Prediction of Post-induction Hypotension with Arterial Pulse wave applied Machine Learning, a non-randomized prospective data collection observational study.

Hypotension during surgery is associated with increased morbidity and
mortality. The majority of patients will have post-induction hypotension (PIH),
a mean arterial blood pressure below 65 mmHg for at least one minute and
occurring during the first 20 min after anesthesia induction. PIH is highly
prevalent and probably occurs more often than intraoperative hypotension (IOH).
Early IOH is described as a mean arterial blood pressure below 65 mmHg for at
least one minute occurring in the first 30 minutes after the start of surgery.
Also, PIH is very common in Intensive Care Unit patients and is likely to have
an equally negative effect on outcome as any other type of IOH. Even short
periods of hypotension are known to contribute to the occurrence of
postoperative renal failure, myocardial injury, stroke and length of hospital
stay.

The early identification and treatment of hypotension is clinically relevant.
Current therapies are reactive and are started after hypotension occurs.
Post-induction hypotension (PIH) is likely to occur in the majority of cases in
the face of boluses of anesthetic agents causing severe vasodilation and even
temporary cardiac depression as a surgical stimulus is missing. Since any type
of hypotension is likely to have negative effects, prevention is warranted. A
machine-learning algorithm based on the arterial pressure signal for the
prediction of post-induction hypotension, in analogy of the recently
FDA-approved intra-operative Hypotension Prediction Index (HPI), would
eventually allow preemptive treatment and prevention of post-induction
hypotension altogether.

Primary Objective:
The primary aim of this study is data collection of continuous noninvasive
arterial pressure waveform signals with the ClearSight (CS) finger cuff,
continuous invasive arterial pressure waveform signals when an arterial cannula
is already available due to standard of care, continuous noninvasive cerebral
oximetry signals and clinical data from patients* electronic medical record
(EMR) in surgical and ICU patients. These data will be used to predict the
likelihood of derangement of physiologic parameters in awake patients before
induction of anesthesia and to predict the occurrence of post-induction
hypotension and IOH using machine learning.

The collected digital pressure waveform will be used to assess the feasibility,
the learning and building of an initial machine learning model using the
CS/EV1000/HemoSphere continuous (non)invasive arterial pressure signal and
internally validate it.

Secondary Objectives:
To correlate pre-induction waveform and patient history characteristics,
alterations in cerebral oximetry, used medication during induction and
maintenance of anesthesia, and the occurrence of events (acute kidney injury,
myocardial injury, stroke, in-hospital length of stay and 30-day mortality)
during the first 30 days after induction of anesthesia, to PIH and IOH.

This is a non-randomized prospective observational data collection study that
will take place only in the Amsterdam UMC, location AMC in the Netherlands.
We estimate that inclusion will take 18 to 45 months from initiation of study
(aim December 2018; actual start of the trial: 7th of January 2019).

Electronic data collection of continuous (non)invasive arterial pressure
waveform signals and cerebral oximetry signals takes places with the
CS/EV1000/HemoSphere system in awake patients before induction of anesthesia.
In addition to noninvasive and invasive arterial pressure waveform data and
noninvasive cerebral oximetry data, the data collection will require
deidentified patient medical records/anesthesia charts.
We evaluate hemodynamic parameters at least 30 minutes before to a minimum of
20 minutes after induction and its correlation with post-induction hypotension
for ICU patients requiring intubation or elective tracheostomy.
We evaluate hemodynamic parameters at least 30 minutes before induction to a
minimum of 30 minutes after start of surgery and its correlation with PIH or
IOH for elective surgical patients.

We aim to include at least 600 and maximally 1100 patients for elective surgery
or for intubation or elective tracheostomy in the ICU.

Phase 1A:
A pilot phase comprising 100 elective surgical patients. Standard of care is
performed and timing and dosing is left to the judgement of the attending
anesthesiologist.

Phase 1B:
Continuation of phase 1A, with the additional inclusion of ICU patients
requiring intubation or elective tracheostomy. Invasive blood pressure data
will be collected when an arterial cannula is already available due to standard
of care and noninvasive measurements of blood pressure and cerebral oximetry in
both surgical and ICU patient groups. Standard of care is performed and timing
and dosing of anesthetics is left to the judgement of the attending
anesthesiologist or intensivist.

After phase 1 we will perform an interim analysis to determine the number of
patients needed in phase 2.

Phase 2:
A non-randomized prospective observational data collection study in at least
600 and maximally 1100 (dependent on the interim analysis) elective surgical
patients or ICU patients requiring intubation or elective tracheostomy. An
amendment will be submitted if there are changes to the protocol.

Phase 3:
Data collection in 100 additional patients to perform an external validation of
the predictive algorithm.

There are no additional risks or benefits associated with participation. There
are no investigational devices used in this study. There are no additional
risks associated with the use of the CS/EV1000/HemoSphere monitor other than
described in the Instructions for Use. There are also no risks associated with
the study procedures. The anesthetic regimes are based on what is currently
used in daily practice and reported in the literature. We give standard
anesthesia on basis of daily practice and established pharmacodynamic models
that have been shown to be both save and effective. *