can no longer provide adequate excretion of CO 2 . Clinically, this is recognized as tachypnea, tachycardia, intercostal muscle retraction, accessory muscle use, diaphoresis and paradoxical breathing.
The metabolic component of the arterial blood gas: bicarbonate
Measurement of bicarbonate refl ects a patient ’ s acid – base status.
The bicarbonate concentration reported with a blood gas is cal- culated using the Henderson – Hasselbalch equation and repre- sents a single ionic species. Total serum CO 2 (tCO 2 ) content is measured with serum electrolytes and is the sum of the various forms of CO 2 in serum. Bicarbonate is the major contributor to tCO 2 , and additional forms include dissolved CO 2 , carbamates, carbonate, and carbonic acid. The calculated bicarbonate con- centration does not include carbonic acid, carbonate, and carbamates.
Frequently, arterial and venous blood samples are obtained simultaneously, making arterial blood gas bicarbonate and venous serum tCO 2 measurements available. Venous serum tCO 2 content is 2.5 – 3 mEq/L higher than arterial blood gas bicarbon- ate, since CO 2 content is higher in venous than arterial blood and all species of carbon dioxide are included in the determination of tCO 2 . If the blood sample is arterial, the tCO 2 content reported on the electrolyte panel should be 1.5 – 2 mEq/L higher than the calculated bicarbonate. The tCO 2 measured directly with serum electrolytes will be higher because it includes the different forms of CO 2 . Since both blood gas bicarbonate and electrolyte tCO 2 determinations are usually available, there is a split of opinion as to the relative clinical utility of each [63] . A recent review, however, concludes that calculated and measured bicarbonate values are close enough in most cases that either is acceptable for clinical use [64] .
sured ions. Na + and K + account for 95% of cations while HCO3− and Cl − represent 85% of anions [66] . Thus, unmeasured anions are greater than unmeasured cations. The anion gap is the differ- ence between measured plasma cations (Na + ) minus measured anions (Cl − , HCO3−) and is derived:
Total anions Total cations unmeasured
anions
=
+ =
Measured anions
m measured cations Cl tCO2
+
[ ]
+[ ]
+unmeasured cations unmeasured
aanions
unmeasured cations Unmeasured
anions
unme
⎛⎝
)
=⎛⎝[ ]
+)
−
Na++
aasured
cations =
[ ]
−( [ ]
+[ ] )
=
[ ]
−[ ]
Na++ Cl tCO2 Anion gap Na++
(
Cl + ttCO[
2] )
A normal anion gap is 8 – 16 mEq/L. Potassium may be included as a measured cation, although it contributes little to the accuracy or utility of the gap. If K + is included in the calculation, however, the normal range becomes 12 – 20 mEq/L [67] .
A change in the gap involves a change in unmeasured cations or anions. An elevated gap is most commonly due to an accumu- lation of unmeasured anions that include organic acids (i.e. keto- acids or lactic acid), or inorganic acids (i.e. sulfate and phosphate) [68] . A decrease in cations (i.e. magnesium and calcium) will also increase the gap, but the serum level is usually life - threatening.
occurs at pH < 7.20. Respiratory rate and tidal volume increase in an attempt to compensate for the acidosis. Maternal acidosis can result in fetal acidosis as H + ions equilibrate across the placenta, and fetal pH is generally 0.1 pH units less than the maternal pH.
The compensatory response to metabolic acidosis is an increase in ventilation that is stimulated by the fall in the pH.
Hyperventilation lowers PCO 2 as the body attempts to return the HCO3− PCO2
[ ]
ratio toward normal. The respiratory response is proportional to the degree of acidosis and allows calculation of the expected PCO 2 for a given bicarbonate level (Table 5.1 ).When the measured PCO 2 is higher or lower than expected for the measured serum bicarbonate, a mixed acid – base disorder must be present. This formula is ideally applied once the patient has reached a steady state, when PCO 2 nadirs 12 – 24 hours after the onset of acidosis [56] .
The classifi cation of metabolic acidosis as non - anion gap or anion gap acidosis helps determine the pathologic process. Once a metabolic acidosis is detected, serum electrolytes should be obtained to calculate the anion gap. Frequently the clinical history and a few additional diagnostic studies can identify the underly- ing abnormality (Figure 5.5 ) [65] .
Electroneutrality in the body is maintained because the sum of all anions equals the sum of all cations. Na + , K + , Cl − , and HCO3− are the routinely measured serum ions while Mg + , Ca 2+ , proteins (particularly albumin), lactate, HPO4− and SO4− are the unmea-
pH, HCO32–
Calculate anion gap
Elevated anion gap Measure:
Serum glucose Serum ketones Serum creatinine Lactate
Serum osmolality Toxin screen Salicylate level
Ethylene glycol ingestion Lactic acidosis
Methanol ingestion Paraldehyde ingestion Propylene glycol ingestion Salicylate toxicity
Renal failure (late acute or early chronic) Ketoacidosis
Diabetic Alcoholic Starvation Normal anion gap
Gastrointestinal bicarbonate loss Diarrhea
Small bowel fistula Renal tubular acidosis Medication
Carbonic anhydrase inhibitors (e.g., acetozolamide)
Amphotericin B Cyclosporine Cholestyramine Acid ingestion Hypoaldosteronism
Renal failure (early acute or mild chronic)
Figure 5.5 Etiology and evaluation of metabolic acidosis.
The following example demonstrates use of the anion gap in a patient who had been experiencing dysuria, polyuria, and poly- dypsia of several days duration. Initial evaluation of this 19 - year - old gravida at 24 weeks gestation was notable for a serum glucose level of 460 mg/dL and 4+ urinary ketones. Further investigation revealed: arterial pH of 7.30, HCO3− of 14 mEq/L, serum Na + of 133 mEq/L, K + of 4.1 mEq/L, tCO 2 of 15 mEq/L, and Cl − of 95 mEq/L. The anion gap was determined:
Anion gap Na Cl tCO
mEq L mEq L mEq L
=
[ ]
−( [ ]
+[ ] )
= −( + )
=
+ − −
2
133 95 15
1333 110 23
mEq L mEq L Anion gap mEq L
−
=
The elevated anion gap is the result of unmeasured organic anions or ketoacids that have accumulated and decreased serum bicarbonate. As this patient with type I diabetes mellitus receives insulin therapy, the anion gap will normalize, refl ecting disap- pearance of the ketoacids from the serum.
The limitations of the anion gap, however, should be recog- nized. Various factors can lower the anion gap, but its importance is not so much in the etiology of the decrease as in its ability to mask an elevated gap. Since albumin accounts for the majority of unmeasured anions, the gap decreases as albumin levels fall. For each 1 g decrease in albumin, the gap may be lowered by 2.5 – 3 mEq/L. The most common cause of a lowered gap is decreased serum albumin. Other less common causes include markedly elevated levels of unmeasured cations (K + , Mg + , and Ca 2+ ), hyper- lipidemia, lithium carbonate intoxication, multiple myeloma, and bromide or iodide intoxication.
Although an elevated anion gap is traditionally associated with metabolic acidosis, it may also occur in the presence of severe metabolic alkalosis. The ionic activity of albumin changes with increasing pH and protons are released. The net negative charge on each molecule increases, thereby increasing unmeasured anions. Volume contraction leads to hyperproteinemia and aug- ments the anion gap.
If an anion gap acidosis is present, the ratio of the change in the anion gap (the delta gap) to the change in HCO3− can be helpful in determining the type of disturbances present:
∆
∆ gap HCO
Anion gap
3 HCO3
12
− = 24 −−
−
[ ]
In simple anion gap metabolic acidosis, the ratio approximates 1.0, since the decrease in bicarbonate equals the increase in anions. The delta gap for the patient with diabetes and ketoaci- dosis previously described is calculated as follows:
∆
∆ gap HCO
Anion gap
3 HCO3
12 24
23 12 24 14 11 10 1 1
−= −−
−
[ ]
= −
−
= = .
The delta gap is 0 when the acidosis is a pure non - anion gap acidosis. A delta gap of 0.3 – 0.7 is associated with one of two mixed metabolic disorders: (i) a high anion gap acidosis and respiratory alkalosis and (ii) high anion gap with a pre - existing normal or low anion gap. A ratio greater than 1.2 implies a meta- bolic alkalosis superimposed on a high anion gap acidosis or a mixed high anion gap acidosis and chronic respiratory acidosis.
The use of the delta gap is, however, limited by the wide range of normal values for the anion gap and bicarbonate, and its accuracy has been questioned [69] .
When a normal anion gap metabolic acidosis is present, the urinary anion gap may be helpful in distinguishing the cause of the acidosis:
urinary anion gap=
[
urine Na+]
+[
urine K+]
−[
urine Cl−]
The urinary anion gap is a clinically useful method to estimate urinary ammonium ( NH4+) excretion. Since the amount of NH4+
excreted in the urine cannot be directly measured, the urinary anion gap helps determine whether the kidney is responding appropriately to a metabolic acidosis [70] . Normally, the urine anion gap is positive or close to zero. A negative gap (Cl − > Na + and K + ) occurs with gastrointestinal bicarbonate loss and NH4+
excretion by the kidney increases appropriately. In contrast, a positive gap (Cl − < Na + and K + ) in a patient with acidosis suggests impaired distal urinary acidifi cation with inappropriately low NH4+ excretion.
A variety of processes can lead to metabolic acidosis and therapy will depend on the underlying condition. Adequate oxy- genation should be ensured and mechanical ventilation instituted for impending respiratory failure. The use of bicarbonate solu- tions to correct acidosis has been suggested when arterial pH is less than 7.10 or bicarbonate is lower than 5 mEq/L. Bicarbonate solutions must be administered with caution since an “ over- shoot ” alkalosis can lower seizure threshold, impair oxygen avail- ability to peripheral tissues, and stimulate additional lactate production.
Metabolic alkalosis
Metabolic alkalosis is characterized by a rise in serum bicarbonate concentration and an elevated arterial pH. The most impressive clinical effects of metabolic alkalosis are neurologic and include confusion, obtundation, and tetany. Cardiac arrhythmias, hypo- tension, hypoventilation and various metabolic aberrations may accompany these neurologic changes.
Metabolic alkalosis results from a loss of acid or the addition of alkali. The development of metabolic alkalosis occurs in two phases, with the initial addition or generation of HCO3− followed by the inability of the kidney to excrete the excess HCO3−. The two most common causes of metabolic alkalosis are excessive loss of gastric secretions and diuretic administration. Once established, volume contraction, hypercapnea, hypokalemia, glucose loading, and acute hypercalcemia promote HCO3− reabsorption by the kidney and sustain the alkalosis.
a responsive disorder, infusion of sodium chloride will correct the abnormality. Conversely, saline administration will not correct a chloride resistant disorder and can be harmful.
Treatment of the primary disease will concurrently correct the alkalosis. Although mild alkalemia is generally well tolerated, critically ill surgical patients with a pH ≥ 7.55 have increased mortality [72,73] .
Respiratory acidosis
Respiratory acidosis is characterized by hypercapnea (a rise in PCO 2 ) and a decreased arterial pH. The development of respira- tory acidosis indicates the failure of carbon dioxide excretion to match CO 2 production. A variety of disorders can contribute to this acid – base abnormality (Table 5.4 ). It is important to remem- ber that the normal PCO 2 in pregnancy is 30 mmHg, and norma- tive data for non - pregnant patients do not apply to the gravida.
The clinical manifestations of acute respiratory acidosis are particularly evident in the central nervous system. Since carbon dioxide readily penetrates the blood – brain barrier and cerebro- spinal fl uid buffering capacity is not as great as blood, PCO 2 elevations quickly decrease the pH of the brain. Thus, neurologic compromise may be more signifi cant with respiratory acidosis than metabolic acidosis [59] . Acute hypercapnia also decreases The degree of respiratory compensation for metabolic alkalosis
is more variable than with metabolic acidosis, and formulas to estimate the expected P a CO 2 have not proven useful [56] . Alkalosis tends to cause hypoventilation but P a CO 2 rarely exceeds 55 mmHg [56,71] . Tissue and red blood cells attempt to lower HCO3− by exchanging intracellular H + ions for extracellular Na + and K + .
Once metabolic alkalosis is diagnosed, determination of urinary chloride concentration can be helpful in determining the etiology (Figure 5.6 ). Urinary chloride is a more reliable indicator of volume status than urinary sodium concentration in this group of patients. Sodium is excreted in the urine with bicarbonate to maintain electroneutrality and occurs independently of volume status. Therefore, low urinary chloride in patients with volume contraction accurately refl ects sodium chloride retention by the kidney.
A urinary chloride concentration < 10 mEq/L that improves with sodium chloride administration is a chloride - responsive metabolic alkalosis. In contrast, a urine chloride > 20 mEq/L indi- cates that the alkalosis will not improve with saline administra- tion and is a chloride - resistant alkalosis. Urine chloride levels must be interpreted with caution since levels are falsely elevated when obtained within several hours of diuretic administration.
Treatment of metabolic alkalosis is aimed at eliminating excess bicarbonate and reversing factors responsible for main- taining the alkalosis. If the urinary chloride level indicates
pH, HCO3– Measure urinary chloride
Chloride resistant (Urine Cl– > 20 mEq/L) Hypertensive:
Mineralocorticoid excess Hyperaldosteronism Normotensive:
Magnesium depletion Diuretic use (current) Profound hypokalemia Alkali ingestion Bicarbonate therapy Antacids
Lactate Acetate Citrate
Massive blood transfusion Hypercalcemia
Medications Carbenicillin Penicillin Sulfates Parathyroid disease Refeeding alkalosis Chloride responsive
(Urine Cl– < 10 mEq/L) Gastrointestinal Vomiting
Nasogastric suction Diuretic use (discontinued) Contraction alkalosis Posthypercapnea
Figure 5.6 Etiology and evaluation of metabolic alkalosis.
Table 5.4 Causes of respiratory acidosis.
Airway obstruction Aspiration Laryngospasm Severe bronchospasm
Impaired ventilation Pneumothorax Hemothorax Severe pneumonia Pulmonary edema
Adult respiratory distress syndrome
Circulatory collapse Massive pulmonary embolism Cardiac arrest
CNS depression Medication Sedatives Narcotics
Cerebral infarct, trauma or encephalopathy Obesity – hypoventilation syndrome
Neuromuscular disease Myasthenic crisis Severe hypokalemia Guillain – Barr é Medication
acid – base disorder in which the compensatory response can return the pH to normal.
Respiratory alkalosis may be diagnostic of an underlying con- dition and is usually corrected with treatment of the primary problem. Hypocapnea itself is not life - threatening but the disease causing the alkalosis may be. The presence of respiratory alkalosis should always raise suspicion for hypoxemia, pulmonary embo- lism, or sepsis. These conditions, however, can be overlooked if the only concern is correction of the alkalosis. Mechanical venti- lation may lead to iatrogenic respiratory alkalosis and the PCO 2 can usually be corrected by lowering the machine - set respiratory rate. [75]
References
1 Kruse JA . Acid – base interpretations . Crit Care 1993 ; 14 : 275 . 2 MacRae DJ , Palavradji. Maternal acid – base changes in pregnancy . J
Obstet Gynaecol Br Cwlth 1967 ; 74 : 11 .
3 Hankins GDV , Harvey CJ , Clark SL , Uckan EM . The effects of mater- nal position and cardiac output on intrapulmonary shunt in normal third - trimester pregnancy . Obstet Gynecol 1996 ; 88 : 327 .
4 Cruikshank DP , Hays PM . Maternal physiology in pregnancy . In:
Gabbe S , Niebyl J , Simpson JL , eds. Obstetrics: Normal and Problem Pregnancies , 2nd edn. New York : Churchill Livingstone , 1991 : 129 . cerebral vascular resistance, leading to increased cerebral blood
fl ow and intracranial pressure.
The compensatory response depends on the duration of the respiratory acidosis. In acute respiratory acidosis, the respiratory center is stimulated to increase ventilation. Carbon dioxide is neutralized in erythrocytes by hemoglobin and other buffers, and bicarbonate is generated. An acute disturbance implies that renal compensation is not yet complete. Sustained respiratory acidosis (longer than 6 – 12 hours) stimulates the kidney to increase acid excretion, but this mechanism usually requires 3 – 5 days for full compensation [74] .
The primary goal in the management of respiratory acidosis is to improve alveolar ventilation and decrease arterial PCO 2 . Assessment and support of pulmonary function are paramount when a patient has respiratory acidosis. Carbon dioxide accumu- lates rapidly, and PCO 2 rises 2 – 3 mmHg/min in a patient with apnea. The underlying condition should be rapidly corrected and may include relief of an airway obstruction or pneumothorax, administration of bronchodilator therapy, narcotic reversal, or a diuretic.
Adequate oxygenation is crucial because hypoxemia is more life - threatening than hypercapnia. In the pregnant patient, hypoxemia also compromises the fetus. Uterine perfusion should be optimized and maternal oxygenation ensured since the com- bination of maternal hypoxemia and uterine artery hypoperfu- sion profoundly affects the fetus. When a patient cannot maintain adequate ventilation despite aggressive support, endotracheal intubation and mechanical ventilation should be performed without delay.
Respiratory alkalosis
Respiratory alkalosis is characterized by hypocapnea (decreased PCO 2 ) and an increased arterial pH. Acute hypocapnea frequently is accompanied by striking clinical symptoms, including pares- thesias, circumoral numbness, and confusion. Tachycardia, chest tightness, and decreased cerebral blood fl ow are some of the prominent cardiovascular effects. Chronic respiratory alkalosis, however, is usually asymptomatic.
Respiratory alkalosis is the result of increased alveolar ventila- tion (Table 5.5 ). Hyperventilation can develop from stimulation of brainstem or peripheral chemoreceptors and nociceptive lung receptors. Higher brain centers can override chemoreceptors and occurs with involuntary hyperventilation. Respiratory alkalosis is commonly encountered in critically ill patients in response to hypoxemia or acidosis, or secondary to central nervous system dysfunction.
The compensatory response is divided into acute and chronic phases. In acute alkalosis, there is an instantaneous decrease in H + ion concentration due to tissue and red blood cell buffer release of H + ions. If the duration of hypocapnea is greater than a few hours, renal excretion of bicarbonate is increased and acid excretion is decreased. This response requires at least several days to reach a steady state. Chronic respiratory alkalosis is the only
Table 5.5 Causes of respiratory alkalosis.
Pulmonary disease Pneumonia Pulmonary embolism Pulmonary congestion Asthma
Drugs Salicylates Xanthines Nicotine
CNS disorders
Voluntary hyperventilation Anxiety
Neurologic disease Infection Trauma
Cerebrovascular accident Tumor
Other causes Pregnancy Pain Sepsis Hepatic failure
Iatrogenic mechanical hyperventilation
28 Bloom SA , Canzanello VJ , Strom JA , Madias NE . Spurious assessment of acid – base status due to dilutional effect of heparin . Am J Med 1985 ; 79 : 528 .
29 New W . Pulse oximetry . J Clin Monit 1985 ; 1 : 126 .
30 Al - Ameri MW , Kruse JA , Carlson RW . Blood sampling from arterial catheters: minimum discard volume to achieve accurate laboratory results . Crit Care Med 1986 ; 14 : 399 .
31 Bhaskaran NC , Lawler PG . How much blood for a blood gas?
Anesthesiology 1988 ; 43 : 811 .
32 Biswas CK , Ramos JM , Agroyannis B , Kerr DNS . Blood gas analysis:
effect of air bubbles in syringe and delay in estimation . Br Med J 1982 ; 284 : 923 .
33 Morganroth ML . Six steps to acid – base analysis: clinical applications . J Crit Ill 1990 ; 5 : 460 .
34 Harsten A , Berg B , Inerot S , Muth L . Importance of correct handling of samples for the results of blood gas analysis . Acta Anesthesiol Scand 1988 ; 32 : 365 .
35 Mueller RG , Lang GE . Blood gas analysis: effect of air bubbles in syringe and delay in estimation . Br Med J 1982 ; 285 : 1659 .
36 Madiedo G , Sciacca R , Hause L . Air bubbles and temperature effect on blood gas analysis . J Clin Pathol 1980 ; 33 : 864 .
37 Schuch CS , Price JG . Determination of time required for blood gas homeostasis in the intubated, post - open - heart surgery adult following a ventilator change . NTI Res Abs 1986 ; 15 : 314 .
38 McHugh RD , Epstein RM , Longnecker DE . Halothane mimics oxygen in oxygen microelectrodes . Anesthesiology 1979 ; 50 : 47 .
39 Douglas IHS , McKenzie PJ , Ledingham IM , Smith G . Effect of halo- thane on PO 2 electrode . Lancet 1978 ; 2 : 1370 .
40 Maekawa T , Okuda Y , McDowall DG . Effect of low concentrations of halothane on the oxygen electrode . Br J Anaesth 1980 ; 52 : 585 . 41 Dent JG , Netter KJ . Errors in oxygen tension measurements caused
by halothane . Br J Anaesth 1976 ; 48 : 195 .
42 Hess CE , Nichols AB , Hunt WB . Pseudohypoxemia secondary to leukemia and thrombocytopenia . N Engl J Med 1979 ; 301 : 363 . 43 Nearman HS , Sampliner JE . Respiratory monitoring . In: Berk JL ,
Sampliner JE , eds. Handbook of critical care , 3rd edn. Boston : Little Brown , 1982 : 125 – 143 .
44 Demling BK , Knox JB . Basic concepts of lung function and dysfunc- tion: oxygenation, ventilation and mechanics . New Horiz 1993 ; 1 : 362 . 45 Huch A , Huch R , Konig V et al. Limitations of pulse oximetry . Lancet
1988 ; 1 : 357 .
46 Dildy GA , Loucks CA , Porter TF , Sullivan CA , Belfort MA , Clark SL . Many normal pregnant women residing at moderate altitude have lower arterial oxygen saturations than expected . Society for Gynecologic Investigation , Atlanta, GA , March 1998 .
47 Dildy GA , Sullivan CA , Moore LG , Richlin ST , Loucks CA , Belfort MA , Clark SL . Altitude reduces and pregnancy increases maternal arte- rial oxygen saturation . Society for Maternal - Fetal Medicine , San Francisco, CA , January 1999 .
48 Richlin S , Cusick W , Sullivan C , Dildy GA , Belfort MA . Normative oxygen saturation values for pregnant women at sea level . The American College of Obstetricians and Gynecologists , New Orleans, LA , May 1998 .
49 Goldberg M , Green SB , Moss ML et al. Computer - based instruction and diagnosis of acid – base disorders . JAMA 1973 ; 223 : 269 . 50 Davenport HW . Normal acid – base paths . In: The ABC of Acid – Base
Chemistry , 6th edn. Chicago : University of Chicago Press , 1974 : 69 .
5 Artal R , Wiswell R , Romem Y , Dorey F . Pulmonary responses to exercise in pregnancy . Am J Obstet Gynecol 1986 ; 154 : 378 .
6 Liberatore SM , Pistelli R , Patalano F et al. Respiratory function during pregnancy . Respiration 1984 ; 46 : 145 .
7 Andersen GJ , James GB , Mathers NP et al. The maternal oxygen tension and acid – base status during pregnancy . J Obstet Gynaecol Br Cwlth 1969 ; 76 : 16 .
8 Dayal P , Murata Y , Takamura H . Antepartum and postpartum acid – base changes in maternal blood in normal and complicated pregnan- cies . J Obstet Gynaecol Br Cwlth 1972 ; 79 : 612 .
9 Templeton A , Kelman GR . Maternal blood - gases, (PA|d2dend - Pa|d2dend (Maternal blood - gases, PaO2 – PaO2, physiological shunt and VD/VT in normal pregnancy) physiological shunt and V|dDdend/
V|dTdend in normal pregnancy . Br J Anaesth 1976 ; 48 : 1001 . 10 Pernoll ML , Metcalfe J , Kovach PA et al. Ventilation during rest
and exercise in pregnancy and postpartum . Respir Physiol 1975 ; 25 : 295 .
11 Awe RJ , Nicotra MB , Newsom TD , Viles R . Arterial oxygenation and alveolar - arterial gradients in term pregnancy . Obstet Gynecol 1979 ; 53 : 182 .
12 Hankins GDV , Clark SL , Uckan EM et al. Third trimester arterial blood gas and acid – base values in normal pregnancy at moderate altitude . Obstet Gynecol 1996 ; 88 : 347 .
13 Sobrevilla LA , Cassinelli MT , Carcelen A et al. Human fetal and maternal oxygen tension and acid – base status during delivery at high altitude . Am J Obstet Gynecol 1971 ; 111 : 1111 .
14 Novy MJ , Edwards MJ . Respiratory problems in pregnancy . Am J Obstet Gynecol 1967 ; 99 : 1024 .
15 Weinberger SE , Weiss ST , Cohen WR et al. Pregnancy and the lung . Am Rev Respir Dis 1980 ; 121 : 559 .
16 Lucius H , Gahlenbeck H , Kleine HO et al. Respiratory functions, buffer system, and electrolyte concentrations of blood during human pregnancy . Respir Physiol 1970 ; 9 : 311 .
17 Cain SM . Peripheral uptake and delivery in health and disease . Clin Chest Med 1983 ; 4 : 139 .
18 Stock MC , Shapiro BA , Cane RD . Reliability of SvO 2 in predicting A - VDO 2 and the effect of anemia . Crit Care Med 1986 ; 14 : 402 . 19 Bryan - Brown CW , Back SM , Malcabalig et al. Consumable oxygen:
oxygen availability in relation to oxyhemoglobin dissociation . Crit Care Med 1973 ; 1 : 17 .
20 Perutz MF . Hemoglobin structure and respiratory transport . Sci Ann 1978 ; 239 : 92 .
21 Rackow EC , Astiz M . Pathophysiology and treatment of septic shock . JAMA 1991 ; 266 : 548 .
22 Shoemaker WC , Ayers S , Grenuik A et al. Textbook of Critical Care , 2nd edn. Philadelphia : WB Saunders , 1989 .
23 Shibutani K , Komatsu T , Kubal K et al. Critical levels of oxygen delivery in anesthetical man . Crit Care Med 1983 ; 11 : 640 .
24 Barron W , Lindheimer M . Medical Disorders During Pregnancy , 1st edn. Mosby - Year Book , St Louis , 1991 : 234 .
25 Gemzell CA , Robbe H , Strom G et al. Observations on circulatory changes and muscular work in normal labor . Acta Obstet Gynecol Scand 1957 ; 36 : 75 .
26 Ueland K , Hansen JM . Maternal cardiovascular hemodynamics:
II. Posture and uterine contractions . Am J Obstet Gynecol 1969 ; 103 : 1 .
27 Walton JR , Shapiro BA , Wine C . Pre - analytic error in arterial blood gas measurement . Respir Care 1981 ; 26 : 1136 .