurate data. And importantly, there are currently no validated – or unvalidated – parameters to evaluate these patients. Physiological parameters are highly impacted by numerous factors like weather and daily schedules. Large sample sizes should reduce this natural variability, but trials involving MDS-patients have thus far been unable to collect more than 38 patients in one study. Combining groups, like thalassemia and myelofibrosis, will increase the sample size. However, subgroup analysis will still lack power. Therefore, international collaborations are necessary to acquire sufficient sample sizes to answer questions like: What parameters are of value to evaluate transfusion strategies? How do these parameters respond to different 1) amounts of RBCs; 2) thresholds; and 3) timing of transfusion? After answering these first questions, we can look at how these parameters change with longer-standing changes, and how deleterious the fluctuations in Hbmass are to the patient. Image:Summary/Conclusion: Depending on the research context, intervention and the desired outcomes, one should carefully choose the required device. While various devices may mork, the trade-off between ease of use and data granularity can make a substantial difference when analysing the data. For sufficient sample sizes, international collaborations are required to answer the many questions in order to improve the care of chronic transfusion dependent patients.