Yg. Weiss et al., Computer assisted physiologic monitoring and stability assessment in vascular surgical patients undergoing general anesthesia - Preliminary data, J CLIN M C, 16(2), 2000, pp. 107-113
Background. Physiologic monitors present an influx of numerical data that c
an be overwhelming to the clinician. We combined several parameters in an e
ffort to reduce the amount of information that must be continuously monitor
ed including oxyhemoglobin saturation by pulse oximetry, end-tidal CO2 conc
entration, arterial blood pressure, and heart rate into an integrated measu
re - the health stability magnitude (HSM). The HSM is computed for a pre-de
termined basal period, the reference HSM (RHSM), and recalculated continuou
sly for comparison with the baseline value. In this study we present the HS
M concept and examine changes in the HSM during abdominal aortic aneurysm s
urgery. Materials and methods. After IRB approval, nine patients were studi
ed. The anesthesiologist recorded all significant intra-operative events. W
ithin a defined time interval, data were recorded and used to calculate a c
ombined parameter, the HSM. The baseline or reference value of this index (
RHSM) was calculated after the induction of anesthesia. Individual HSM valu
es were repeatedly calculated for ten second periods after the RHSM value w
as established. A > 30% deviation of the HSM from the RHSM was considered s
ignificant. Deviations in the HSM were compared with events recorded by the
anesthesiologist on a paper record and with the record from an electronic
record-keeping system. The deviation observed between two consecutive HSMs,
called dHSM, was plotted against HSM to construct a contour diagram of dat
a from all patients to which individual cases could be compared. Results. T
he plot showed that dHSM vs. HSM values were tightly clustered. The inner c
ontour on the distribution plot contained 90% of values. Individual patient
's time course, projected on this diagram, revealed deviations form "normal
" physiology. Fifty-nine events led to > 30% deviations in the HSM; 27 were
anticipated events and 32 were unanticipated. Conclusion. The correlation
between HSM and dHSM depicts changes in multiple monitored parameters that
can be viewed using a single graphical representation. Projection of indivi
dual cases on the contour diagram may help the clinician to distinguish rel
ative intraoperative stability from important events. Data reduction in thi
s manner may guide clinical decision-making in response to unanticipated or
unrecognized events.