Identification of critical patient-ventilator complex interactions

Identification of critical patient-ventilator complex interactions 1080 810 Guillem Cebrian

Recently, research has begun that aims to predict the behavior of critically ill patients during mechanical ventilation and their response to the machine. The first step has been to develop a method based on a series of entropy sample calculations in order to identify complex patient-ventilator interactions during mechanical ventilation.

This initial phase of the research has already given rise to a first posting al Scientific Reports, a very important Q1 journal in the multidisciplinary science category, which in addition publishes in open access.

Then we talk to Leonardo Sarlabos (on the left of the image, along with Rudy's Pomegranates, co - author of the publication and member of Better Care). Sarlabos is dated scientist ofI3PT and first author of the publication, and will tell us broadly what the complex interaction between patient and ventilator is, how they have identified it, what calculations they have been based on, and what the future lines are, among others.

What is the purpose of your research?

(Leonardo Sarlabos). This study has 3 well-defined objectives. Firstly the definition of complex interactions (CP-VI), secondly the validation of these by experts in mechanical ventilation, and finally the development of an automatic algorithm capable of detecting the same interactions.

Invasive mechanical ventilation is a means of life support administered to patients who cannot breathe on their own. During mechanical ventilation there are mismatches between the ventilator and the patient's respiratory pattern leading to asynchronies. Asynchronies are difficult to define based solely on visual analysis performed by non-expert staff. Very recent studies developed by different high-level international groups, including our own group, highlight the impact that asynchronies have on clinical outcomes.

In our research we have defined the term complex patient-ventilator interaction as the change in respiratory rate by more than 50% and / or having more than 30% of cycles with asynchronies of any kind in a time period of 3 minutes.

What type of sample did you study?

(LS) We analyzed traces of pressure and respiratory flow of 27 critically ill patients who self-extubated, where patients experience complex interactions. The registrations were made at the Parc Taulí Hospital in Sabadell and at the Hospital Althaea of Manresa through the connectivity platform Better Care which allows interoperate signals from different ventilators and monitors, and subsequently apply algorithms to diagnose patient-ventilator asynchronies.

This research was also developed in the framework of the SCHEDULE CHALLENGES (RTC-2017-6193-1) and in collaboration with the CIBER (Research 33). This represents a clear example where public-private collaboration is of vital importance for the implementation of findings from scientific research. In this case, our research into complex interactions has also led to a European patent (pending resolution) which, once acquired, will be licensed to thespinoff Better Care for exploitation.

Where did you extract all the calculations and values ​​presented in the article?

(LS) These calculations are based on entropy. It is a non-invasive method that measures theother times and the predictability of stochastic processes. That is, a very irregular process will have high entropy while a very regular process will have low entropy. A clear example where entropy has been successfully used as a monitoring tool has been the measurement of the degree of depth of anesthesia.

Our hypothesis is that in periods with complex interactions the entropy values ​​will be higher than in periods with regular patient-ventilator interactions ”. says Sarlabous

In this sense, we analyze in 15-minute windows the changes that exist in the value of entropy with respect to the value of basal entropy that the same patient has. In this way we ensure that the algorithm is customized for each patient. Our algorithm is able to detect a complex interaction when a 25% increase in entropy calculated from the flow signal or a 30% increase in entropy calculated from the pressure signal is experienced.

What are the future lines of research and how far do you want to go?

(LS) We firmly believe that the events of complex interactions depending on when they occur during mechanical ventilation could give us different types of information.

On the one hand we have the hypothesis that the events of complex interactions could be closely related to the patient's ability to successfully pass the spontaneous breathing test (SBT). That is, the fact that a patient has such events could alert us that he is ready to perform the SBT if the test's own criteria are met.

On the other hand, if these events occur in stages where the patient does not meet the criteria to perform the SBT it could indicate to us that the patient has a certain probability of self-extubation. It would be a very useful tool to alert medical staff of possible self-extubation.

Hence, it may be necessary to explore the clinical relevance and usefulness of identifying complex patient-ventilator interactions in different clinical settings. In any case, this represents one more step towards the precision and predictive personalized medicine.

Guillem Cebrian

Graduate in Information and Documentation (UB) and Master in Management and Direction of Libraries and Information Services (UB). At I3PT I am in charge of the Knowledge Management Unit and I am in charge of collecting and disseminating its scientific production. I am passionate about new technologies, data management and open science.

All stories by: Guillem Cebrian

Leave a Reply

Your email address Will not be published.

Privacy preferences

When you visit our website, your browser may store information about specific services, usually in the form of cookies. Here you can change your privacy preferences. Please note that blocking certain types of cookies may affect your experience on our website and the services we offer.

Enable / disable Google Analytics tracking code
Enable / disable Google Fonts
Turn Google Maps on / off
Enable / disable embedded videos
This website uses cookies, mainly from third party services. You can edit your privacy preferences and / or accept the use of cookies.