PPG in the OR and ICU

Photoplethysmography (PPG), often performed with a pulse oximeter, is an
optical method providing a measurement of the changes in blood volume in the
measurement site (e.g., the finger tip) as a result of cardiac output-induced
blood pressure pulses [WebsterJG1997-Book-a]. With each cardiac cycle the heart
pumps blood around in the body and the pressure wave induced by heart
contraction results in a transient increase in vascular diameter, which is
shown in the PPG signal as a transient decrease in the light transmission
through e.g. the finger. Since the late 1980's, pulse oximeters are widely used
to measure SpO2 and pulse rate.

Various reviews in the last decade advocate to investigate novel advanced
PPG-derived hemodynamic parameters beyond pulse rate and SpO2 [ShelleyKH2007a,
AllenJ2008a, SahniR2012a]. Some recent PPG-derived parameters under
investigation are predictors of fluid responsiveness [CannessonM2007a,
MonnetX2012a], respiration frequency [AddisonPS2012a], detection of congenital
heart disease in newborn infants [EwerAK2012a], blood pressure [SolaJ2013a],
cardiac output [IshiharaH2004a,IshiharaH2012a], and detection of return of
spontaneous circulations in patients with cardiac arrest [WijshoffRWCGR2013a].
However, little is known about the performance of these advanced PPG-derived
parameters in clinical practice, and these new parameters may put more
stringent requirements on the PPG signal than before [MonnetX2013a]. There are
two aspects of the PPG signal that require deeper understanding.

The first aspect is the reliability and quality of the PPG signal. The
performance of pulse oximeters can be severely compromised under low-perfusion
conditions, motion, and arrhythmias [WebbRK1991a, ReichDL1996a,
BransonRD2004a, BatchelderBP2007a, HummlerHD2006a]. The performance of pulse
oximeters is traditionally measured in terms of the precision, bias, and
accuracy of the SpO2 reading [ISO9919:2005, VanDeLouwA2001a, SeguinP2000a], but
drop-out rates, frozen readings, and fall-off events are also common problems
of pulse oximeters [ReichDL1996a, BatchelderBP2007a]. Some studies have dealt
with SpO2 in low-perfusion conditions and motion [ShahN2012a, GehringH2002a].
These investigations, however, have been done in laboratory settings and the
studies intrinsically depend on undisclosed algorithms for the computation of
the SpO2 value.

The second aspect refers to the characteristics of the raw PPG signal depending
on the type of PPG sensor and sensor location. Due to the complex origin of the
photoplethysmography signal, a deep understanding of the PPG signal in relation
to various physiological conditions and sensor location/geometry is still
lacking [SinexJE1999a, ShelleyKH2007a, ReisnerA2008a]. A certain change in the
hemodynamic state will almost certainly manifest itself differently in the PPG
signals measured at different sites on the body.

There is no literature specifically targeting the quality of and information
contained in the raw PPG signals measured at different body sites during
surgery and recovery. The PPG-derived hemodynamic parameters (beyond heart rate
and SpO2) should reflect established methods based on the arterial blood
pressure (ABP) waveform. For example, the PPG-derived measure for the fluid
responsiveness corresponds to the ventilation-induced modulation of the pulse
pressure, commonly known as the pulse pressure variability (PPV). Therefore,
investigating the quality of and information in the raw PPG signals from
different body sites can best be done by correlating the raw PPG signal from
different sites with the intra-arterial blood pressure, fluid status changes,
and vasoactive medication.We aim to investigate (1) whether features in the PPG
signal correlate with features in the ABP signal and (2) whether this
correlation is affected by specific clinical conditions encountered during
surgery (i.e. mechanical ventilation, anesthesia, blood loss, and blood
transfusion), intensive care (i.e. mechanical ventilation, anesthesia, fluid
therapy, and vasopressor administration), and less intensive care (i.e.
spontaneously breathing, no anesthesia, less medication).

The primary objective of the study is to correlate the effects of blood
transfusion, fluid therapy, and vasoactive medication during surgery, recovery,
and intensive care with and without mechanical ventilation on the invasively
measured arterial blood pressure waveforms and the PPG waveforms non-invasively
measured using commercial PPG sensors at the finger, ear, and forehead.This
clinical investigation is needed, because there is not sufficient knowledge
available on the clinical use of PPG for non-invasive hemodynamic monitoring
other than its use for heart rate and SpO2 assessment.

Observational study, compliant with ISO 14155.

Not applicable