Purpose: Carbon dioxide (CO2) is a major end-product of metabolism and can have marked
effects on central nervous system function. It can also be easily manipulated during
general anesthesia via controlled ventilation. High levels of carbon dioxide
(hypercapnia) are associated with sedation and have been shown to produce selective
suppression of thermal and ischaemic pain in animals and humans. This effect was
attenuated by dexamethasone and naloxone, indicating that stress pathways and endogenous
opioids may be implicated. Hypercapnia during anesthesia may have additional benefits,
including reduced levels of wound infection due to improved tissue oxygenation reduced
incidence of postoperative nausea and vomiting and reduced recovery time from volatile
anesthetic.
It is a known phenomenon for high levels of CO2 to be associated with reduced levels of
consciousness in humans, known as CO2 narcosis. A 1927 paper described narcosis of
animals when breathing 30-40% CO2 in oxygen, with prompt recovery upon removal. The
authors described a 'sharp sour taste' and associated hypertension when the same solution
was administered to humans. However, few studies investigate the impact of carbon dioxide
on anesthetic requirements. An animal study from 1967 demonstrated that very high levels
of CO2 (>95 mmHg) offset halothane requirements in dogs. Most recently, increased carbon
dioxide levels during surgery (40 - 45 mmHg) were shown to reduce the Minimal Alveolar
Concentration to Blunt Adrenergic Response (skin incision; MAC-BAR) of sevoflurane in
adult patients undergoing gastric carcinoma resection.
Total intravenous anesthesia (TIVA), an alternative to inhalational anesthesia, is a
commonly used anesthetic technique in the investigator's institution. This is due to its
many benefits, including reduced emergence delirium, reduced environmental impact and
reduced post-operative nausea and vomiting. Administration can be guided by depth of
anesthesia monitoring such as the Bispectral Index (BIS), which measures the patient's
level of consciousness derived from electroencephalogram readings. BIS has been shown to
help guide propofol dosing in children regardless of whether the TIVA technique was
target controlled or a manual infusion regimen, and to correlate well with both modelled
and measured propofol levels in children.
The investigator's study aims to determine whether differing levels of CO2 affect the
anesthetic depth in anesthetized children, as measured by BIS.
Hypothesis: Hypercarbia is associated with a reduction in BIS readings, in anesthetized
children.
Justification: The impact of EtCO2 on BIS has not been studied in children. If
discovered, a correlation between the two could significantly change anesthetic practice.
It is straightforward to increase EtCO2 levels in anesthetized patients, and if this was
found to reduce their anesthetic requirements it could enable lower rates of anesthetic
drug administration. This would benefit the patient by exposing them to less medication
and fewer associated side effects, as well as benefitting the hospital and wider
environment by reducing cost and use of disposable equipment such as ampoules, packaging
and syringes.
Objectives: (1) To determine the effect of EtCO2 on the depth of anesthesia in children,
as measured by BIS.
(2) Patient movement as detected clinically by the surgical or anesthetic team.
Research Design: The investigators plan to conduct a randomized, prospective, crossover
trial. The within-subject design allows patients to act as their own controls. The order
in which the EtCO2 levels are tested will be randomized between patients using sealed
envelopes. The anesthesiologist in the room will be blinded to the BIS reading but will
be informed by a research assistant if it reads persistently high (>60) for over one
minute.
Statistical analysis: Physiological data will be collected in real-time using
purpose-built software. Patient demographics and characteristics will be collected by a
research assistant. Time-series plots of BIS and EtCO2 for each participant will be made
in R (R Foundation for Statistical Computing, Vienna, Austria). Generalized estimating
equations (GEE), using the geepack package, will be applied to estimate the effect of
changes in target EtCO2 on serial BIS measurements during the anesthesia maintenance
phase. Each participant will be considered their own data cluster to provide an
appropriate grouping structure for the analysis. An independent correlation structure
will be applied, and a robust (sandwich) estimator method will be used to obtain standard
errors.