Biomedical Engineering Reference
In-Depth Information
the mask and the patient's face. The leakage likely contributes to instability
since the inspiratory phase of pressure support terminates when ow falls to
a predetermined fraction of peak inspiratory ow
5
. Instabilities predicted
by the above mentioned studies were found to be entirely independent of
patient's eort or volition.
Our study highlights the complex dynamic behaviors that occur even
in the absence of a mask leak. Although the mathematical model we em-
ployed is a simple one-compartment linear system, the model simulations,
in which we allowed the total breathing time t
tot
to vary from one breath
to another (in the case of skipped breaths), agree with experimental ob-
servations, illuminating factors that inuence patient ventilator synchrony
and tolerance. Regions of variable (unstable) V
T
delivery are related to
respiratory frequency, resistance and compliance, pressure sensitivity, and
pressure triggering levels.
As can be seen in Figure 4, for xed frequency, pressure sensitivity and
resistance, elevated compliance C increases the unstable ow cut-o values,
enlarging the region of instability. Figure 4 also indicates that lower pressure
sensitivity exacerbates the V
T
instability. On the other hand, for a xed
pressure sensitivity, P
sen
, and resistance, increasing respiratory frequency
results in larger region of V
T
instability for each pressure triggering level.
Figure 6 indicates that higher ow resistance leads to greater V
T
in-
stability. It pushes the curve for each pressure triggering level further to
the left and higher, and therefore enlarging the region of instability. At a
low resistance level of 15 cmH
2
O/L/s, with P
sen
= 5 cmH
2
O, and f = 14
breaths/min as in Figure 6(a), stability is predicted for all pressure trigger-
ing levels if C < 0:8.
Previous results indicated that the stability of support during PSV de-
pends critically on complex dynamic interactions. As we have shown here,
many of the determinants of instability are amenable to both mathemati-
cal analysis and clinical manipulation. This fact suggests clear potential for
mathematical analysis to guide and improve patient care.
Acknowledgments
The authors, Rattanamongkonkul and Lenbury, would like to thank the
Thailand Research Fund and the National Research Council of Thailand
for the nancial support. All of the authors are indebted to Dr. Li Chen for
data that she collected from the mechanical lung experiments at Regions
Hospital in St. Paul, MN.
Search WWH ::
Custom Search