Show Summary Details
Page of

Pathophysiology and therapeutic strategy of respiratory acidosis 

Pathophysiology and therapeutic strategy of respiratory acidosis
Pathophysiology and therapeutic strategy of respiratory acidosis

Luciano Gattinoni

and Alfredo Lissoni

Page of

PRINTED FROM OXFORD MEDICINE ONLINE ( © Oxford University Press, 2021. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a title in Oxford Medicine Online for personal use (for details see Privacy Policy and Legal Notice).

date: 26 January 2021

Key points

  • Respiratory acidosis is a process in which the low pH is primarily due to a rise in PCO2. In critically-ill patients the CO2 load may be greatly increased due to hypermetabolism, excessive parenteral nutrition, or titration of HCO3 by fixed acids.

  • CO2 buffering in vivo leads to an increase in HCO3 of 1 mmol/L for every increase of 10 mmHg (1.33 kPa) in PCO2 above 40 mmHg (5.33 kPa). In chronic conditions the kidney-related increase of strong ion difference (reduction in plasma chloride) leads to a plasma HCO3 increase of approximately 3.5 mmol/L for every increase of 10 mmHg (1.33 kPa) in PCO2 above 40 mmHg (5.33 kPa).

  • If respiratory acidosis is present, identifying the precipitating factors and assessing whether the associated signs (hypoxaemia, hypercapnia, and acidosis) are themselves life-threatening is fundamental. Assessment of the precipitating factors and the time required for their correction is of paramount importance in planning therapeutic strategy.

  • Since hyperinflation of the lung, with its cardiovascular consequences, is the most harmful complication of mechanical ventilation in chronic respiratory acidosis, it is essential to set a small tidal volume associated with an adequate expiratory time. Non-invasive ventilation appears to be a promising treatment of chronic respiratory acidosis.

  • The artificial lungs to remove CO2 extracorporeally and are increasingly employed during bridge to lung transplantation, acute respiratory distress syndrome, and the hyperinflation states of COPD re-exacerbation or status asthmaticus.



Although carbon dioxide (CO2) is not an acid per se, it becomes a proton (H+) donor when it is combined with water:

C O 2 + H 2 O H 2 C O 3 H + + H C O 3
[eqn 1]

When PCO2 increases above its physiological levels (i.e. above 40–45 mmHg (5.33–6 kPa), corresponding to 1.2–1.35 mmol/L), acidaemia usually develops. Thus, respiratory acidosis can be defined as a process in which the tendency for the arterial pH to decrease is primarily due to an increase in PCO2.

The CO2 load

In normal man, the amount of CO2 produced metabolically is approximately 200–250 mL/min, corresponding to an acid load of 13,000–16,000 mmol/day. This amount may be increased by 30–100% because of increased metabolism, burns, infection, and an excessive calorific intake, particularly during parenteral nutrition [1]‌. An additional source of CO2, not metabolically produced, is the titration of HCO3 by fixed acid, as lactic acid (LAH), particularly relevant in critically-ill patients.

Excretion of CO2

The acid load expired by the lungs can be expressed as:

V C O 2   =   K   ×   P a C O 2 ×   V A
[eqn 2]

where VCO2 is the expired CO2, VA is alveolar ventilation, PaCO2 is the partial pressure of CO2 in arterial blood, and K is a factor converting pressure to volume fraction. Of note, we will refer always to the alveolar ventilation, which is the fraction of the minute ventilation, which actually reaches the perfused alveoli. The difference between the total ventilation and the alveolar ventilation is modelled as physiological dead space, in part anatomical (ventilation of regions not designed for gas exchange) and in part alveolar, i.e. the fraction of alveolar ventilation ventilating unperfused alveoli.

VCO2 can also be expressed according to Fick’s law as:

VCO 2   =   Q ( C v C O 2     CaCO 2 )
[eqn 3]

where Q is the cardiac output, and CvCO2 and CaCO2 are, respectively, the CO2 contents of mixed venous and arterial blood. Assuming, for simplicity, that the relationship between CO2 content and partial pressure is linear in the range of physiological interest, and combining and rearranging equations 2 and 3, it follows that:

P a C O 2 ( Q / ( V A   +   Q ) ) P v C O 2
[eqn 4]

This equation shows that PaCO2, for a given acid load (which increases the PvCO2) is directly proportional to cardiac output and inversely proportional to alveolar ventilation.

Body response to respiratory acidosis

Buffer system (minutes)

Of the 15,000 to 30,000 mmol of protons delivered daily to the extracellular fluid, only 40–60 nmol/L (i.e. only 1 mmol of H+ out of 1 million) are found free in the blood. The remainder are bound by the buffer system. Buffers that are present in extracellular and intracellular fluids, are substances characterized by pK values (i.e. –log 10 K, where K is the acid dissociation constant) close to the pH value and which are able to bind or release H+, thus preventing large changes in free H+ concentration. In general, buffers are weak acids, which may be present electrically neutral (AH) or electrically charged (A). At a given pH their electrical status is defined by the Henderson–Hasselbalch equation:

pH = pK + ( A / A H )
[eqn 5]

where [A] and [AH] are the concentrations of dissociated base and undissociated acid, respectively. The pK values and the concentrations of the various extracellular buffers are listed in Table 113.1.

Table 113.1 The blood buffers

Buffer pairs

Normal concentration (dissociated + undissociated)


Base-to-acid ratio at pH 7.4 (dissociated/undissociated)

Bicarbonate–carbonic acid (mmol/L)



[HCO3]/[CO2] = 20/1

Dibasic–monobasic phosphate (mmol/L)



[HPO42–]/H2PO4 = 4/1

Proteins (mmol/L)

14 (7 g/dL)


Pr–/HP = 4/1

Haemoglobin (mmol/L)

90 (15 g/dL)


Hb–/HHb = 4/1

Of note, only proteins and phosphates are available to buffer the CO2 load. It is important to note that the buffer mechanism described remains entirely in the ‘buffer base dominion,’ this means that the total amount of dissociated buffers does not change as the free H+ deriving from CO2 load hydration are buffered by A, forming AH, while CO2 becomes HCO3. In vivo, this immediate buffering mechanism produces ≈ 1 mmol/L HCO3 increase every ≈10 mmHg of increased PCO2.

Kidneys response (days)

The normal kidneys compensate the low pH by increasing plasma strong ion difference (SID). To perform this variation the kidneys must excrete the appropriate amount of strong ions. In the urine the electrolyte equilibrium can be written as:

Na +   +   K + +   UI + =   Cl
[eqn 6]

where UI+ and UI are the unmeasured positive or negative strong urinary ions. The most important unmeasured positive strong ion is NH4+, the most important unmeasured negative ions is SO42–, which derives from sulphur acid metabolism. To manipulate the difference of strong ions that are relevant in generating the plasma SID (and pH) it follows that:

Na + + K + +   CI = UI   UI +
[eqn 7]

To increase plasma SID (traditionally called compensatory metabolic alkalosis), urinary SID must decrease, and this occurs by increasing Cl excretion. To maintain electroneutrality, the difference UI – UI+ must decrease and this occurs by rising NH4+ excretion, which is the best way to excrete an increased amount of Cl while retaining Na+ [2,3]. Therefore, immediately after the PCO2 rise and the pH decrease, the kidneys start to increase the ammonium production, thus increasing the chloride excretion. Indeed, during respiratory acidosis, plasma chloride decreases (increasing plasma SID), while urine chloride and the ammonium increase (with associated reduction in urinary pH). It takes approximately 2–3 days to excrete a sufficient amount of chloride in order to significantly affect the plasma SID. When plasma SID is increased the buffer base (HCO3 + A) increases equally. In vivo this process leads to an increase of ≈3.5 mmol/L of HCO3 every ≈10 mmHg of PCO2 chronically increased.

Acid–base regulation in critically-ill patients

Every step of acid–base regulation may be affected in critically-ill patients:

  • The acid load may be altered by both the underlying disease (hypoperfusion, shock, increased metabolism, etc.) and therapeutic intervention (excessive parenteral feeding).

  • The buffer system is frequently abnormal in critically-ill patients; decreased levels of albumin, the most important component of the A–AH buffer pair, are not unusual. A decreased concentration of both A–AH and HCO3–CO2 buffer pairs implies a greater change in pH for a given acid load.

  • Transport of blood from the venous to the arterial side may also be affected (low-flow states). In these conditions, the CO2 clearance may be altered, resulting in a large difference between the acid–base status of the venous and arterial blood, with increases in the PCO2 and pH gradients.

  • The physiological response to increased PCO2 and decreased pH, i.e. increased ventilation, is usually impaired in critically-ill patients for two main reasons. The underlying disease may affect the lung, thus preventing increased ventilation (due to reduced minute ventilation or increased dead space), and these patients are usually on mechanical ventilation. If volume-controlled ventilation is used, PCO2 will rise as the patient (usually sedated and sometimes paralyzed) cannot change either the tidal volume or the respiratory rate. If pressure-support ventilation is used, the patient may respond to the increase in PCO2 by triggering mechanical breathing more frequently. However, the depth of breath (i.e. the tidal volume) is often out of the patient’s control. Thus it is obvious that mechanical ventilation has a deleterious effect on the physiological response to changes in acid–base status.

  • The physiological response to decreased pH in the renal tubular cells, in which HCO3 losses are replaced, may also be impaired in the critically ill, whose renal function is often affected. It is also important to realize that some forms of renal support, as haemofiltration, may lead to HCO3 losses, with consequent additional derangement of the acid–base balance.

To summarize, it is important to remember that the physiological control of the acid–base equilibrium is often impaired in critically-ill patients, and physicians must understand which mechanisms are altered so that an adequate substitute can be provided for the physiological control which has been lost.

Therapeutic strategies

Approach to respiratory acidosis

In patients presenting with hypercapnia, acidosis, and hypoxaemia, two lines of action are required:

  • Identification of the causes of respiratory acidosis, with particular focus on the correction of precipitating factors which may be reversible.

  • Treatment of the symptoms and signs if they are themselves a possible cause of unfavourable outcome.

These two actions should be pursued together, as correction of the precipitating factors may lead to almost immediate resolution of the respiratory acidosis. However, the first goal in intensive care is the maintenance of homeostasis, and correction of life-threatening conditions is the priority. Thus, the indications for symptomatic treatment will be discussed first.

Hypoxaemia, hypercapnia, and acidosis


Hypoxemia unavoidably occurs during CO2 retention when the patient breathes room air (FiO2 = 21%) since:

F A O 2   =   F I O 2   F A C O 2
[eqn 8]

where FaO2 is the alveolar fraction of oxygen, FiO2 is the inspired fraction of oxygen, and FaCO2 is the alveolar fraction of CO2. This type of hypoxaemia can easily be corrected by increasing FiO2. For example, when PaCO2 = 80 mmHg (10.66 kPa) (FaCO2 = 11.2%), FaO2 can be restored to its normal values by increasing FiO2 from 21 to 26.6%. Therefore, the hypoxaemia due to hypercapnia is easily corrected by increasing the inspired oxygen fraction, differently from the hypoxaemia due to the right to left shunt. The pathophysiological meaning of different PO2/PCO2 combinations are listed in Table 113.2 (conditions of high altitude or inhalation of hypoxic gas mixtures are excluded).

Table 113.2 Relationship between PO2 and PCO2

Normal PCO2

Low PCO2

High PCO2

Normal PO2


Pure hyperventilation if FiO2 = 21% (stress + anxiety, metabolic acidosis); if FiO2 > 21%, the above conditions plus shunt

Ventilatory impairment while inhaling FiO2 > 21%

Low PO2

Oxygenation impairment due to shunt plus relative ventilatory impairment

Oxygenation impairment due to shunt with physiological ventilatory response

Ventilatory impairment if FiO2 = 21%; oxygenation impairment due to shunt if FiO2 > 21% associated with ventilatory impairment

High PO2

FiO2 > 21% without ventilatory impairment

Hyperventilation during inhalation of FiO2 > 21%

Ventilatory impairment while inhaling FiO2 > 21%

Hypercapnia and acidosis

Similarly to PO2, there is no threshold value of PCO2 or pH which is ‘harmful’ per se. Many factors may influence the response to increased CO2, such as the rate of increase in PCO2 (acute or chronic), age, and cardiovascular conditions. If associated with normoxia, near-normal pH, consciousness, and haemodynamic stability, a high PCO2 does not need any therapeutic intervention. The indications for mechanical assistance should be based on a global clinical assessment, considering the three main consequences of increased PCO2, i.e. tissue acidosis, impairment of the central nervous system, and the cardiovascular response [4]‌.

Tissue acidosis

As molecular CO2 enters the cell membrane faster than HCO3, it is generally believed that intracellular pH decreases more than extracellular pH. However, there is increasing evidence that the intracellular buffers limit tissue acidosis and hypercapnia is well tolerated. If cellular acidosis develops, cell function and viability are impaired.

Effects on the central nervous system

Increasing PCO2 may have a severe effect on central nervous system activity. Experimentally, the brain excitability first decreases, then increases, with associated seizures, and finally decreases to anaesthesia and coma. CO2 is one of the major determinants of the cerebral vascular reactivity, both directly and indirectly (through pH changes), and acute hypercapnia may result in an increased cerebral blood flow and intracranial pressure.

Circulatory response to hypercapnia

The effect of hypercapnia on the cardiovascular system depends on the balance between the direct depressant effects of PCO2 on heart and peripheral vascular smooth muscles, and the increased plasma levels of epinephrine and norepinephrine due to activation of the sympathetic nervous system. In normal conditions, the net result is an increase in cardiac output and a slight decrease in peripheral resistance. The arterial pressure tends to rise and the pulmonary artery pressure may increase substantially. It is important to remember that these reactions are observed in intact subjects. In patients given β‎-blockers, for example, hypotension and decreased cardiac output may be observed.

Causes of respiratory acidosis and precipitating factors

The most common causes of respiratory acidosis and the time required for their correction are summarized in Table 113.3. They can be classified into three groups:

  • In this group the cause of hypercapnia can be removed easily. If hypoxaemia can be corrected by supplemental oxygen administration, it is better, after removal of the precipitating factors, to wait for a spontaneous increase in alveolar ventilation. Hypercapnia does not require any treatment if associated with a stable pH, high HCO3, and haemodynamic stability in a conscious patient.

  • In this group, the correction of the precipitating factors will probably require hours or days. The need for mechanical ventilation should be determined on the basis of a global clinical assessment. In patients in whom hypercapnia is associated with clinical signs of severely increased work of the respiratory muscles, mechanical support should be introduced before the development of respiratory fatigue, which may lead to a sudden deterioration of PCO2 and pH.

  • In the final group the cause cannot be corrected (e.g. late-stage neuromuscular disease). In most cases, the issue is more ethical than medical, and the therapeutic plan should be discussed with the patients and relatives (e.g. planning for home ventilation).

Table 113.3 Causes of respiratory acidosis

Immediate reversibility

Reversibility within hours/days


Respiratory drive

  • Drugs (antidotes available)

  • Nutritional insufficiency

  • Chronic loading

  • Metabolic alkalosis

  • Endocrine disturbances



  • Secretions

  • Foreign bodies

  • Bronchospasm

  • Airways apparatus

  • Asthma

  • Bronchial stenosis

Terminal COPD


Drugs (antidotes available)

  • Neuromyopathies

  • Endocrine/electrolyte disorders

  • Abdominal distension

  • Hyperinflation

  • Quadriplegia

  • Terminal neuromuscular

  • Disease

Chest wall

Flail chest

  • Kyphoscoliosis

  • Thoracoplasty

Lung parenchyma

  • Pneumothorax

  • Pleural effusion

  • Pulmonary oedema

  • Pulmonary embolism

Terminal obstructive and restrictive lung disease

Assessment and correction of hypercapnia

Knowledge of the previous respiratory status is of paramount importance in determining the goal of respiratory support. A reasonable goal in a previous healthy subject could be normal blood gases (PaCO2 = 40 mmHg (5.33 kPa), PaO2 > 80 mmHg (10.66 kPa)). In a subject with previous chronic respiratory impairment, the goal should be the blood gas values present before the superimposed acute derangement (e.g. PaO2 = 60 mmHg (8 kPa), PaCO2 = 50 mmHg (6.66 kPa)). An accurate history is essential for differentiating chronic and acute respiratory impairment (i.e. to define the target blood gases). Measurement of the blood gases alone may be misleading. In fact, the normal increase in HCO3 of ≈1 mmol/L (acute conditions) or ≈3.5 mmol/L (chronic condition) for every increase of 10 mmHg (1.33 kPa) in PaCO2 may be offset by the concurrent presence of metabolic acidosis, which ‘consumes’ the HCO3. Under these conditions the HCO3 level does not discriminate between acute and chronic respiratory acidosis.

In typical chronic respiratory acidosis, hypoxaemia is usually corrected by increasing FiO2. The risk of high PaO2 in chronic patients breathing spontaneously (coma) has probably been overestimated [5]‌, and reasonable oxygenation is a mandatory target.

Since PaCO2 ∝ VCO2/Va, it is evident that it can be decreased by either decreasing VCO2 or increasing the alveolar ventilation Va. Methods of decreasing VCO2 include withdrawal of the excessive load of glucides delivered by parenteral nutrition, control of temperature in a hyperthermic patient, and artificial removal of part of VCO2 by extracorporeal methods such as dialysis or artificial lungs. However, the usual way of correcting hypercapnia is to increase alveolar ventilation, and this is usually achieved by mechanical ventilation. When the causes of hypercapnia are extrapulmonary, such as central nervous system or neuromuscular diseases, and the lung parenchyma is normal, mechanical ventilation causes no more problems than during general anaesthesia in a normal patient.

Major problems may occur in patients in whom hypercapnia is due to dysfunction of the small airways (e.g. bronchospasm or asthma), parenchymal lesions (e.g. emphysema), or diseases involving both airways and parenchyma (e.g. severe chronic obstructive pulmonary disease (COPD)). If mechanical support of these patients is unavoidable, it is essential to avoid worsening hyperinflation, which may lead to devastating haemodynamic consequences including cardiac tamponade. Small tidal volumes and prolonged expiratory time must be maintained when ventilating these patients, even at the cost of a relatively high PCO2. The advantages of this setting have been shown in asthmatic patients [6]‌ and suggested in other patient populations [7]. As a general rule, non-invasive ventilation is preferable to mechanical ventilation with intubation in the hypercapnic patient [8]. If the patient is hypercapnic, hyperinflated, and performing excessive respiratory work, the use of continuous positive airway pressure may decrease respiratory work and enable the patient to maintain spontaneous breathing. It must be noted, however, that when mechanical ventilation is per se harmful to correct hypercapnia, the artificial lungs are the appropriate solution. This form of support is increasingly used worldwide as it allows to remove varying amounts of VCO2, up to removal of all the metabolically-produced CO2, therefore abolishing the need for mechanical ventilation [9,10]. The main indications are, to date, bridge to lung transplant, acute respiratory distress syndrome and COPD re-exacerbation.


1. Covelli HD, Black JW, Olsen MS, and Beekman JF (1981). Respiratory failure precipitated by high carbohydrate loads. Annals of Internal Medicine, 95(5), 579–81.Find this resource:

2. Kellum JA. (2000). Determinants of blood pH in health and disease. Critical Care, 4(1), 6–14.Find this resource:

3. Ring T, Frische S, and Nielsen S. (2005). Clinical review: renal tubular acidosis—a physicochemical approach. Critical Care, 9(6), 573–80.Find this resource:

4. Bidani A, Tzouanakis AE, Cardenas VJ, Jr, and Zwischenberger JB. (1994). Permissive hypercapnia in acute respiratory failure. Journal of the American Medical Association, 272(12), 957–62.Find this resource:

5. Aubier M and Dombret MC. (1993). Acute exacerbation of chronic airflow obstruction. In: Pinsky MR and Dhainaut JFA. (eds) Pathophysiology: Foundations of Critical Care, pp. 427–46. Baltimore, MD: Williams & Wilkins.Find this resource:

6. Darioli R and Perret C. (1984). Mechanical controlled hypoventilation in status asthmaticus. American Reviews in Respiration Diseases, 129(3), 385–7.Find this resource:

7. Hickling KG, Henderson SJ, and Jackson R. (1990). Low mortality associated with low volume pressure limited ventilation with permissive hypercapnia in severe adult respiratory distress syndrome. Intensive Care Medicine, 16(6), 372–7.Find this resource:

8. Brochard L, Mancebo J, Wysocki M, et al. (1995). Noninvasive ventilation for acute exacerbations of chronic obstructive pulmonary disease. New England Journal of Medicine, 333(13), 817–22.Find this resource:

9. Gattinoni L, Carlesso E, and Langer T (2011). Clinical review: Extracorporeal membrane oxygenation. Critical Care, 15(6), 243.Find this resource:

10. Gattinoni L, Carlesso E, and Langer T. (2012). Towards ultraprotective mechanical ventilation. Current Opinions in Anaesthesiology, 25(2), 141–7.Find this resource: