Page
22
of
24
subject provided the basis for building models to estimate
CRI, the proportion of remaining volume loss
the individual can tolerate before collapse (Equation 1)
. A robust CRI estimate requires analysis of the
continuous pulsatile waveform signal and comparison to a reference database. The estimated CRI is
determined by the most similar signal in the database.
Clinical Validation of CRI
To characterize the relationship of CRI to significant blood loss and to support regulatory clearance,
Flashback Technologies conducted a prospective validation study that enrolled 42 healthy volunteers (19
to 36) undergoing stepwise removal and replacement of approximately 20% of total blood volume while
in the supine position at rest (non-motion). The methodology and preliminary results from the first 20
subjects have been reported by (Convertino, Howard, et al. 2015) and the full dataset by (S. Moulton, et
al. 2017).
Results from the dataset of 42 volunteers demonstrate:
•
Trending Intravascular Volume Changes: 32 of the 42 subjects completed the full protocol. For
these subjects, the correlation between CRI and blood volume loss is greater than 0.9 (p<0.05).
o
Correlation between average CRI values (at constant intravascular volume) and
intravascular volume loss is greater than 0.94 (95% confidence 0.92 to 0.96).
o
Correlation between one-minute average CRI and intravascular volume loss is greater
than 0.9 (95% confidence 0.87 to 0.93).
•
In this population ‘normal’ CRI values (obtained from all 42 subjects) before intravascular volume
loss were significantly above 0.7 (see box and whisker plots in Figure 13, and Figure 14).
o
Before intravascular volume loss 95% of CRI values were above 0.75 and 98% of CRI values
were above 0.7.
o
Before intravascular volume loss the mean CRI value was 0.9 (95% confidence 0.89 to
0.91) with a standard deviation of 0.077 (95% confidence 0.053 to 0.078).
•
At 20 % intravascular blood loss volume, CRI values (obtained from the 32 subjects that completed
the full protocol) were significantly lower than those before blood loss (see box and whisker plots
in Figure 13 and Figure 14).
o
At 20% intravascular blood loss volume the mean CRI value was 0.6 (95% confidence 0.57
to 0.63) with a standard deviation of 0.17 (95% confidence 0.16 to 0.17).
•
When comparing one-minute average CRI values before blood loss to those at 20% intravascular
blood loss volume (for the 32 subjects that completed the full protocol), the average drop was
0.35 (95% confidence 0.29 to 0.4) with a standard deviation of 0.15 (95% confidence 0.11 to 0.2).
•
CRI values when subjects were experiencing symptoms associated with hemodynamic
decompensation were significantly lower than when subjects had no symptoms (see box and
whisker plots in Figure 13 and Figure 14).
o
Seventeen percent of the subjects (7/42) experienced
‘Hemodynamic
Decompensation’
during conduct of the study.
▪
As evidenced by a drop in systolic BP from 119 mmHg to <80 mmHg and a
significant drop in Mean Arterial Pressure (average decrease in MAP was 43
mmHg; average MAP prior to intravascular volume loss = 93 mmHg, range 91-95;
average MAP at the point of hemodynamic decompensation was 50 mmHg, range
44-58).
o
During symptoms the mean CRI value was 0.15 (95% confidence 0.12 to 0.17) with a
standard deviation of 0.08 (95% confidence 0.05 to 0.09).
o
During symptoms more than 97% of CRI values were below 0.3, while when no symptoms
were present 98% of CRI values were above 0.3.
