Relating to “Automating the weaning process with advanced closed-loop systems” by Dr. Burns et al. [1], we agree that clinical investigations are required to asses these systems in various patient populations; but with SmartCare™ there’s one additional concern to worry about.

Computer-driven Protocolized Weaning from Mechanical Ventilation (SmartCare™, Dräger, Lübeck, Germany), a strategy designed for shortening the weaning period, now-a-days is used around the world in several intensive care units and it’s been recently introduced in México. The protocol was found to be efficient in a multicenter randomized trial by Lellouche et al. [2]; however, it was made on cities that not exceed 100 m above the sea level (m.s.l.), and the so-called “comfort zone” may differ strongly in its ideal limits at levels considered of “low”, “medium” or “high” altitude (2,000, 3,000 and 4,000 m.s.l.) [3], so it’s the case of Mexico City (2,300 m.s.l.)

In normal highland subjects, the induced hypobaric-hypoxia produces chronic hyperventilation that lowers plasma bicarbonate concentration, early on primarily by inhibiting the urinary net acid excretion. For each decrease of 1 mmHg in PaCO2, the plasma bicarbonate decrease by 0.41 mmol/L and hydrogen ion concentration decrease by about 0.4 nmol/L [4], until the lost of bicarbonate reach an equilibrium that allows a normal pH. Significant hypocapnia could be seen at 2,000 m.s.l. [5].

Arterial blood samples in highland subjects of all ages allowed the modification of the Siggaar-Andersen Chart for different levels of altitude [3]. For Mexico City the expected PaCO2 is on the range of 32–34 mmHg. Starting at this point, an adequate renal compensation in a normal subject will bring the serum bicarbonate to 19–22 mmol/L. If we put this subject under the computer-driven weaning protocol which tolerates a PetCO2 of 55 mmHg, or its equivalent PaCO2 of 59 mmHg, it will allow pH values of about 7.15 before alerting the physician with the “Insufficient Ventilation” alarm (if the patient keeps its respiratory frequency between 15 and 30 per minute). The “Tachypnea”, “Persistent Tachypnea” or “Severe Tachypnea” alarms will alert if the response to hypercapnic stimuli is present with a rise in respiratory frequency above 30 per minute.

Anyone of these hypothetic scenarios would imply risk, with a delayed software response on raising the pressure support level, a delay on the alert that calls to check the patient’s clinical condition or allowing levels of respiratory acidosis that could reverberate in other systems. The computer-driven protocol seems not too safe in highlanders, and the altitude level considered “critic” will depend on the change on acid-base balance it causes. That’s why the “comfort zone” limits should be adjusted according to the altitude-expected PaCO2 levels.