Goals of Ventilation and Choosing a Mode

Educational reference — a way to reason about what a ventilator mode is for and how to choose among modes. It is not a bedside directive, a substitute for clinical judgment, or a device’s operator’s manual.

Start with the question, not the dial

A ventilator has a dizzying array of modes, and the temptation is to learn them the way you learn a phone menu — memorize which button does what on which machine. Mireles-Cabodevila and Chatburn argue this gets the problem exactly backward. The right first question is not which mode? but what am I trying to achieve for this patient right now? Their 2013 “rational framework” reframes mode selection as a mapping problem: the patient sits between what the research knows and what the industry sells, and the clinician’s job is to relate the diagnosis to the treatment — to match a mode’s technological capabilities to the patient’s current goal [Mireles-Cabodevila 2013].

That framework rests on a claim that is almost startling in its economy: every indication for mechanical ventilation reduces to just three goals [Mireles-Cabodevila 2013].

1. Safety — provide gas exchange without doing harm (primum non nocere).
2. Comfort — keep the patient in sync with the machine.
3. Liberation — get the patient off the ventilator.

Everything else in this chapter hangs off those three words. First we make each one concrete; then we turn them into a way of choosing.

The three goals, operationalized

flowchart LR
  G1["Safety"] --> C["Match mode<br/>capabilities to the goal"]
  G2["Comfort"] --> C
  G3["Liberation"] --> C
  C --> Sel["Choose the mode by its tag,<br/>not its brand name"]

Figure 1 — goal-driven mode selection: match a mode’s capabilities to the goal, then name it by its tag rather than its brand name. Adapted from [Mireles-Cabodevila 2013].

The strength of the framework is that it does not leave the goals as slogans. The 2013 paper’s Table 4 unfolds each goal into a chain — objectives → clinical aims → ventilator capabilities → specific mode features — so that a goal becomes a checklist of things a mode can or cannot do [Mireles-Cabodevila 2013].

Safety breaks into two physiologic objectives plus one that is admittedly aspirational [Mireles-Cabodevila 2013]:

• Optimize ventilation–perfusion (V/Q). Maximize alveolar ventilation (an acceptable PaCO₂) and minimize intrapulmonary shunt (an acceptable PaO₂). The capabilities that serve this range from the simplest — a clinician manually setting tidal volume and rate to hit a minute ventilation — up to modes that adjust a CO₂ or minute-ventilation target automatically.
• Optimize the pressure–volume curve. Ventilate on the steep, high-compliance part of the P/V curve, at minimal tidal volume and optimal PEEP, to minimize stress and strain and avoid ventilator-induced lung injury (VILI) — atelectrauma from too little volume, volutrauma from too much. Capabilities here include automatic lung-protective limits and automatic tidal-volume adjustment.
• Optimize alarm settings. Warn of unsafe conditions. The authors flag this as genuinely under-studied — no mode yet has intelligent alarming beyond default thresholds [Mireles-Cabodevila 2013].

Comfort is a single objective — optimize patient–ventilator synchrony, because asynchrony is common and is associated with discomfort and longer ICU and hospital stays [Mireles-Cabodevila 2013]. Its aims: maximize trigger and cycle synchrony (maximal when all breaths are spontaneous, as in pressure support, PAV, or NAVA), maximize flow synchrony (favoring pressure-control modes with unrestricted flow), minimize autoPEEP, and — the hardest — coordinate the ventilator’s work output with the patient’s demand. Truly matching work to demand requires the ventilator’s inspiratory pressure to be proportional to the patient’s effort, something achieved only by proportional assist ventilation and NAVA [Mireles-Cabodevila 2013]. Reading that coordination at the bedside is the domain of Bedside Monitoring of Respiratory Mechanics.

Liberation is a prime goal from the moment ventilation begins, because every extra day carries cost and risk. Its objective is to optimize the weaning experience, with aims to minimize both the duration of ventilation and adverse events [Mireles-Cabodevila 2013]. The framework distinguishes patient-driven support reduction (adaptive pressure-targeting modes that lower pressure as effort rises — “automatic weaning modes”) from ventilator-driven reduction (a mode that intermittently drops support, evaluates the response, and even recommends separation) [Mireles-Cabodevila 2013].

A crucial caveat travels with all three: they are not equally weighted at all times. In acute respiratory failure, safety dominates and liberation is irrelevant; as the patient stabilizes, comfort matters more; and safety versus comfort can genuinely conflict [Mireles-Cabodevila 2013]. The three goals are a frame for deciding which one matters most now, not a fixed ranking.

Mode selection as deduction, not tradition

Once the goals are concrete, the selection logic follows. The framework’s paradigm reasons deductively, from goals down to features: a mode has design features that implement general capabilities; each capability serves a clinical aim; each aim serves an objective and ultimately a goal [Mireles-Cabodevila 2013]. Because each capability is, by definition, beneficial in the scenario it is meant for, the framework’s working rule is that — other things equal — the mode with more of the capabilities that serve your chosen goal is the better tool for that goal [Mireles-Cabodevila 2013].

Why deduce from first principles rather than simply consult the trial evidence? Because the trial evidence effectively cannot exist. The authors do the arithmetic: head-to-head randomized comparison of the ~22 unique modes on four common ICU ventilators would require 231 comparisons, which — scaled from the ARDS Network’s cost of roughly four years and $38 million per study — works out to more than 900 labor-years and nearly $9 billion [Mireles-Cabodevila 2013]. So a mode can be evaluated at three levels — theoretical (first principles), performance (how a given device actually behaves), and clinical outcome (survival or physiology in a population) — but only the theoretical level is practically achievable for comparing modes as a class [Mireles-Cabodevila 2013]. The framework therefore deliberately compares “the tools in the toolbox” independent of who wields them, ignoring operator skill and assuming each mode runs under best-case conditions where its design assumptions are not violated [Mireles-Cabodevila 2013].

This yields concrete, if illustrative, best-in-class answers by goal [Mireles-Cabodevila 2013]:

• Safety — a mode that both assures ventilation and applies expert-rule lung protection, e.g., IntelliVent-ASV (a PC-IMV mode with optimal/intelligent targeting).
• Comfort — a mode that lets every breath be spontaneous while unloading the muscles in proportion to effort, e.g., proportional assist ventilation or NAVA (PC-CSV with servo targeting).
• Liberation — a mode that reduces support and tests readiness, e.g., SmartCare/PS (PC-CSV with intelligent targeting).

The tags in parentheses — PC-IMV, PC-CSV — are not decoration. They are the device by which this whole comparison becomes possible, and that is the next piece of the story.

The proliferation problem, and why names must become tags

There is a reason mode selection feels overwhelming, and it is not the clinician’s. Recent years have seen a dizzying proliferation of modes, driven by technological advances and market pressures rather than clinical data [Mireles-Cabodevila 2013]. The numbers are concrete: one respiratory-care equipment textbook lists 174 unique mode names, and four common ICU ventilators alone carry 47 unique names [Mireles-Cabodevila 2013]. The 2014 maxims paper makes the same point at larger scale — 290 unique mode names across 30 ventilators, with one textbook listing 174 names across 34 machines [Chatburn 2014].

Two things make this unmanageable. First, the names obscure rather than reveal: different brand names often describe the same underlying behavior, and a name frequently does not describe the behavior at all [Mireles-Cabodevila 2013; Chatburn 2014]. Second — and this is the framework’s most human observation — working memory holds only about four variables at once [Mireles-Cabodevila 2013]. Several hundred proprietary names simply exceed what a clinician can hold, compare, and reason over. No amount of study fixes a representation that is the wrong size for the mind using it.

The remedy is to stop thinking in names and start thinking in tags — the short, generic classifications produced by the taxonomy for modes of mechanical ventilation, where every mode is named by its control variable, breath sequence, and targeting scheme [Chatburn 2014]. The collapse is dramatic and has been shown repeatedly:

• 2013: across four ICU ventilators, 52 mode names (47 unique) reduce to 17 tags — about 22 unique modes once targeting varieties are counted [Mireles-Cabodevila 2013].
• 2014: a catalog of 290 names across 30 ventilators reduces to 45 tags [Chatburn 2014].

Several hundred names become a few dozen tags — a set small enough to fit in the working memory the problem overflowed, and, more importantly, a set you can finally compare. Once “PRVC,” “AutoFlow,” and “VC+” all resolve to the same tag, their redundancy becomes visible and the genuinely distinct modes stand out [Mireles-Cabodevila 2013; Chatburn 2014]. This is why the shift from names to tags is not cosmetic: it is the precondition for matching capabilities to a goal at all. (The taxonomy is a [proposed] standard — its authors advocate it and ECRI adopted it for describing and comparing ventilators — not universally adopted nomenclature; manufacturers still name modes as they please [Chatburn 2014].)

The pointed observation: the popular modes are unsophisticated

Here the framework says something that ought to give any teacher pause. The three most common adult modes worldwide are Volume Assist/Control, Pressure Assist/Control, and Pressure Support, with Volume A/C the oldest and still the most widely used [Mireles-Cabodevila 2013]. Run them through the framework and the verdict is unflattering: these popular modes are technologically unsophisticated, and they are not obviously the safest or the most comfortable [Mireles-Cabodevila 2013]. Pressure A/C and Pressure Support deliver variable flow (good for comfort) but do not assure a minimum minute ventilation (so safety may suffer) [Mireles-Cabodevila 2013]. By a simple capability tally, the sophisticated closed-loop modes — the ASV/IntelliVent family, PAV, NAVA, SmartCare — outscore them on every goal.

So why does Volume A/C endure? The framework’s answer is disciplined and worth holding onto. Its popularity is defensible for one specific reason: of all the capabilities a mode might have, the only one with hard outcome evidence linking it to survival is that lower tidal volume reduces mortality — and Volume A/C lets you set tidal volume directly and simply [Mireles-Cabodevila 2013]. The authors put it bluntly: after decades of research, “the only thing we seem to know is that smaller tidal volumes are better than larger ones” [Mireles-Cabodevila 2013]. Every other capability in the framework is argued at the theoretical level — beneficial by design, not by demonstrated outcome.

This is the framework at its most honest. A more capable mode is the better tool for a goal; that is not the same as the better outcome for a patient. The one place the two coincide with hard evidence is low tidal volume, and a simple, unsophisticated mode delivers it perfectly well.

A gap, marked honestly. Broader outcome evidence — whether any mode (as opposed to a single setting like tidal volume) changes survival — is exactly what this corpus does not, and by its own argument largely cannot, supply [Mireles-Cabodevila 2013]. This wiki is single-school (Mireles-Cabodevila & Chatburn); an outcome trial pitting one mode against another would come from outside that school, and none is cited here. Treat mode-versus-mode outcome claims as an [open] question this framework deliberately leaves at the theoretical level.

A practical, goal-driven way to choose

Assemble the pieces and a workable bedside habit falls out — not a protocol, but a way of thinking [Mireles-Cabodevila 2013; Chatburn 2014]:

1. Diagnose the goal, not just the disease. Ask which of safety, comfort, and liberation matters most for this patient at this moment, remembering that the answer moves as the illness evolves — safety early, comfort as they stabilize, liberation as soon as it is feasible [Mireles-Cabodevila 2013].
2. Translate the mode in front of you into its tag. Do not trust the brand name; resolve the mode to its control variable, breath sequence, and targeting scheme so you actually know what it does — see The Taxonomy of Ventilator Modes [Chatburn 2014].
3. Match capabilities to the chosen goal. Prefer the mode whose capabilities serve that goal — assured ventilation plus lung protection for safety, proportional unloading for comfort, support-reduction-with-testing for liberation [Mireles-Cabodevila 2013].
4. Weigh sophistication against evidence. More capabilities help a goal in theory; only lower tidal volume helps survival by demonstrated outcome. A simple mode a team knows well can beat a sophisticated one it does not [Mireles-Cabodevila 2013].
5. Re-diagnose as the patient changes. Because the goals shift, the best mode shifts with them; selection is a repeated decision, not a one-time setting [Mireles-Cabodevila 2013].

Notice what this is not: a claim that the fanciest mode wins. Whether a capability yields the best outcome depends on the patient’s condition, the timing, and the operator’s skill — the very variables the theoretical comparison deliberately holds aside [Mireles-Cabodevila 2013].

How this chapter connects

• Choosing rests entirely on being able to name what a mode does — the machinery of that is The Taxonomy of Ventilator Modes, which turns brand names into three-part tags.
• The breath-sequence family that most directly trades safety against comfort and liberation is IMV; its variants are treated in The Five Types of Intermittent Mandatory Ventilation.
• Judging whether a chosen mode is actually meeting the comfort goal — synchrony, effort, work — is done at the bedside in Bedside Monitoring of Respiratory Mechanics.
• Teaching this whole way of reasoning to mastery is the mission of Learning Mechanical Ventilation - The SEVA Program.

Open

• Mode-versus-mode outcomes. By the framework’s own arithmetic, head-to-head outcome trials of modes are essentially unattainable, so the comparison lives at the theoretical level; the only capability tied to survival is low tidal volume [Mireles-Cabodevila 2013]. Whether any mode changes mortality is [open] and would require evidence from outside this single-school corpus.
• Capabilities without a goal. The framework notes some capabilities serve no current goal — biologically variable (“noisy”) ventilation improves oxygenation and lowers peak pressure but fits neither safety nor comfort as a direct goal, hinting at a possible future goal of “biocompatibility” [Mireles-Cabodevila 2013]. [open], by the authors’ own admission.
• Intelligent alarms and adverse-event prediction. The framework calls optimal alarm design under-studied and speculates about modes that estimate the probability of VILI, hypoventilation, or asynchrony — but notes no such monitoring research yet exists [Mireles-Cabodevila 2013]. [open].

Sources

Every claim above is drawn from the frozen, human-reviewed primary literature in the machine layer (raw/literature/, one artifact per paper). The short keys used inline resolve to:

• [Mireles-Cabodevila 2013] — Mireles-Cabodevila E, Hatipoğlu U, Chatburn RL. A rational framework for selecting modes of ventilation. Respir Care 2013;58(2):348–366. PMID 22710796.
• [Chatburn 2014] — Chatburn RL, El-Khatib M, Mireles-Cabodevila E. A taxonomy for mechanical ventilation: 10 fundamental maxims. Respir Care 2014;59(11):1747–1763. PMID 25118309.

This chapter is the human-layer synthesis of two machine-layer records — Goals of Mechanical Ventilation and Modes of Mechanical Ventilation — and the frozen source summaries they cite (see the derived_from field).