Hemodynamic Instability Index: The Impact to Care With a New Predictive Indicator

Philips has developed a new algorithm, called Hemodynamic Instability Rules Based Advisor
(HIRBA 2.0) system, for the prediction of hemodynamic instability in critically ill patients,
who were not previously known to be at risk. This algorithm provides with the numerical index
- Hemodynamic Instability Index (HII) and its trend which are calculated from existing real
time data derived from patient's current hospital stay such as heart rate and blood pressure,
labs, and Admission - Discharge- Transfer (ADT) data to help healthcare providers obtain
information about patient's status and make clinical decisions.

The aim of the study is to determine if the patient clinical information, displayed in the
form of HII, can prompt possible patient hemodynamic status change and trigger earlier
clinical care team response. Study will also identify if the use of the Philips new algorithm
called Hemodynamic Instability Rules Based Advisor (HIRBA 2.0) can improve outcomes for
patients admitted to the intensive care units (ICUs).

During this study, upon patient admission to the ICU clinical data collection will begin.
Data that will be pulled during the study includes: vital signs, medications, reason for
admission to the ICU, results from blood work, treatments that patient may receive.

Research subjects will not be asked to do anything as far as specific activities expected to
be accomplished during study participation. Subjects will be involved in the study only
during their stay in ICU. HII information will be visible to clinicians, taking care of this
patient only during his/ her stay in the study unit. Patients will be randomized into HII arm
(treatment group) and into no-HII - standard treatment arm (control group).

Study team will collect patient information and have the ability to view the HII for subjects
randomized into HII group arm. For subjects randomized into the control arm study team will
collect patient's clinical information but the clinical care team will not have the ability
to view the HII.

Once patient leaves the ICU the study team will collect information on the remainder of
patient's hospital course including date of hospital discharge. All collected data will be
then fully de-identified and released to Philips for analysis.

This clinical investigation is designed to answer several key questions including:

1. Are the performance metrics the same when the algorithm operates in the live host as it
does on retrospective data?

2. Are there opportunities for improvements in the host implementation; including graphical
rendering of the HII information and creation of an alert?

3. Are the predicted workflows which result from the use of the HII clinically acceptable?

4. Is there a difference in the clinical measures for patients who have HII displayed when
compared to those who do not?

5. What is the optimal care process which should be followed based on the presence of the
indicator?