Intra-operative gas exchange and post-operative hypoxaemia.

The relation between intra-operative gas exchange and post-operative oxygen saturation was examined using pulse oximetry (SpO2) 30 h after anaesthesia for upper abdominal surgery. The inspired oxygen partial pressure (PIO2) was varied during anaesthesia to produce a plot of PIO2 vs. SpO2. Gas exchange during anaesthesia was inferred from the relation between PIO2 and SpO2 which was described by its lateral and vertical displacement away from the position of a standard oxygen dissociation curve. With patients breathing air, the SpO2 was measured over a 10-min period at 30 h after surgery. The correlation coefficient of the lateral displacement with post-operative SpO2 was 0.7, and that of the vertical displacement with post-operative SpO2 was 0.66. The correlation coefficient of intra-operative SpO2 at PIO2 = 21 kPa with post-operative SpO2 was 0.69. It can be concluded that gas exchange abnormalities during anaesthesia deduced from the displacement of the PIO2 vs. SpO2 curve correlate with SpO2 30 h post-operatively and may help to identify patients at risk of post-operative hypoxaemia.

A model to describe the rate of oxyhaemoglobin desaturation during apnoea.

We have developed a model to describe the rate of oxyhaemoglobin desaturation during apnoea. This model takes into account the non-steady-state kinetics which pertain to this situation. We first derived a mathematical expression for instantaneous oxygen flux rate from the alveolar compartment. We then derived an expression to describe the effect of shunt on this flux. The effect of circulation time on real-time arterial mixed venous oxygen content difference and oxygen flux in the lung was determined graphically. We finally described a manoeuvre to accommodate the effect of the Bohr shift which is related to the increase in FACO2 during apnoea. We present plots of arterial oxyhaemoglobin saturation (SaO2) vs duration of apnoea to illustrate the individual effects of the initial fractional concentration of oxygen in the alveolus (FAO2initial), alveolar volume (VA), shunt fraction (QS/QT), oxygen consumption rate (VO2), total blood volume (QT) and haemoglobin concentration (Hb). The model is illustrated by examples of paediatric, morbidly obese and post-operative scenarios. The postoperative scenario is particularly notable for the effect of a combination of small changes in individual variables leading to a large overall effect on the rate of oxyhaemoglobin desaturation.

Causes of oxyhaemoglobin saturation instability in the postoperative period.

Healthy, young patients tend to have stable patterns of oxygen saturation (SaO2) in the perioperative period, but unstable patterns of oxygen saturation data in the postoperative period have been described which are associated with wide fluctuations in the value of SaO2 over short periods of time and with a propensity for severe desaturation. This investigation was designed to study the cause of these unstable, hypoxaemic patterns. SpO2 was measured by pulse oximetry in an "at risk" patient group (n = 20) before and after operation and displayed as sequential distribution diagrams and centile plots. In 16 of these patients SpO2 was determined also at different values of PlO2 (21-50 kPa) before and after operation. Instability of SpO2, assessed over periods of 1 h, was common (n = 18) after operation; it was caused by both acute desaturations and a fluctuating baseline SpO2. The PlO2 vs SpO2 relationship was displaced after operation, with both lateral (rightwards) and downwards displacement. The result was to bring the PlO2 to or near to the steep part of the curve. In patients who displayed an unstable pattern of SpO2 over a short period of time (< 5 min), this was abolished by increasing PlO2. We postulate that the displacement of the PlO2 vs SpO2 curve increases the propensity to desaturation in those patients with impaired control by placing the steep part of the curve near to a PlO2 value of 21 kPa.

Analysis of factors which affect the relationship between inspired oxygen partial pressure and arterial oxygen saturation.

The adoption of pulse oximetry in anaesthesia and intensive care has resulted in oxyhaemoglobin saturation (SaO2) becoming an important and widely used clinical descriptor of gas exchange, although its full potential has not been realized. We have explored the effect of impaired gas exchange on the relationship between PIO2 and SaO2. A curve of oxygen content vs PIO2 for an ideal lung was constructed using the oxygen dissociation curve and allowing for the effect of dissolved oxygen in plasma. A plot of PIO2 vs SaO2 was derived from this content curve. The effect of impaired gas exchange (shunt and V/Q effects) on this relationship was then modelled using for shunt the equation Qs/Qt = (Cc' O2-CaO2)/(Cc' O2-CvO2) and for V/Q, a graphical method using the equation V/Q = k(Cc' O2-CvO2)/ (PIO2-PAO2) and the oxygen content curve to model the effect of a spread of V/Q ratios. A total of nine compartments were used to model the scatter of V/Q, and the size of these compartments were determined by their relative blood supply. Plots of PIO2 vs SaO2 were derived for different values of shunt and V/Q. The PIO2 vs SaO2 curve reflected the shape of the oxygen dissociation curve but lay to the right by PIO2-PAO2. Shunt caused a depression of the plateau of this curve with relatively little lateral movement, whereas a low mean V/Q ratio caused a marked rightwards lateral displacement. Increased IogSD Q caused a flattening of the relationship. The combined effect of shunt and a spread of V/Q ratios caused both lateral and vertical displacements. Thus a series of simple measurements of PIO2 and SaO2 gave information about both shunt and V/Q abnormality.

Liver function tests in the critically ill patient.

The liver has a wide range of functions that may be disturbed in different ways by the many diseases which affect it and, in consequence, there are a large number of tests which look at different aspects of its function. Specific diagnoses are made using a range of clinical, biochemical, histological and radiological methods. Measurement of the plasma concentration of alanine aminotransferase (ALT, SGPT), gamma-glutamyl transpeptidase (gammaGT) and albumin are particularly valuable as these substances are specifically affected by liver disease. Their elevation can reveal increases in the membrane permeability of hepatocytes (ALT), cholestasis and toxic damage (gammaGT), or an impairment of liver protein synthesis and secretion (albumin), respectively. If their activities remain within the normal range the likelihood of significant liver disease is less than 2%. A series of quantitative liver function tests are described which each examine one aspect of hepatic function. Table 1 lists several situations where the assessment of hepatic function is necessary. Liver function tests are classified in Table 2.

Patterns of oxygenation after thoracotomy.

We have studied patterns of oxygen saturation (SpO2) before and after thoracotomy in 20 patients monitored nightly from the preoperative night to the fourth postoperative night. After operation, 10 patients received paravertebral bupivacaine (PVB) infusion and 10 received paravertebral saline (PVS) infusion. Papaveretum was given as required. Before operation the SpO2 profiles formed two groups: stable with SpO2 greater than 94% and stable with a median SpO2 less than 94% (hypoxaemia). During the first night after operation SpO2 profiles formed four groups: stable, not hypoxaemic (2/20); stable, hypoxaemic but improving (8/20); stable and constant hypoxaemia (5/20); unstable, hypoxaemic and deteriorating (5/20). Eleven patients remained hypoxaemic as late as the fourth night after operation. All patients who were hypoxaemic before operation were hypoxaemic after operation. Postoperative hypoxaemia was predicted in only 50% of cases. Papaveretum requirement was reduced in the PVB group, but regional analgesia did not affect the proportion of patients showing each SpO2 profile. Papaveretum caused a decrease in SpO2 in both analgesic groups.

Oxygen analysers. An evaluation of five fuel cell models.

Five currently available fuel cell oxygen analysers were studied with a view to their use in anaesthesia. The accuracy, response time and safety features of these analysers are discussed. Fuel cell analysers appear to be suitable oxygen monitors for routine anaesthetic use.