|
L. STRACHAN and D.W. NOBLE
Department of Anaesthetics, Intensive Care and Hyperbaric Medicine, Aberdeen Royal
Infirmary, Foresterhill, Aberdeen, UK
|
|
Hypoxia is a common phenomenon peri-operatively. Although mild hypoxaemia of short duration is likely to have little effect more severe and prolonged hypoxaemia can seriously affect surgical outcome. Rational use of oxygen therapy may limit these adverse effects.
Keywords: Bernouille principle, endoscopy, functional residual capacity, hypoxaemia, oxygen, pulse oximetry, venturi
J.R.Coll.Surg.Edinb., 46, October 2001, 297-302
There are many practices, which at various times are under the supervision of the anaesthetist, that can affect the outcome of surgical patients. These include participation in pre-operative assessment, intra-operative temperature management, choice of anaesthetic given, peri-operative drug management, post-operative pain management and oxygen therapy.
Mild to moderate hypoxaemia is common in the postoperative period. It is often unrecognised and its potential to contribute to poor surgical outcome is often underestimated.
The focus of this review is to provide an understanding of the reasons why post-operative oxygen therapy is necessary, with emphasis on the practicalities of delivering oxygen to the patient. Post-operative oxygen therapy need not remain the domain of the anaesthetist, as surgical staff have the most direct contact with their patients in this time period.
It is widely appreciated that extreme hypoxaemia can produce severe and permanent brain injury and/or cardiac arrest. By the time cardiac arrest has occurred, the supervening severe hypoxaemia results in further brain damage, even if successful cardiopulmonary resuscitation has taken place.
However, it is perhaps less readily appreciated that mild to moderate hypoxaemia can also contribute to post-operative morbidity, and in combination with other factors, to increased likelihood of mortality as a result of wide ranging pathophysiological disturbances, as outlined below:
Certain patient groups are predisposed to hypoxaemia, and some patients are at particular risk from hypoxaemia. The former include patients at extremes of age, pregnancy, obesity, smoking and those with co-existing cardio-respiratory disease. 14 The latter will include those with heart disease (both ischaemic and non-ischaemic), cerebrovascular disease, anaemias, haemoglobinopathies and head injured patients.
Even in the absence of complications, the site of surgery, the residual effects of anaesthesia and post-operative analgesia can contribute to hypoxaemia. Superimposed pulmonary complications such as atelectasis, sputum retention, pneumonia and pulmonary thrombo-embolism may occur in patients that are already hypoxaemic from these factors.
Normally 1000 mls/minute (550 mls/min/m2) of oxygen is transported from the lungs to the periphery by the circulation. Only 25% of this is utilised in a resting person.
Satisfactory delivery to tissues depends on a number of factors:
Inhalational anaesthetic agents, even in sub-anaesthetic concentrations, depress compensatory responses to hypoxaemia, hypercapnia and obstruction of the airway.
Opioid analgesia will cause hypoventilation and depress central control of ventilation.
Anaesthetics, sedatives and opioid analgesia may all also contribute to sleep disordered breathing. Direct sedative effects exacerbate the tendency to hypopnoea and apnoeic episodes (overweight male subjects being particularly prone). Additionally, opioids are powerful REM sleep suppressants and there is a dramatic rebound increase in REM sleep when they are withdrawn. During REM sleep, muscular relaxation occurs, including those muscles involved in maintaining airway patency, predisposing to hypopnoea and apnoeic episodes. Many post-operative complications such as myocardial infarction and post-operative delirium have their peak incidence on the third post-operative day, and a link between these complications and sleep-disordered breathing has been proposed. 16
Chronic sleep disordered breathing may be a causal factor in the development of congestive cardiac failure. 17
Gas exchange abnormalities in the post-operative period occur early or late. Early post-operative hypoxaemia is due to anaesthetic factors such as alveolar hypoventilation (above), ventilation/perfusion mismatching, decreased cardiac output and increased oxygen consumption due to shivering (induced by volatile agents or recovery from intra-operative hypothermia). Also, ‘diffusion hypoxia’ may transiently contribute to early hypoxaemia as a result of the very soluble nitrous oxide diffusing out of the circulation into the alveoli when anaesthesia is terminated, reducing the concentration of oxygen in the alveolar gas.
The later onset gas exchange problems are due to alterations in functional residual capacity (FRC) and factors that affect the patient’s ability to inspire deeply or cause the patient to be immobilised in bed.
Functional residual capacity (the volume of air left in the lung after a normal expiration) decreases immediately on induction of anaesthesia.18 This is important because the small airways in the lung periphery are not supported by cartilage and so are influenced by transmitted pleural pressures. Normally, pleural pressures are less than atmospheric, producing a positive transpulmonary pressure, which distends all the airways. Breathing at a low FRC raises atmospheric pleural pressures in gravity-dependent areas of the lung resulting in small airway closure, leading to atelectasis and ventilationperfusion mismatch and hypoxaemia. Patients with pre-existing lowered FRC include the obese, the pregnant, the elderly, neonates and infants.
Site of surgery and, perhaps, the type of incision influence respiratory mechanics. Upper abdominal/ thoracic surgery leads to a reduction in FRC, which is most marked at 24 hours and takes up to 2 weeks to recover. The reduction in PaO2 parallels the FRC changes. In comparison, equally traumatic lower abdominal/pelvic surgery produces less respiratory effect and orthopaedic surgical procedures on the limbs have a still less significant effect. Thus, the site of surgery is an important determinant of post-operative hypoxaemia.
The type of incision may also be important. There is some evidence that horizontal incisions may be less deleterious in this regard compared with vertical incisions. However, the evidence for this is not conclusive.
Careful attention to post-operative fluid balance is required in order to ensure that the patient is normovolaemic to facilitate adequate cardiac output and oxygen carriage to tissues and cells. Factors causing vasoconstriction such as hypovolaemia, hypothermia and pain not only increase cardiac afterload and work but lead to a defective microcirculation and tissue perfusion with resultant tissue hypoxia.
The mechanism responsible for hypoxia may not be immediately obvious but supplemental oxygen should be provided whilst looking for the underlying cause.
Importantly, because of the sigmoid shape of the oxygenhaemoglobin dissociation curve, a small increase in PaO2 in hypoxic patients will usually produce a valuable increase in oxygen saturation and content. In contrast, administration of oxygen in concentrations that produce a higher than normal PaO2, will only marginally increase oxygen content of blood.
MONITORING FOR HYPOXAEMIA CLINICAL ASSESSMENT
Severe hypoxaemia may manifest itself clinically in one or all of the following ways:
It is clinically difficult to detect mild to moderate hypoxaemia. A high index of clinical suspicion is required and a low threshold for the use of pulse oximetry or blood-gas analysis to confirm this hypoxic state.
Figure 1: Sinusoidal breathing pattern. Diagram from Leigh J. Oxygen therapy. Scientific Foundations of Anaesthesia. pp 235-242
Objective measures of monitoring for hypoxaemia include pulse oximetry. This is a good bedside monitor if its limitations are recognised. It is a continuous and non-invasive monitor. Its principal limitation is that, in patients who are receiving supplemental oxygen, it will not reliably detect hypoventilation. Hypoventilation must, in the clinical environment, usually be confirmed by measurement of the PaCO2 by arterial blood gas analysis.
Infrequently, inadequate oxygenation with normal oxygen saturation may occur in cases with very gross anaemia or in situations where the cells are unable to utilise oxygen such as severe sepsis or cyanide poisoning. Mixed venous oxygen saturation measurements may be helpful in these situations but this is only practical in an intensive care setting with a pulmonary artery catheter in situ. Inaccurate readings may also be obtained in patients who have high carboxyhaemoglobin or methaemoglobin concentrations, high concentrations of endogenous or exogenous pigments such as bilirubin or methylene blue as well as with cold extremities and movement artifact.
In most circumstances, the trend in oxygen saturation is more important than the value per se as this can indicate whether the patient is responding to therapy or deteriorating.
This is the ‘gold standard’ monitor of ventilation. Arterial blood gases are needed to obtain accurate data, in particular, evidence of hypoventilation (raised PaCO2) as a reason for hypoxaemia. Arterial blood gases may also give an indication of the metabolic effects of clinically important hypoxaemia. Formal blood gas analysis may also afford accurate estimates of carboxyhaemoglobin and methaemoglobin, the former being particularly important in patients rescued from fires. However, a blood gas is a painful, invasive and intermittent procedure that is time consuming in the setting of a busy ward.
A spectrum of treatments exist for the hypoxic patient. These range from supplemental oxgyen therapy and simple measures such as altering posture. Even sitting a patient up improves FRC, compared with the patient lying down. Physiotherapy can be useful, but most specifically in those patients with copious airways secretions. 19 If the patient is still hypoxic after these ward-based treatments, measures such as continuous positive airway pressure, non-invasive ventilation or invasive ventilation may be required, usually in the setting of an intensive care unit.
Increasing alveolar oxygen concentration (PAO2) increases the pressure gradient for diffusion of oxygen into the pulmonary capillary blood stream.
An important fact is that the flow rate of gas into and out of the lungs is not constant during the respiratory cycle, more approximating a sinusoidal waveform (Figure 1). During normal inspiration, gas entry into the alveoli is, at its fastest (peak inspiratory flow rate), around 35 L/minute. Oxygen flow meters at patients’ bedsides are only calibrated to deliver 15 L/minute, which does not meet this demand and ambient air makes good the deficit (see below).
Oxygen delivery devices (Figure 2) can be divided into two categories, variable performance devices and fixed performance devices. The latter deliver a predictable oxygen concentration to the patient, whereas the former variable devices do not. 20
Variable performance devices include simple masks (e.g. a Hudson mask). During respiration the patient entrains air around the sides of the mask (Figure 3), the amount of air drawn being related to peak inspiratory flow rate. A vigorously breathing patient will draw larger amounts of air around the sides of the mask, therefore, diluting the inspired oxygen concentration. For the patient taking smaller, shallower breaths, the fixed oxygen flow will be supplemented with relatively less air and the patient will breathe a higher concentration of oxygen. This is a welcome feature in the face of a hypoventilating surgical patient who will, as a consequence, breath a higher oxygen-enriched mixture.
By adding a reservoir bag of oxygen to the mask (e.g. a trauma mask or Laerdel bag), an extra supply of oxygen will be provided, thereby, decreasing the need to entrain atmospheric air. This will result in the patient’s peak inspiratory flow rate having a larger amount of oxygen than if there was no reservoir bag. However, unless there is a non-rebreathing valve in situ, the patient will be re-breathing some expired gas and, thus, receiving an unknown amount of oxygen. The other variable performance device, which is commonly used on the wards, is nasal cannulae. This can be a useful device to obtain supplemental oxygen in patients who are intolerant of masks or who require to remove the mask to eat and drink. It is unpleasant for patients to have nasal cannulae and flows above 2 L/minute. High flows of dry oxygen cause dessication of the nasal mucosa, which can lead to crusting and even ulceration of the mucosa.
Figure 2: Oxygen delivery devices. 1. Venturi mask. 2. Hudson mask; 3. Trauma mask; 4. Nasal cannulae
Figure 3: Variable performance devices
Figure 4: Venturi device
Using a fixed performance device, such as theVentimask, the oxygen concentration delivered to the patient is more accurate than that achieved with a simple mask. Traditionally, this type of Venturi device (Figure 4) is used for patients with chronic obstructive pulmonary disease who retain carbon dioxide. However, it is also useful for any patient who has respiratory disease, as the results of blood gas analysis can be interpreted more meaningfully in the knowledge of the inspired oxygen concentration.
A Ventimask works on the Bernouille principle. Energy cannot be created or destroyed but simply converted from one form to another. A Venturi device is a tube with a constriction. At this constriction point, kinetic energy (speed of movement) of flowing gas is increased, therefore, causing a drop in potential energy (or gas pressure). If there is a hole in the tubing at this point of sub atmospheric pressure, air will be entrained in a fixed ratio to the oxygen gas-flow. This means that inflowing oxygen is diluted by atmospheric air prior to entering the patient’s facemask. The total flows through these masks are in the region of 40 to 60 L/minute and, thus, usually exceed the patient’s peak inspiratory flow rate, therefore negating the need to draw variable amounts of air around the sides of the mask.
The Venturi devices are colour-coded and each one is designed to deliver a particular predetermined concentration of oxygen dependant upon setting a given flow of oxygen through a standard flow meter. The recommended flow rate to allow delivery of a defined concentration of oxygen is stated on the packaging.
If dry piped or cylinder oxygen is delivered at a flow rate of 1-4 L/minute by mask, the oropharynx and nasopharynx will provide adequate humidification. At higher flow rates, or if delivered directly to the trachea, humidification may be desirable to prevent drying of airway secretions and disruption of ciliary function, which exacerbate inability to expectorate. Finally, it should be noted that the only way to guarantee delivering 100% oxygen, is by a cuffed tracheal tube.
HOW MUCH AND FOR HOW LONG?
There are no didactic rules as to which patients should receive a certain amount of oxygen. Oxygen therapy should always be monitored, and the period for which it is prescribed should take into account the surgery performed and the patient’s preexisting medical problems, as previously discussed. Dodd et al (2000) suggests that a specific oxygen prescription chart should be used as their audit showed that this improved the process of care. 21
As a guideline, young, fit healthy patients having peripheral surgery should receive oxygen for about 30 minutes in recovery to allow resolution of the effects of diffusion hypoxia, and until they are awake and comfortable and protecting their airway. There is no need to administer high dose oxygen, 4 L/minute being adequate.
A patient having major surgery should receive at least 72 hours of oxygen at concentrations of 28-60%.
Our suggestion is that, in the case of fit patients with no coexisting diseases, a pulse oximeter could be used to decide when to discontinue oxygen therapy. Oxygen saturations should exceed 90% on air before supplemental oxygen is withdrawn. However, if the patient has important co-existing diseases that increase the likelihood of hypoxaemia, is at increased risk of the consequences of hypoxaemia, or clinically significant hypoventilation is a potential problem, then invasive arterial blood gases may give additional useful information to direct oxygen therapy.
A special mention must be made of patients who chronically retain carbon dioxide. These are patients with respiratory disease in whom alveolar ventilation is no longer regulated by arterial carbon dioxide tension, as is the norm. They depend on a hypoxic drive to regulate breathing, as they are believed to be desensitised as a result of chronically high levels of carbon dioxide. Thus, in patients with chronic obstructive pulmonary disease and hypercapnic respiratory failure, oxygen therapy will need fine titration to ensure adequate oxygenation without significant worsening of hypercapnia. These patients will often require advanced respiratory support in an intensive care unit environment post-operatively, particularly following major surgery, when the residual effects of anaesthetics, opioid analgesia and pulmonary effects of surgery significantly compromise respiration in such patients.
For most other post-operative or critically ill patients, however, harm is much more likely to result from lack of oxygen than too much oxygen. Hypercapnia is not uncommon in surgical patients and this is most often due to opioid analgesia. In the vast majority of such patients, without chronic respiratory disease, reducing or withdrawing oxygen to induce hypoxaemia to treat mild hypercapnia is irrational and inappropriate.
Patients with chronic obstructive airways disease and chronic hypercapnia can be identified pre-operatively from venous bicarbonate concentrations as well as arterial blood gas analysis. Such patients, at risk of decompensating into hypercapneic ventilatory failure, are likely to have a high venous and arterial bicarbonate level as bicarbonate is retained to maintain blood pH at the most optimal level.
Discussion of oxygen therapy in a surgical journal would be incomplete without the mention of hypoxia during sedation for endoscopy. Wang et al (2000) found that sedation significantly increased the incidence of desaturation (SaO2 94% or less for up to 15 seconds) and hypoxia (SaO2 92% or less for over 15 seconds). 22 Supplementary nasal oxygen at 4 L/minute, which was well tolerated for this relatively short length of time, abolished these episodes. As one in 2000 patients die following endoscopy, we concur that oxygen supplementation and pulse oximetry should now be routine practice during these procedures and accepted as a standard of care. 23
In summary, this review has sought to illustrate why oxygen therapy is pertinent to good surgical outcome. It is essential that all personnel involved in day-to-day care of surgical patients understand the principles of oxygen therapy in surgical patients.
1. Allen DB, Maguire JJ, Mahdavian M, Wicke C, Marcocci L, Scheuenstuhl H, Chang M, Le
AX, Hopf HW, Hunt TK. Wound hypoxia and acidosis limit neutrophil bacterial killing
mechanisms. Arch Surg 1997; 132: 991-6
2. Hopf HW, Hunt TK, West JM, Blomquist P, Goodson WH, Jensen A et al. Wound tissue
oxygen tension predicts the risk of wound infection in surgical patients. Arch Surg 1997;
132: 997-1005
3. Greif R, Akca O, Horn EP, Kurz A, Sessler DI. Supplemental peri-operative oxygen to
reduce the incidence of wound infection. N Engl J Med 2000; 342: 161-7
4. Van Esbroeck C, Eve T, Hubons A. Evaluation of tissue oximetry in peri-operative
monitoring of colorectal surgery. Br J Surg 1992; 79: 584-7
5. Gennari R, Alexander JW. Effects of hyperoxia on bacterial translocation and mortality
during gut-derived sepsis. Arch Surg 1996; 131: 57-62
6. West M, Li MH, Seatter SC, Bubrick MP. Pre-exposure to hypoxia or septic stimuli
differentially regulates endotoxin release of tumour necrosis factor, interleukin-6,
interleukin-1, prostaglandin E2, nitric oxide, and superoxide by macrophages. J Trauma 1994;
37: 82-90
7. Bertges DJ, Fink MP, Delude RL. Hypoxic signal transduction in critical illness. Crit
Care Med 2000; 28 [Suppl.]: N78-N86
8. Knight PR, Holm BA. The three components of hyperoxia. Anesthesiology 2000; 93:
3-5
9. Gibson GE, Pulsinelli W, Blass JP, Duffy TE. Brain dysfunction in mild to moderate
hypoxia. Am J Med 1981; 79: 1247-54
10. Aakerland LP, Rosenberg J. Post-operative delerium: treatment with supplementary
oxygen. Br J Anaesth 1994; 72: 286-90
11. Davies SW, Wedzicha JA. Hypoxia and the heart. Br Heart J 1993; 69: 3-5
12. Rosenburg-Adamsen S, Lie C, Bernhard A, Kehlet H, Rosenberg J. Effect of oxygen
treatment on heart rate after abdominal surgery. Anesthesiology 1999; 90: 380-4
13. Greif R, Laciny S, Rapf B, Hickle RS, Sesslet DI. Supplemental oxygen reduces the
incidence of postoperative nausea and vomiting. Anesthesiology 1999; 91: 1246-52
14. Craig DB. Post-operative recovery of pulmonary function. Anesth Analg 1981; 60:
46-52
15. Townsend P, Webster NR. Why is this patient still hypoxic despite being ventilated? J
R Coll Surg Edinb 2000; 45: 235-40
16. Hanning CD. Prolonged post-operative oxygen therapy. Br J Anaesth 1992; 69: 115-6
17. Somers VK. To sleep perchance to breath. Implications for the failing heart. Am J
Respir Crit Care Med 1999; 160: 1077-8
18. Wabha RWH. Peri-operative functional residual capacity. Can J Anaesth 1991; 38:
384-400
19. Selsby D, Jones JG. Some physiological and clinical aspects of chest physiotherapy. Br
J Anaesth 1990; 64: 621-31
20. Waldan T, Larsen VH, Bonde J. Evaluation of five oxygen delivery devices in
spontaneously breathing subjects by oxygraphy. Anaesthesia 1998; 53: 256- 63
21. Dodd ME, Kellet F, Davis A, Simpson JCG, Webb AK, Hawsworth CS, Niven RM. Audit of
oxygen prescribing before and after the introduction of a prescription chart. BMJ 2000;
321: 864-5
22. Wang CY, Ling LC, Cardosa MS, Wong AK, Wong NW. Hypoxia during upper gastrointestinal
endoscopy with and without sedation and the effect of pre-oxygenation on oxygen
saturation. Anaesthesia 2000; 55:654-8
23. Charlton JE. Monitoring and supplemental oxygen during endoscopy: one death per 2000
procedures demands action. BMJ 1995; 310: 886-7
FURTHER READING
1. Heys SD, Noble DW, Broom J, Edmond P and Eremin O. Surgical Practice. In: The
Scientific and Clinical Basis of Surgical Practice. Eremin O, Ed. Oxford Medical
Publications: Oxford, 2001: 141-99
2. Nunn JF. Applied Respiratory Physiology, 4th edn. Cambridge:
Butterworth-Heinemann, 1993
Copyright date: 5th June 2001
Correspondence: D.W. Noble, Department of Anaesthetics, Intensive Care and Hyperbaric
Medicine, Aberdeen Royal Infirmary, Foresterhill, Aberdeen AB25 2ZN, UK