Intraoperative implementation of the hypotension probability indicator (HPI) algorithm
a pilot randomized controlled clinical trial

Intraoperative hypotension occurs often. In a recent not yet published study
(5, 6), we found in the AMC that up to 60% of patients endured hypotension
(defined as mean arterial pressure below 65 mmHg) during anaesthesia for an
average of 10% of surgery time.
Hypotension during surgery is associated with increased morbidity and
mortality. Studies have shown that even short durations of intraoperative
arterial blood pressure below 65 mmHg significantly increased the postoperative
risk of myocardial ischemia, neurological deficits, renal insufficiency, and
30-day mortality (1-4).
Not only the time spent in hypotension but also the severity of hypotension is
important for associations with postoperative outcome. The time weighted
average (TWA) combines the time and depth of hypotension (7) and is therefore a
good outcome parameter to study. The TWA calculates the area under the
threshold (in this study we use a threshold of 65 mmHg) per unit total time of
measurement.
Current treatment of intraoperative hypotensive episodes is not proactive and
rather occurs with some delay. There is need for a method to prevent
hypotension. Edwards Lifesciences has developed an algorithm that by analysing
continuously invasively measured arterial waveforms with the FlotracIQ is able
to predict hypotension with high accuracy minutes before blood pressure
actually decreases (5). Theoretically this would enable the treating
anaesthetist to proactive treat the predicted hypotension and thereby reducing
the TWA.

The algorithm developed is named the hypotension probability index (HPI), for
example a HPI of 85 translates approximately to a 85% chance of hypotension to
occur in the following minutes (unpublished data, see Chapter 6). The secondary
screen of the HPI shows variables such as dp/dt, dynamic elastance, systemic
vascular resistance as well as stroke volume, cardiac output and stroke volume
variation. These variables provide insight in the pathophysiology of the
predicted hypotension and in this way enable the treating anaesthetist to
provide the correct therapy to prevent this hypotension from occurring. The
device Flotrac is CE approved and has been in use in the operation room for
years. The FlotracIQ with HPI software is also CE approved but this will be the
first clinical study evaluating its possible beneficial effect. In this
two-phased trial we will test our hypothesis that the guided use of the HPI
will alter hypotension treatment and can reduce the time weighted average spent
in hypotension during surgery.

Study phase A: Prospective Data Collection
Objectives:
* What is the time weighted average spent in hypotension (defined as MAP<65
mmHg) during surgery?
* What is normal treatment behaviour of hypotension during surgery?
(Treatment choice, treatment dose, time to treatment)
* Hemodynamic parameters, measured beat tot beat (CO, SV, SVV, SVR,
elastance,
dP/dT)
* What is the time weighted average spent in hypertension (defined as MAP>100
mmHg) during surgery?

Study phase B: Prospective randomized controlled clinical trial.
Objectives:
* Does availability of HPI together with suggestions of the cause of
hypotension change treatment behaviour and correctness of the assessment among
anaesthetists and do they alter time towards treating hypotension?
* Can availability of Hypotension Probability Index (HPI) reduce the time
weighted average, the incidence and the percentage of time spent in hypotension
(defined as MAP<65 mmHg) during surgery?
* Does early treatment based on HPI increase the incidence, the TWA and the
percentage of time spent in hypertension (defined as MAP >100) during surgery?
* Feasibility of working with HPI (assessed by number of non-treated alarms)
* Explore the relation of HPI alarms to changes in hemodynamic parameters,
measured beat-to-beat (CO, SV, SVV, SVR, elastance, dP/dT)

* Can availability of Hypotension Probability Index (HPI) reduce the time
weighted average, the incidence and the percentage of time spent in hypotension
(defined as MAP<65 mmHg) after surgery, before being discharged to a normal
nursing ward?
* Does early treatment based on HPI increase the incidence, the TWA and the
percentage of time spent in hypertension (defined as MAP >100) after surgery?

This study is divided into two parts:
A. Prospective data collection (3-6 months)
B. Prospective randomized controlled clinical trial (6-8 months)

A. Prospective data collection
This will be a baseline data collection. We aim to include 40 patients and we
estimate this will take 3-6 months. We use this phase to be able to correct for
the possible Hawthorne effect.
We expect a Hawthorne effect during the RCT phase simply because more attention
will be drawn to prevent hypotension during surgery.
The baseline data collection enables us to prospectively calculate the time
weighted average spent in hypotension and provides data on normal treatment
behaviour of anaesthetists in the AMC.

B. RCT
We aim to include 60 patients and we estimate this will take 6-8 months. This
phase consists of a randomized controlled clinical trial in which patient and
statistician are blinded. Additionally the treating anesthetist in the
conventional arm is blinded to FlowtracIQ variables. The two study arms are
described below:

1) Treatment arm: Flotrac with HPI with guidance. HPI will be calculated via
FlotracIQ connected to the radial arterial line. The treating anaesthetist is
trained to understand Flotrac parameters and the meaning of HPI. The treating
anaesthetist is provided with guidance concerning timing of treatment (HPI
>85%) and the causes of hypotension. See flowchart in Appendix I. HPI values
above 85 translate to a 85% of hypotension to occur in the following minutes,.
HPI values between 50-85% translates approximately to a 50-85% change of
hypotension to occur in the following minutes (see chapter 6). When HPI falls
in the 50-85% range the study investigator starts diagnosing the cause of
pending hypotension via the secondary screen. If HPI reaches 85%, treatment is
suggested to be started and the investigator informs the anesthesiologist of
the most likely cause of hypotension (see Appendix I) The anesthesiologists
treats the patient based on these suggestions but may choose to deviate from
the study protocol if deemed necessary in both timing as well as treatment.
Specific scenarios can be formulated with use of the secondary screen. The
secondary screen of the FlotracIQ provides the investigator and treating
anaesthetist with variables such as dp/dt, dynamic elastance, systemic vascular
resistance as well as stroke volume, cardiac output and stroke volume
variation. In Appendix II the pathophysiological meaning of these variables is
explained. For more information about the FlotracIQ device and or the HPI
software please read chapter 5 and chapter 6.

2) Conventional arm: Institutional Standard of Care (excluding perioperative
goal directed therapy) with intention to keep MAP> 65 mmHg. The FlotracIQ will
be connected but fully covered. We connect the device to be able to calculate
the TWA similarly to the intervention group.

In the RCT phase of this study we look at the effect of introduction of a
protocol (based on FloTracIQ) with blood pressure targets and predictive
analytics to reduce total TWA spent in hypotension during surgery.

This is a pilot study. We can use this data to set up a larger multicentre
study in the future.

We look at the effect of introduction of a protocol (based on FloTracIQ) with
blood pressure targets and predictive analytics (HPI) to reduce the time
weighted average (TWA) spent in hypotension during surgery.
Flotrac device is a CE certified product, in use in European theatres and
Intensive Care Units (8, 9). It is used to monitor hemodynamic parameters
derived from waveforms, such as cardiac output, stroke volume, stroke volume
variation and dynamic elastance. We do not use it outside the scope of the
intended use.
Using the Flotrac device means attaching the device to the arterial line, it
does not induce any harm or risks for the participants. The Flotrac is
completely safe to use.
In our centre the Flotrac device without HPI software is already in use when
cardiac output needs to be monitored.

The FlotracIQ device with HPI (hypotension probability index) software is our
investigational product (10). The HPI is able to predict hypotension before it
occurs, i.e. a HPI of 85 translates approximately to a 85% chance of
hypotension to occur in the following minutes (see unpublished data in chapter
6). The variables in the secondary screen (appendix II) provide information on
the pathophysiology of the predicted hypotension. Broadly, hypotension can be
explained by a preload, afterload or contractility *problem*. Or otherwise
stated, the treatment of hypotension can consist of: fluids, vasopressor and/or
inotropes. The variables given by the FlotracIQ HPI are dp/dt, dynamic
elastance (Eadyn), stroke volume (SV), stroke volume variation (SVV), cardiac
output (CO) and systemic vascular resistance (SVR). By knowing the values of
these variables the anaesthetist can *diagnose* the cause of the hypotension
and thereby administer the correct treatment.

We use the FlotracIQ device HPI additionally to conventional intraoperative
monitoring. Furthermore the treatment suggestions we will provide is not
different from the standard treatment options used to treat intraoperative
hypotension.

The difference, however, can be found in treating the hypotension proactive
rather than
reactive. See appendix I for the treatment and guidance flow diagram.

We aim to test if guided use of the HPI will alter hypotension treatment
and can reduce the
time weighted average spent in hypotension during surgery.
the HPI is CE certified (11) and has been validated on prospectively
collected patient data by means of offline analysis (5), however this will
be the first clinical trial in humans.

The study participants allocated to diagnosis according to de HPI algorithm
will receive diligently titrated vasopressors, inotropes or fluids minutes
before hypotension is predicted to occur. Theoretically this may lead to
(transient) hypertension (MAP>100). Besides hypertension no adverse events are
expected during this study. The anaesthetist in charge is free to deviate from
the HPI algorithm protocol suggestions when he/she feels necessary. Monitoring
with Flotrac HPI is used additionally to standard monitoring. We use the CE
approved Flotrac HPI device within the intended indication of use. Therefore
the risks for study participants are expected to be low.