Approach to hypercapneic respiratory failure

Several past paper SAQs demand the candidate either generate a list of differentials or define a sensible approach to the assessment  and management of acute hypercapneic respiratory failure.

• Question 16 from the first  paper of 2014 (generate differentials)
• Question 14 from the second paper of 2011 and Question 1a from the second paper of 2000 are virually identical, and discuss a non-specific "approach" to the management of  a comatose fat man with a PaCO₂ in excess of 70mmHg.

Also related to this topic are past paper questions on the ventilation strategies in COPD, which include the following SAQs:

• Question 29 from the second paper of 2005
• Question 30 from the second paper of 2005
• Questions 1a, 1b, 1c  and 1d from the first paper of 2001 

Causes of Hypercapnia

Decreased minute ventilation	Central nervous system Drugs affecting respiratory drive, eg. opiates Brainstem or cortical lesion affecting consciousness Central sleep apnoe Spinal cord injury Neuromuscular Neuropathy, eg. Guillain-Barre NMJ disorder, eg. myasthenia gravis Myopathy Respiratory Decreased lung compliance, eg. pulmonary oedema Decreased chest wall compliance, eg. kyphosis or obesity Increased airway resistance, eg. COPD or asthma Metabolic, endocrine and environmental Metabolic alkalosis Hypothyroidism Hypothermia
Increased dead space	Increased anatomical dead space Unusually long ventilator circuit (eg. while down in MRI) Increased alveolar dead space (i.e. ventilated but not perfused) Bullous emphysema, COPD Interstitial pulmonary fibrosis Large pulmonary embolism
Increased CO2 production	Increased metabolic rate Hyperthermia (including malignant hyperthermia) Hyperalimentation Hyperthyroidism Seizures, status epilepticus Muscle exertion

Features of clinical examination that assist in making a diagnosis:

Observation:

• Obesity
• Short fat neck of OSA
• Cushingoid appearance (OSA, but also suspicious of long term steroids for some sort of autoimmune condition, or COPD)
• Wasting and cachexia of severe CCF, end-stage COPD or cancer
• Abnormal breathig pattern (eg. the abdominal breathing of a C-spine quad)

Start with the hands.

• Clubbing (suggestive of chronicity)
• Cyanosis
• Unilateral small muscle wasting (lung mass invading brachial plexus)
• Pulse (collapsing pulse of AR?)

Axillae and neck

• Lymph nodes
• JVP (cardiac causes of ventilation failure)
• Dissection scars from lymph node clearance; radiotherapy tattoos

Face and cranial nerves

• Plethoric "mitral facies"
• Droop, cranial nerve signs of stroke
• Horner's syndrome (malignancy or stroke)
• Temporalis wasting (malnutrition)

Chest

• Abnormal chest wall movement (eg. flail segment or unilateral phrenic nerve paralysis)
• Subcutaneous emphysema on palpation, suggestive of pneumothorax
• Percussion findings (eg. dullness of an effusion)
• Auscultation findings of wheeze or creps (spasm or APO)

Abdomen

• Recent abdominal wounds (is pain or infection preventing diaphragm excursion?)
• Distension (Gas? Poop? Ascites?)

Lower limbs

• Oedema of CCF or prolonged bed stay
• Muscle wasting of quads (another feature of malnutrition)

It would be nice to ignore the comatose fat man, but two SAQs have been repeated involving him. The college wanted an "approach" to his problem. The approach listed below is directly copied from those SAQs, with no modification.

1. - Assess airway patency;
2. - if airway is not protected, introduce simple airway devices and assess their effect on airway patency
3. - if airway reflexes are intact, commence NIV paying attention to the dangers of the high pressures which will be required in an obese patient
4. - if airway reflexes are intact, assess for intubation and solicit expert help for intubation.
5. - assess any rapidly reversible causes of obtundation such as opioid intoxication and hypo/hyperglycaemia
6. - if no rapidly reversible cause is found, mechanical ventilation must commence while the process of investigation continues
7. - once airway patency is established and some form of mechanical ventilation is in progress, other causes for the reduced level of consciousness must be pursued and managed, including intracranial causes, thromboembolism, electrolyte abnormalities, cardiac failure, hypothyroidism etc.
8. - at the same time, management of the possible causes of hypercapneic respiratory failure must commence (therapies specifically directed at COPD and OSA)
9. - at the same time, standard management protocols for the care of an obese ICU patient must be followed, including thromboprophylaxis, pressure area care, the use of specialised bariatric equipment, and ulcer prophylaxis.

Causes of unusually decreased CO₂ elimination

Leaving aside weird amphibian stuff and ECMO, it is fair to say that ventilation is the main mechanism of removing CO₂, and the failure of ventilation is the main mechanism of hypercapnia. The reader preparing for CICM exams should therefore be prepared to readily assess the cause of hypercapnia using some kind of cognitive schema that covers all the possible causes. Of these, there are many, and none is superior to others, except insofar as its coverage of the usual suspects, which are:

• Failure of centrasl respiratory drive
• Failure of some effector between the CNS and the muscle
• Failure of the muscle or the mechanical coupling between the muscle and ventilation (eg. something being fundamentally wrong with the chest wall)
• Insufficiency of blood flow to the gas exchange surface
• Insufficiency of the gas exchange surface itself (rarely a diffusion issue, more commonly a surface area problem)
• Impediment to exchange from an increase in the CO₂ content of the  ambient gas mixture

There are literally one million methods to represent this for the user to remember and consolidate. What follows is an application of a tired old alphabetic system: 

• A - obstructed airway
  • Upper:
    • Depressed consciousness
    • Obstructive sleep apnoea
    • Laryngeal oedema, laryngospasm or angioedema
    • Tumour, haematoma or foreign body
    • Obstruction of a tube, eg. ETT or tracheostomy
  • Lower:
    • Bronchospasm (asthma, COPD)
    • Excessive secretions
  • Increased anatomical or apparatus dead space:
    • Excessive connectors, catheter mounts or HMEs between the patient and the circuit separation point
    • Masks, snorkels, MRI extension tubing, etc
• B - lung pathology:
  • Increased physiological dead space (eg. PE, emphysema)
  • Rebreathing or increased inspired CO₂, eg.  failure of CO₂ absorption in closed-circuit breathing apparatus (Deng et al, 2015)
  • Terrible choice of mechanical ventilation settings, eg. inappropriately low respiratory rate or tidal volume
  • Excessive circuit leak
  • Decreased effective minute ventilation
  • Decreased compliance
  • Respiratory muscle fatigue
• C – cardiovascular and circulatory
  • Reduced pulmonary blood flow, producing alveolar dead space, eg.  severe shock, cardiac arrest
  • Washout of accumulated tissue CO₂ following ROSC
• D – neurological and neuromuscular
  • Failure of central respiratory drive (opioids, TBI, post ictal,  etc)
  • Failure of descending motor pathways (eg. high cervical spinal cord injury, demyelinating disease, etc)
  • Peripheral nerve or NMJ disease, eg. Guillain–Barré syndrome, phrenic nerve palsy, myasthenia gravis, botulism
• E – electrolyte, acid–base and endocrine
  • Respiratory muscle weakness from severely low phosphate,  hypokalaemia, etc
  • Metabolic alkalosis, which reduces ventilatory drive and produces compensatory hypoventilation
  • Administration of bicarbonate or other buffer, increasing the CO₂ load
• F – renal
  • Fluid overload, pulmonary oedema
  • CO₂ delivery during bicarbonate haemodialysis or CRRT, particularly when ventilatory reserve is limited; transfer of CO₂ from dialysate into blood has been demonstrated directly (Sombolos et al, 2005)
• G – gastrointestinal, abdominal and hepatic
  • Impaired diaphragmatic excursion, eg. obesity, abdominal compartment syndrome, ascites, pregnancy
  • Pneumoperitoneum (systemic absorption of insufflated CO₂)

Causes of unusually increased CO₂ production

Question 24.1 from the first paper of 2022 presented the candidates with a PaCO₂ of 192 mmHg, in the context of an unconscious post-arrest patient ventilated via an LMA on the way to the ED. Could a tremendous increase in CO₂ production have contributed to the hypercapnia, a reader asked; and would such an answer get marks? These might include heat stroke,  cocaine toxicity,  malignant hyperthermia and status epilepticus (status epilepticus presumably of the fully convulsive variety, to maximise muscle activity). To these, one might have also added thyroid storm and hypermetabolic states in sepsis trauma and burns. 

There are several sources for confusion here. CO₂ is generated by the tissues during aerobic metabolism,  and so the appearance of CO₂ in the blood is conditional on the perseverance of several normal physiological mechanisms, those being:

• the delivery of oxygen to the tissues
• the extraction of metabolic wastes from the tissues
• the ability of the tissues to transform metabolic fuel into CO₂ and water

As such, CO₂ may be:

• Not produced in the tissues because the metabolism there is not aerobic (eg. during an exceptionally prolonged seizure) 
• Produced in the tissues, but stored there and not carried out because of poor circulatory function (eg. during a prolonged resuscitation of cardiac arrest)

However, we can dismiss some of these concerns in the service of greater amusement. Consider: any metabolic secrets created during the arrest are revealed at ROSC, undesupplied metabolism still yields CO₂ by eating into myoglobin stores, and post arrest metabolism may actually be more vigorous because of post-arrest inflammatory cascades. In short, we should now seriously engage in a discussion of the causes of hypercapnia that could be atttributed to increased production, even in the face of a normal or increased ventilation.

What could produce this sort of situation? In essence, the metabolic rate needs to be increased to a level beyond the normal, and the usual mechanisms for coping with this need to be hindered in some serious way, so that they cannot compensate normally, permitting the CO2 to accumulate. This can happen in any number of situations that can all be broadly described as "hypermetabolic thing plus mandatory mechanical ventilation". The aforementioned heat stroke,  sympathomimetic toxidrome,  malignant hyperthermia and status epilepticus could definitely fit the description. In brief:

• Sympathomimetics can increase CO₂ production, but not massively, and usually by means of increasing the body temperature. For example, Amoroso et al (1993) were able to slightly raise the VO₂ of asthmatics with salbutamol (from ~180-190 ml/min to ~220 ml/min with twelve puffs of an MDI), and they remained euthermic, whereas Freedman et al (2005) were able to increase the VO₂ of healthy volunteers from ~200 to ~250 ml/min using 2mg/kg  of MDMA along with a ~1 degree increase in body temperature. 
• Status epilepticus can produce increased VCO₂ by increasing skeletal muscle VO₂ as well as cerebral metabolism, and apparently by quite a significant amount, though much of the work in this space is present in the form of case reports and short letters to editors. Literally the only study dredged up by a search on this subject turned out to be a letter that could not even be upgraded to a case report, including mobile phone snaps of the monitoring screens, where a spike in VCO₂ can be seen. In general, it seems like seizures could produce vast amounts of CO₂ if they were sustained, but often the natural destination for status epilepticus is to become nonconvulsive (so called "electroclinical dissociation"). 
• Hyperthyroidism increases VCO₂ chronically, but again not by a dramatic amount - Jansson et al (2001) measured the pre and post β-blocker VCO₂ and reported a modest drop from ~210-220 to ~170-180 ml/min
• Sepsis and trauma increase oxygen consumption and VCO₂ production; Voerman et al (1993) reported a VCO₂ of around 280ml/min in septic patients, and referred to even older studies suggesting that something similar is seen with trauma and burns. Dickerson et al (2002). directly reports around 300-400 mL/min at about post-burn day 8 in 24 adults with ~20% TBSA burns, which could represent a doubling of their metabolic rate.
• CO₂ excess during hyperthermia (eg. heat stroke) is mostly known from measurements collected from passively heated volunteers.  Saxton (1981) raised the temperature of resting men by 2°C and found an 18% rise in V̇CO₂, and  Pranskunas et al (2015) passively heated healthy volunteers to a rectal temperature of 39.5°C, which increased their V̇CO₂  from 175 ml/min to around 300. In 12 febrile mechanically ventilated patients, Manthous et al (1995) also measured a V̇CO₂ around 300ml/min at 39.4 and something closer to 240 mL/min after cooling to 37.0°C. 

So, it is plausible that PaCO₂ might increase in these people, and that they could develop hypercapnia if their minute volume was for some reason restricted to some normal value, eg. where they are ventilated mechanically with a mandatory mode set for a normal set of circumstances. And it should be possible to estimate the rate of increase mathematically, using the ideal gas law. Any individual at a steady state has a minute volume which maintains a stable PaCO₂, proportional to their rate of production, which for a normal RQ of 0.8 ends up being 200ml/min. If the rate of production increases but the minute volume remains the same, it logically follows that 200ml/min will remain the stable elimination rate, and the minutely accumulation of CO₂ will occur at a rate of x - 200ml/min, where x is the increased rate of production. For x = 300 ml/min, like the hyperthermic patients from Manthous et al and  Pranskunas et al, this means an accumulation of an extra 100ml of CO₂ per minute. For gas volumes expressed at standard temperature and pressure, 100 mL occupies 0.100 L and corresponds to:

Applying the ideal gas law, 
 

where the gas is distributed into a fixed notional distribution volume V (say, 42 litres, the total body water)  at constant temperature T (body temperature in Kelvins, 310.15°K)  gives a pressure change P of:

Multiplying by the net molar accumulation rate, dn/dt = 0.00446 mol/min, gives:

or,  given that 1 kPa = 7.5006 mmHg,  

In short, if the patient's rate of metabolic CO2 production increases by 50% because they are hyperthermic to ~39.5º C and one completely ignores their ventilator, their PaCO₂ will increase by something like 2mmHg per minute. Given that normally the rate of increase is said to be something like 3mmHg/min, this is substantial.

But forget the gentle increases in metabolism seen in hyperthyroidism and sepsis; the readership of Deranged Physiology would surely want to know: what is the maximum metabolic CO2 production that anyone has seen in the wild human? Some tremendous values are reported in the literature, mostly from field reports issued by traumatised anaesthetists following malignant hyperthermia crises, and this makes sense if the only way one can think of truly exploding one's metabolic activity is through some mad and mindless metabolic cycles. For example, in a cursed species of hyperthermia-susceptible pigs, Gronert et al (1977) found a maximum rise in whole-body oxygen consumption of approximately 6 mL O₂/min/kg, suggesting something in the order of 400-500 ml/min for an equivalent sized human.  Liebenschütz, Mai, and Pickerodt (1979) calculated an index from respiratory minute volume and PaCO₂ in two human malignant hyperthermic crises, where assuming a dead-space fraction of 40%, the highest CO₂ output in one patient was estimated to be of the order of 1,000 mL/min.  More recently, Lin et al (2014) described a crisis in which 10 L/min minute ventilation coexisted with PaCO₂ 107.7 mmHg and end-tidal CO₂ above the capnograph's 99-mmHg limit, which suggests that double the usual rate of ventilation still yielded a 2.5-fold increase in the PaCO₂.  (i.e. the increase in production was theoretically 10-fold).  Even a relatively modest 800ml/min rate of CO₂ accumulation amounts to

or an increase from 40 mmHg PaCO₂ to something around 200 mmHg over 20 minutes.

But these are pathological states. If this rate of increase seems insane, it is only because the reader has not considered the possibility of measuring a VCO2 from a healthy volunteer pushed to their physiological exercise limit. Kristian Blummenfelt the Norwegian triathlete and 2020 Olympic gold medallist famously clocked in a V̇O₂ of 101.1 mL/kg/min, which he immediately shared on Instagram to flaunt his gallantry in the face of teeming hordes of sports physiologists, snarling their opprobrium on his feed. If this 74kg athlete was indeed consuming ~100ml/kg/min of oxygen, then at a usual RQ of 0.8 the total V̇CO₂ would have to be 80 ml/kg/min or 5920 ml (0.2641 mol)/ min, giving:

Or, a rate of accumulation of CO₂ that is approximately 15 times faster than those anaesthetised patients with malignant hyperthermia.