FUNDAMENTALS OF MECHANICAL VENTILATION

Published on 20/03/2015 by admin

Last modified 22/04/2025

Print this page

This article have been viewed 2048 times

CHAPTER 86 FUNDAMENTALS OF MECHANICAL VENTILATION

Soumitra R. Eachempati, Marc J. Shapiro, Philip S. Barie

Mechanical ventilation is often required to manage trauma or critical illness, whether for airway protection, administration of general anesthesia, or management of acute respiratory failure (ARF) (Table 1). New technology now provides several modes by which a patient may be ventilated, with the goals of improved gas exchange, better patient comfort, and rapid liberation from the ventilator. Moreover, noninvasive positive-pressure ventilation permits some cases of ARF to be managed without insertion of an artificial endotracheal airway, and some patients who are extubated with marginal reserves to avoid reintubation. Nearly all ventilators can be set to allow full support of the patient on the one hand, and periods of exercise on the other. Thus, the choice of ventilator settings is a matter of physician preference for the majority of patients (Table 2). Controlled ventilation with suppression of spontaneous breathing leads rapidly to respiratory muscle atrophy; therefore modes of assisted ventilation are preferred wherein machine-delivered breaths are triggered by the patient’s own inspiratory efforts. Basic modes of assisted ventilation include assist-control ventilation (ACV), synchronized intermittent mandatory ventilation (SIMV), and pressure support ventilation (PSV).

Table 1 General Indications for Mechanical Ventilation

Airway maintenance or protection

Airway obstruction

General inhalational anesthesia

Hemodynamic instability

Hypoxemia

Metabolic acidosis

Pulmonary toilet (excessive secretions)

Table 2 Glossary of Basic Terminology of Mechanical Ventilation

Control: regulation of gas flow

Volume-controlled—volume limited, volume targeted; airway pressure is variable

Pressure-controlled—pressure limited, pressure targeted; volume delivered is variable

Dual-controlled—volume targeted (guaranteed), pressure limited; seldom used in practice

Cycling: ventilator switching from inhalation to exhalation after volume or pressure target (or limit) has been reached

Time-cycled—example: pressure control ventilation

Flow-cycled—example: pressure support ventilation

Volume-cycled—example: volume-controlled ventilation

Triggering: causes the ventilator to cycle to inhalation

Time-triggered—ventilator cycles at a set frequency as determined by the controlled rate

Pressure-triggered—ventilator senses the patient’s inspiratory effort from a decrease of airway pressure

Flow-triggered—constant flow of gas is delivered throughout the respiratory cycle (flow-by). Altered flow caused by patient inhalation is detected by the ventilator, which delivers a breath. Less work is done by the patient than with pressure-triggering

Breaths: cause the ventilator to cycle from inhalation to exhalation

Mandatory—controlled; determined by the respiratory rate

Assisted—examples: assist control, synchronized intermittent mandatory ventilation, pressure support

Spontaneous—no additional assistance, as in continuous positive airway pressure (CPAP)

Flow pattern: constant, decelerating, or sinusoidal

Sinusoidal—examples: spontaneous breathing and CPAP

Constant— flow continues at a constant rate until the set tidal volume is delivered, seldom used in practice

Decelerating—inhalation slows from a high initial flow rate as alveolar pressure increases; example: pressure-targeted ventilation, often used in volume-targeted ventilation, as it causes lower peak airway pressures than constant flow

Mode (breath pattern)

Controlled mechanical ventilation—controlled ventilation, without allowances for spontaneous breathing, typical of anesthesia ventilators

Assist-control—assisted breaths simulate controlled breaths

Intermittent mandatory ventilation—admixes controlled and spontaneous breaths, which may also be synchronized to prevent “stacking”

Pressure support—patient controls all aspects of breath except pressure limit

Most patients are started on mechanical ventilation for management of ARF, during which the work necessary to initiate a breath increases by a factor of four to six. The most common reason to initiate mechanical ventilation is to decrease the work of breathing by the patient. Additional potential benefits of mechanical ventilation include improved gas exchange, enhanced coordination between support and the patient’s own efforts, resting of respiratory muscles, prevention of deconditioning, and prevention of iatrogenic lung injury while promoting healing. However, unless settings are chosen carefully to synchronize with the patient’s own central respiratory drive, mechanical ventilation can cause an increase in work. Regardless of the mode chosen, all mechanical ventilation is a modification of the manner in which positive pressure is applied to the airway, and the interplay of the mechanical support and the patient’s own efforts.

NONINVASIVE VENTILATION

Ventilatory support delivered without establishing an endotracheal airway is “noninvasive ventilation” (NIV). Noninvasive ventilation was administered previously with intermittent negative pressure, but the current technique utilizes positive-pressure ventilation delivered through a nasal or face mask, and usage is expanding in the management of acute and chronic respiratory failure and possibly for some patients with heart failure.

Putative benefits of NIV are numerous, owing to avoidance of the complications of endotracheal intubation. Noninvasive ventilation preserves swallowing, feeding, speech, cough, and physiologic air warming and humidification by the nasooropharynx. Nonintubated patients communicate more effectively, require less sedation, and are more comfortable. In addition, patients are often able to continue with standard oral nutrition. Noninvasive ventilation eliminates complications such as trauma with tube insertion, mucosal ulceration, aspiration, infection (e.g., pneumonia, sinusitis), and dysphagia after extubation.

In a randomized, prospective trial following pulmonary resection of 48 patients with acute hypoxemic respiratory insufficiency, Auriant et al. compared standard invasive mechanical ventilation with nasal mask NIV. The need for postoperative reintubation and mortality were clearly reduced in patients receiving NIV as a part of respiratory support. Similarly, Squadrone et al. randomized 209 patients with respiratory failure in the postanesthesia care unit after major abdominal surgery to oxygen alone, or with continuous positive airway pressure (CPAP) via a mask. Patients who received oxygen plus CPAP had a significantly lower intubation rate, and also lower rates of pneumonia, infection, and sepsis.

Contraindications to Noninvasive Ventilation

Crucial to successful NIV is an awake, cooperative, spontaneously breathing patient. Airway, electrocardiographic, or hemodynamic instability argues against the use of NIV. An additional requirement is an intact cough reflex and ability to clear secretions, the absence of which is a common reason for failure of NIV. Relative contraindications include the inability to fit and seal the mask adequately, inability to cough with prompting, or inability to remove the mask in the event of emesis. A hypothetical contraindication is recent gastrointestinal surgery with aerophagia and gut distention. If pressures used to ventilate the patient are kept below 30 mm Hg, the closing pressure of the lower esophageal sphincter should not be exceeded, and aerophagia should be avoided. Morbid obesity is also a relative contraindication secondary to increased ventilatory pressure requirements arising from body habitus and the weight of the chest wall and abdominal viscera while the patient is supine.

Complications of Noninvasive Ventilation

The most common complication of NIV is focal skin necrosis, which is most common over the bridge of the nose but may also occur over the zygoma. The incidence is 7%–10% among patients receiving full-face-mask NIV. Other complications (incidence, 1%–2% each) include gastric distention, aspiration, and pneumothorax. Conceptual concerns with gastric distention are subsequent vomiting, aspiration, and pneumonia. Conjunctivitis may develop secondary to air leaks near the eyes in around 2% of patients.

The most serious complication is failure to recognize when noninvasive ventilation is not providing a patient with adequate ventilation, oxygenation, or airway patency. Delayed placement of an artificial airway, or failure of placement thereof, may cause continued deterioration or the death of a patient.

PRESSURE SUPPORT VENTILATION

Pressure support is a method of assisting spontaneous breathing in a ventilated patient, either partially or fully. The patient controls all parts of the breath except the pressure limit. The patient triggers the ventilator, which delivers a flow of gas in response up to a preset pressure limit (for example, 10 cm H₂O) depending on the desired minute ventilation (V_E). Gas flow cycles off when a certain percentage of peak inspiratory flow (usually 25%) has been reached. Tidal volumes (V_T) may vary, just as they do spontaneously.

Positive end-expiratory pressure (PEEP, also called continuous positive airway pressure [CPAP]) is added to restore functional residual capacity (FRC) to normal for the patient. When lung volumes are low, the work of breathing during early inhalation is reduced. Noncompliant lungs require higher intrapleural pressures to inflate to a normal V_T, even with CPAP. The addition of pressure support (PS) assists the patient to move up the pressure–volume curve (larger changes in volume for a given applied pressure, i.e., increased lung compliance). The term “pressure support ventilation” describes the combination of pressure support and PEEP (or CPAP). Although useful in the patient breathing spontaneously, PS may be used to assist spontaneous breaths in SIMV. Weaning may be facilitated using this combination, as the backup (SIMV) rate is weaned initially, and then the PS.

HELIOX

Helium has significantly lower density than air or nitrogen. Substituting helium for nitrogen reduces the density of the gas in direct proportion to the amount of helium admixed. Breathing heliox (the concentration in clinical use ranges from 80:20 to 60:40) results in more laminar flow, reduced airway resistance, and reduced work of breathing. Heliox may be useful in clinical situations where resistance to airflow is high, including asthma, acute exacerbations of chronic bronchitis or chronic obstructive pulmonary disease, other causes of bronchospasm, and upper airway obstruction with stridor. Breathing heliox is well tolerated, without significant adverse effects. Disadvantages include high cost and limited utility when a high F_IO₂ is required. Ventilators must also be recalibrated when heliox is delivered by ventilator rather than nebulizer (the usual route), to ensure that the flow of gas is measured correctly.

MODES OF MECHANICAL VENTILATION

Assist Control Ventilation

The ACV mode is the most commonly used mode in medical/surgical critical care units. Set parameters in ACV mode are inspiratory flow rate, frequency (f), and V_T. The ventilator delivers a set number of equal breaths per minute, each of a given V_T. Tidal volume and flow determine inspiratory (I) and expiratory (E) time and the I:E ratio. Plateau or alveolar pressure is related to V_T and respiratory system compliance. The patient has the ability to trigger extra breaths by exerting an inspiratory effort exceeding a preset trigger level. Typically, each patient will display a preferred rate for a given V_T and will trigger all breaths when f is set a few breaths per minute below the patient’s rate. In this mode, the control rate serves as adequate support should the patient stop initiating breaths. When high inspiratory effort continues during a ventilator-delivered breath, the patient may trigger a second superimposed breath. Patient effort can be increased, if desired, by increasing the triggering threshold or lowering V_T.

Synchronized Intermittent Mandatory Ventilation

In a passive patient, SIMV cannot be distinguished from ACV. Ventilation is determined by f and V_T. However, if the patient is not truly passive, respiratory work may be performed during mandatory breaths. In addition, the patient may trigger additional breaths by spontaneous effort. If the triggering effort comes in a brief, defined interval before the next mandatory breath, the ventilator will deliver the mandatory breath ahead of schedule to synchronize with patient inspiratory effort. If a breath is initiated outside the synchronization window, V_T, flow, and I:E are determined by patient effort and respiratory system mechanics, not by ventilator settings. These spontaneous breaths tend to be of low V_T and are variable from breath to breath. The SIMV mode is often used to augment patient work of breathing gradually by lowering the mandatory breath frequency or V_T, compelling the patient to breathe more rapidly in order to maintain adequate V_E. Some ventilators allow combinations of modes. A useful combination is SIMV plus PSV as a means to add “sigh” breaths and decrease atelectasis. Because SIMV plus PSV guarantees some backup V_E that PSV alone does not, this combination may be particularly useful for patients at high risk for deteriorating central respiratory drive, and it is also popular as an adjunct to weaning the ventilator.

Positive End-Expiratory Pressure

Although it is a ubiquitous form of ventilatory support, positive end-expiratory pressure (PEEP) can be confusing because the positive pressure is actually applied throughout the respiratory cycle and is more correctly termed CPAP. Using PEEP accomplishes three goals: prevention of alveolar derecruitment by restoring FRC, which is decreased in acute lung injury (ALI) and atelectasis, to the physiologic range; protection against injury during phasic opening and closing of atelectatic units; and assisting cardiac performance during heart failure, by increasing mean intrathoracic pressure.

The FRC is the lung’s physiologic reserve; loss of chest wall or lung compliance (the rate of change of volume in response to pressure) causes reduced FRC. The FRC is the volume of gas that remains in the lungs at the end of a normal tidal breath (∼2.5 liters); gas exchange does occur. At FRC, the tendency for the lungs to collapse is balanced by the tendency for the chest wall to move outward. A small vacuum in the pleural space assists in maintaining equilibrium, which is lost when pneumothorax is present.

The FRC is determined by the compliance of the lung and chest wall. Anything that constrains chest wall expansion reduces its compliance; likewise, anything that reduces lung volume reduces lung compliance. The FRC is composed of two volumes, the expiratory reserve volume (ERV) and the residual volume (RV). Below FRC, exhalation is active; lung tissue must be compressed to express gas. The RV (∼1 liter in adults) is the point where no more gas can be expressed from the lungs regardless of the pressure applied, because alveolar pressure exceeds atmospheric pressure and the gradient along the airway is reversed. Being filled with gas and coated with surfactant, alveoli are difficult to compress, but airways are compressible. When intrathoracic pressure exceeds pressure in the small airways, “dynamic airway collapse” occurs and gas is trapped in alveoli. Airway collapse increases the work of breathing and leads to ventilation-perfusion (V/Q) mismatch. Collapsed airways are difficult to reinflate, leading to a huge increase in the work of breathing and oxygen consumption.

The concept behind PEEP is to increase FRC; in essence, to allow alveoli to deflate only to the point just above where inflation remains easy (called the lower inflection point of the pressure–volume curve). The patient requires sufficient PEEP to prevent alveolar derecruitment, but not so much PEEP that alveolar overdistension, dead space ventilation from collapse of the alveolar microcirculation, or hypotension due to reduced right ventricular preload, right ventricular output, and ultimately cardiac output occur.

Auto-PEEP is caused by gas trapped in alveoli at end-expiration. This gas is not in equilibrium with the atmosphere and is at positive pressure, increasing the work of breathing. In patients with obstructive airways disease, increased bronchial tone leads to resistance to both inhalation and exhalation. Shortening of E (e.g., small airways disease, mucus plugging, pressure-controlled ventilation with inverted I:E) results in gas trapping at end-expiration, hyperinflation, and increased intrathoracic pressure, which abolishes the alveolar pressure gradient. Auto-PEEP can be ameliorated by lengthening E, shortening I, or decreasing the respiratory rate.

The ideal level of PEEP is controversial (Table 3). It may be that which prevents derecruitment of the majority of alveoli, while causing minimal overdistension; alternatively, application of PEEP is a recruitment maneuver, arguing for higher pressures to be applied to overcome alveolar collapse. Applying PEEP to put the majority of lung units on the favorable part of the pressure–volume curve will maximize gas exchange and minimize overdistention, but is easier said than done because the lower inflection point is sometimes indistinct. Undoubtedly, the combination of PEEP and low V_T prevents volutrauma, but the exact amount of PEEP to apply is controversial. The reason for this is hysteresis—the tendency of the lungs, due to surfactant, to exist at higher volumes in exhalation than in inhalation.

Table 3 Protocol Summary for Institution of Mechanical Ventilation for Acute Lung Injury/Acute Respiratory Distress Syndrome

Initial ventilator settings

Use a volume-controlled mode initially to ensure that VT is delivered.

Initial VT 8 ml/kg; reduce by 1 ml/kg/2 hours until 6 ml/kg is reached. Minimum VT is 4 ml/kg.

Set ventilator rate at 12–20 breaths/minute. Maximum ventilator rate 35 breaths/minute.

Adjust ventilator subsequently based on goals of arterial pH (7.25–7.45; ventilator rate) and end-inspiratory plateau pressure (Pplat) (<30 cm H₂O; VT).

Measure arterial pH upon admission to the ICU, every morning, and 15 minutes after each change in respiratory rate or VT.

Alkalemia is managed by decreasing ventilator rate by at least two breaths/minute.

Mild acidemia (pH 7.15–7.25) is managed by increasing ventilator rate until pH > 7.25 or PaCO₂ < 25 mm Hg up to 35 breaths/minute. If ventilator rate = 35 or PaCO₂ < 25 mm Hg, give sodium bicarbonate.

Severe acidemia (pH < 7.15) is managed by increasing the ventilator rate up to 35 breaths/minute. If ventilator rate = 35 and pH < 7.15, and sodium bicarbonate has been given, increase VT in increments of 1 ml/kg until pH > 7.15. It may be necessary to exceed the target Pplat under these conditions.

Keep Pplat < 30 cm H₂O. Measure Pplat at least every 8 hours, and 5 minutes after each change in PEEP or VT, and more frequently when changes in lung compliance are likely. Accurate measurement of Pplat requires a patient who is not moving or coughing.

If Pplat cannot be measured because of an air leak, peak inspiratory pressure may be substituted.

Target ranges for PaO₂ are 55–80 mm Hg, or SaO₂ > 88%. The combination of PEEP and F_IO₂ is discretionary, but F_IO₂:PEEP should generally be < 5 if F_IO₂ > 0.45.

When increasing PEEP above 10 cm H₂O, increase by 2–5 cm increments up to a maximum of 35 cm H₂O, until target ranges for PaO₂ are reached. Reduce PEEP to the previous level of PEEP if the change does not increase PaO₂ > 5 mm Hg or if decreased oxygen delivery results from a decrease of cardiac output.

Assess arterial oxygenation by blood gas determination or oximetry at least every 4 hours.

If arterial oxygenation is below the target range, increase F_IO₂ incrementally (up to 1.0), then PEEP (up to 35 cm H₂O within 30 min). Reassess every 15 min after each adjustment until target ranges for PaO₂ are regained. Brief periods of SaO₂ < 88% (<5 min) may be tolerated. F_IO₂ = 1.0 may be used transiently (<10 min) for arterial desaturation or during suctioning or bronchoscopy.

PEEP, Positive end-expiratory pressure.

Adapted from Nathens AB, Johnson JL, Mine JP, et al. Inflammation and the Host Response to Injury Investigators: Inflammation and the host response to injury, a large-scale collaborative project: patient-oriented research core—standard operating procedures for clinical care. I. Guidelines for mechanical ventilation of the trauma patient. J Trauma 59:764–769, 2005.

Ventilator “Bundle”

Care of the patient who requires mechanical ventilation is more than just a matter of providing ventilation and oxygenation. Such patients are often critically ill and at risk of numerous complications, not all of which are related directly to acute respiratory failure or mechanical ventilation. Therefore, it is important for the clinician to bear in mind the total patient. The patient may be at prolonged bed rest, and at risk for deconditioning, venous thromboembolic complications, and the development of pressure ulcers. Neurologic compromise from disease or sedative/analgesic drugs may impair the sensorium sufficiently that the patient cannot protect his or her airway, increasing the risk of pulmonary aspiration of gastric contents. Oversedation may be one component aspect of prolonged mechanical ventilation, which is a definite risk factor for development of ventilator-associated pneumonia (VAP). Prolonged mechanical ventilation (>48 hours) is itself a marker of critical illness, specifically the development of stress-related gastric mucosal hemorrhage, a rare but serious (∼50% mortality) harbinger of adverse outcomes of critical illness.

Using the principles of evidence-based medicine, several “best practices” have been combined into a “ventilator bundle” to optimize the outcomes of mechanical ventilation. The bundle consists of four maneuvers: Keeping the head of the patient’s bed up at least 30 degrees from level at all times unless contraindicated medically; prophylaxis against venous thromboembolic disease; prophylaxis against stressrelated gastric mucosal hemorrhage; and a daily “sedation holiday” to assess for readiness to liberate from mechanical ventilation through assessment of a trial of spontaneous breathing. Careful adoption and adherence to all facets of the bundle can decrease the substantial risk of VAP, along with other maneuvers such as adherence to the principles of infection control.

Routine Settings

Ventilator settings are based on the patient’s ideal body mass and medical condition. The risk of oxygen toxicity from prolonged exposure to a fraction of inspired oxygen (F_IO₂) greater than 60% is minimized by using the lowest F_IO₂ that can oxygenate arterial blood satisfactorily (e.g., arterial oxygen tension [PaO₂] of 60 mm Hg or an oxygen saturation [SaO₂] of 88%) (see Table 3).

The normal lung (e.g., during general anesthesia) may be ventilated safely with V_T 8–10 ml/kg for prolonged periods. Historically, critically ill patients with acute lung injury (ALI)/acute respiratory distress syndrome (ARDS) have been ventilated with V_T 10–15 ml/kg of ideal body mass, which is now considered inappropriate due to convincing data from experiments indicating that alveolar overdistention can produce endothelial, epithelial, and basement membrane injuries associated with increased microvascular permeability and iatrogenic lung injury (VILI). Direct monitoring of alveolar volume is not feasible. A reasonable substitute is to estimate peak alveolar pressure as obtained from the plateau pressure (P_plat) measured in a relaxed patient by occluding the ventilator circuit briefly at end-inspiration. In patients with pulmonary dysfunction, there is a growing tendency to reduce the V_T delivered to 4–6 ml/kg or less in order to achieve a P_plat no higher than 35 cm H₂O. The incidence of VILI increases markedly when P_plat is high. Low V_T ventilation may lead to an increase in PaCO₂. Acceptance of elevated CO₂ tension in exchange for controlled alveolar pressure is termed “permissive hypercapnia.” It is important to focus on pH rather than PaCO₂ if this approach is employed. If the pH falls below 7.25, increase V_E or administer NaHCO₃.

The f that is set depends on the mode. With ACV, the backup rate should be about four breaths per minute less than the patient’s spontaneous rate to ensure that the ventilator will continue to supply adequate V_E should the patient have a sudden decrease in spontaneous breathing. With SIMV, the rate is typically high at first and then decreased gradually in accordance with patient tolerance.

An inspiratory flow rate of 60 l/min is used with most patients during ACV and SIMV. In patients with chronic obstructive pulmonary disease, better gas exchange may be achieved at a flow rate of 100 l/min, probably because the resulting increase in E allows for more complete emptying of trapped gas. If the flow rate is insufficient to meet the patient’s requirements, the patient will strain against his or her own pulmonary impedance and that of the ventilator, with a consequent increase in the work of breathing.

In the ACV, SIMV, and pressure control modes (discussed in the next chapter), the patient must lower airway pressure below a preset threshold (usually minus 1–2 cm H₂O) in order to trigger the ventilator to deliver a tidal breath. In most situations, this is straightforward; the more negative the sensitivity the greater the effort demanded of the patient. When auto-PEEP is present, the patient must lower alveolar pressure by the amount of auto-PEEP in order to have any impact on airway opening pressure, and then further by the trigger amount to initiate a breath, increasing dramatically the work of breathing. Flow triggering systems have been used to reduce the work of triggering the ventilator. In contrast to the usual approach in which the patient must open a demand valve in order to receive assistance, continuous flow systems maintain a high continuous flow, and then further augment flow when the patient initiates a breath. These systems reduce the work of breathing slightly below that present using conventional demand valves, but do not solve the triggering problem when breath stacking occurs.

Sedation

Most patients who require mechanical ventilation will require sedation, but only a minority (∼10%) will also require neuromuscular blockade. A panoply of agents are available for both (Table 4), so the choice of agent can be individualized for the patient, but caution must be exercised so that patients receive only what they need and are not oversedated. Titration of sedation such that patients are comfortable is facilitated by ordering sedation to be titrated to a sedation score of 3–4 points on the Ramsay or Riker scale (Table 5). Intermittent doses of sedatives are preferred to continuous infusions, also to attempt to minimize the amount of sedation. Neuromuscular blockade should be avoided whenever possible.

Table 4 Selected Formulary for Analgesia, Anesthesia, and Sedation in the Intensive Care Unit

Agent	Initial IV Adult Dose	Comments
		Induction Agents
Etomidate	≥6 mg	Maintains CO and BP. Reduces ICP but maintains CPP. Short T_½; use infusion for maintenance. Possible adrenal suppression.
Ketamine	1–2 mg/kg	R apid-onset, short-duration agent for induction of anesthesia. Can be given by maintaining continuous infusion, and at lower dose for sedation without anesthesia. Transiently increases BP and HR, raises ICP and intraocular pressure. Usually does not depress breathing. Generally safe in pregnancy and for neonates and children. Concurrent narcotics or barbiturates may prolong recovery. Can cause anxiety, disorientation, dysphoria, and hallucinations, which may be reduced by a short-acting benzodiazepine during emergence. Atropine pretreatment is recommended to decrease secretions, but may increase incidence of dysphoria. Hepatic metabolism.
Propofol	1.5–2.5 mg/kg	Provides no analgesia. Potent amnestic effect. Causes apnea and loss of gag reflex. Can cause marked low BP. Infuse at 0.05–0.3 mg/kg/min for prolonged sedation. Minimal accumulation (hepatic insufficiency) facilitates rapid elimination. Account for 1 kCal/ml (lipid infusion) in nutrition prescription. Use of same vial >12 hours associated with bacteremia. Safety for children still debated.
		Intravenous Sedatives/Analgesics
Midazolam	0.5–4 mg	Short T_½, but accumulates during infusion owing to active metabolites. Only benzodiazepine with potent amnestic effect. Can cause low BP and loss of airway. Primary use is short-term sedation for ICU procedures. Renal elimination.
Lorazepam	1–4 mg	Effective anxiolytic. Preferred agent for continuous infusion of benzodiazepine (starting dose 1 mg/hr). Can cause low BP, especially with hypovolemia, and paradoxical agitation. Hepatic elimination.
Morphine	2–10 mg	Analgesic and sedative effects. Can cause low BP, CO, and apnea. Tolerance and withdrawal possible after long-term use. Can be given as IV infusion or by PCA for analgesia or to facilitate prolonged mechanical ventilation or withdrawal of care. Hepatic elimination.
Hydromorphone	0.5–2.0 mg	Hydrated ketone of morphine with similar use and risk profiles. Approximately eight-fold more potent than morphine. Hepatic elimination.
Fentanyl	50–100 mcg	Approximately 50-fold potency compared with morphine, but less likely to cause low BP in appropriate dosage (less histamine release). Versatile for ICU use given IV or by epidural infusion or PCA. Less potent than local anesthetics for epidural analgesia or abrogation of surgical stress response. Can cause truncal rigidity and apnea with inability to ventilate by hand (use neuromuscular blockade to facilitate intubation in that setting). Hepatic elimination.
		Neuromuscular Blocking Agents
Succinylcholine	0.75–1.5 mg/kg	Only dep olarizing NBMA (occupies ACh receptor). Rapid onset, effect dissipates within 10 minutes of single dose. Causes hyperkalemia. Can precipitate malignant hyperthermia. Increases ICP and intraocular pressure. Contraindicated in TBI, spinal cord injury, neuromuscular disease, and burns. Metabolized by plasma cholinesterase; absence of enzyme (relatively common) causes prolonged paralysis.
Atracurium	0.2–0.5 mg/kg	Short-acting nondepolarizing NMBAs (competitive inhibitor of ACh).
Cisatracurium	0.2–0.5 mg/kg	Relatively slow in onset (also competitive inhibitor of ACh). Atracurium and cisatracurium are similar, except that the former causes histamine release and can cause high HR and low BP. Cisatracurium now used preferentially. Short acting, requires IV infusion for prolonged effect. Effect potentiated by hypokalemia. Many drug interactions. Metabolized by Hoffman elimination/ester hydrolysis, and is thus used for patients with renal/hepatic insufficiency.
Pancuronium	0.05–0.1 mg	Rapid onset, prolonged effect. Causes increased BP and HR. Used for induction of neuromuscular blockade, but should be converted to a drug such as continuousinfusion cisatracurium for maintenance. Renal/hepatic elimination, accumulates in organ dysfunction.
Vecuronium	0.08–0.10 mg/kg	Nondepolarizing NMBA with rapid onset and short duration of action. Less potential for histamine release. Can cause malignant hyperthermia. Metabolized by liver.
		Miscellaneous Agents
Dexmedetomidine	1 mcg/kg load, then 0.2–0.7 mcg/kg/hr	Central selective α₂ agonist used for short-term (<24 hours) sedation. Sympatholysis lowers HR and BP. Can achieve light sedation, does not depress respirations. No anamnestic effect. Useful for drug/alcohol withdrawal and sedation when liberation from mechanical ventilation is imminent. Expensive.
Haloperidol	2–5 mg	Used commonly for anxiolysis (often over lorazepam), especially when respiratory depression is undesirable. IV administration, not FDA-approved, is commonplace. Antidopaminergic properties contraindicate use in Parkinson disease. Causes extrapyramidal effects. Hepatic elimination.
Ketorolac	0.5–1.0 mg/kg	Parenteral NSAID used in lieu of opioids or for opioid-sparing effect in combination. Irreversible platelet dysfunction; can cause incisional or GI hemorrhage and acute renal failure. Use strictly limited to less than 5 days in postoperative period.
		Reversal Agents
Flumazenil	0.1–0.2 mg	Benzodiazepine antagonist. Rapid onset and short duration. Adverse effect of benzodiazepine can persist after drug wears off. Repeated doses of up to 0.8 mg can be used. Abrupt antagonism of chronic benzodiazepine use can precipitate seizures.
Naloxone	0.4 mg	Opioid antagonist. Rapid onset and short duration. Often diluted 0.4 mg/10 ml and titrated 0.04–0.08 mg at a time to reverse undesirable side effects while preserving analgesia. Repeated doses of up to 0.4 mg or continuous IV infusion can be used. Abrupt opioid antagonism can precipitate hypertension, increased HR, pulmonary edema, or myocardial infarction.
Edrophonium with atropine	0.5–1.0 mg/kg 0.007–0.014 mg/kg

Anticholinesterase inhibitor with antidysrhythmic properties. Rapid onset, short duration; therefore used usually in concert with atropine, which counteracts the increased secretions, decreased HR, and bronchospasm. Not effective for reversal of neuromuscular blockade caused by depolarizing agents. Renal and hepatic elimination (edrophonium). Atropine may cause fever. Neostigmine with glycopyrrolate

0.5-2.0 mg

0.1-0.2 mg

Neostigmine causes salivation and severe low HR. May cause laryngospasm or bronchospasm. Renal metabolism. Not effective for reversal of neuromuscular blockade caused by depolarizing agents. Because of profound low HR, given in same syringe with glycopyrrolate (or sometimes atropine). Glycopyrrolate counteracts low HR, and unopposed causes increased HR. May cause fever. Glycopyrrolate is contraindicated in GI ileus/obstruction and in neonates.

ACh, Acetylcholine; BP, blood pressure; CO, cardiac output; CPP, cerebral perfusion pressure; ICP, intracranial pressure; FDA, U.S. Food and Drug Administration; GI, gastrointestinal; HR, heart rate; IV, intravenous; NBMA, neuromuscular blocking agent; NSAID, nonsteroidal anti-inflammatory drug; PCA, patient-controlled analgesia; T½, elimination half-life; TBI, traumatic brain injury; VO₂, oxygen consumption.

Table 5 Sedation Scales in Common Usage

	Value	Clinical Correlate
Ramsay Sedation Score
Awake scores 1–3	1	Anxious, agitated, or restless
	2	Cooperative, oriented, tranquil
	3	Responsive to commands
Asleep scores 4–6	4	Brisk response to stimulus ^a
	5	Sluggish response to stimulus
	6	No response to stimulus
Riker Sedation-Agitation Scale
Dangerous agitation	7	Pulling at catheters, striking staff
Very agitated	6	Does not calm to voice, requires restraint
Agitated	5	Anxious, responds to verbal cues
Calm and cooperative	4	Calm, awakens easily, follows commands
Sedated	3	Awakens to stimulus
Very sedated	2	Arouses to stimulus, does not follow commands
Unarousable	1	Minimal or no response to noxious stimulus

a Stimulus is light glabellar tap or loud auditory stimulus.

Adapted from Ramsay M, Savege T, Simpson B, et al: Controlled sedation with alphaxalon-alphadolone. BMJ 2:656–659, 1974; and Riker RR, Picard JT, Fraser GL: Prospective evaluation of the Sedation-Agitation Scale for adult critically ill patients. Crit Care Med 27:1325–1329, 1999.

Prolonged or excessive sedation increases the duration of mechanical ventilation and increases the likelihood of tracheostomy. Protocolized weaning of sedative medications and daily sedation “holidays” to permit spontaneous breathing trials (see following) shorten the duration of mechanical ventilation and decrease the risk of VAP and other complications.

MONITORING

Blood Gases

Blood gas analyzers report a wide range of results, but the only parameters measured directly are the partial pressures of oxygen (pO₂) and carbon dioxide (pCO₂), and blood pH. The arterial blood hemoglobin saturation (SaO₂) is calculated from the pO₂ using the oxyhemoglobin dissociation curve, assuming a normal P₅₀ (the pO₂ at which SaO₂ is 50%, normally 26.6 mm Hg), and that hemoglobin is normal structurally. Some blood gas analyzers incorporate a co-oximeter that measures the various forms of hemoglobin directly, including oxyhemoglobin, total hemoglobin, carboxyhemoglobin, and methemoglobin. The actual HCO₃^–, standard HCO₃^–, and base excess are calculated from the pH and pCO₂.

A freshly drawn, heparinized, bubble-free arterial blood sample is required. Heparin is acidic; if present to excess, the measured pCO₂ and calculated HCO₃^– are reduced spuriously. Delayed analysis allows continued metabolism by erythrocytes, reducing pH and pO₂ and increasing pCO₂; keeping the specimen iced preserves accuracy for up to 1 hour. Air bubbles cause a decrease in pCO₂ and an increase in pO₂.

The solubility of all gases in blood, including CO₂ and O₂, increases with a decrease in temperature. Thus, hypothermia causes the pO₂ and pCO₂ to decrease and pH to increase. As analysis of a sample taken from a hypothermic patient occurs at 37° C, the pO₂ and pCO₂ results are artificially high, but the error is usually too small to be meaningful clinically.

Pulse Oximetry

Pulse oximetry calculates SaO₂ by estimating the difference in intensity of signal between oxygenated and deoxygenated blood from the red (660 nm) and near-infrared (940 nm) regions of the light spectrum. To function accurately, pulse oximetry must detect pulsatile blood flow, but all things being equal, pulse oximetry data can be obtained from a detector on the finger, the earlobe, or even the forehead. Pulse oximetry is generally accurate (+2%) over the range of SaO₂ 70%–100%, but less accurate below 70%.

Several aspects of the technology and patient physiology limit the accuracy of pulse oximetry. If the device cannot detect pulsatile flow, the waveform will be damped and unable to provide an accurate estimation. Consequently, patients with hypothermia, hypotension, hypovolemia, or peripheral vascular disease, or who are treated with vasoconstrictor medications (e.g., norepinephrine), may have inaccurate pulse oximetry readings. Additionally, an elevated carboxyhemoglobin concentration will lead to falsely elevated SaO₂ because reflected light is absorbed at the same wavelength as oxygenated hemoglobin. Other situations contributing to inaccurate pulse oximetry include the presence of ambient light and motion artifact.

Capnography

Capnography measures changes in the concentration of CO₂ in expired gas during the ventilatory cycle. This technique is most reliable in ventilated patients and employs either mass spectroscopy or infrared light absorption to detect the presence of CO₂. The gas may be collected by sidestream or mainstream sampling; the former is most common and has the advantage of a lightweight analyzer. However, sidestream sampling is susceptible to accumulation of water vapor in the sampling line. In the ICU, where respiratory gases are humidified, mainstream sampling may be preferable.

The peak CO₂ concentration occurs at end-exhalation and is regarded as the patient’s “end-tidal CO₂” (ETCO₂), at which time ETCO₂ is in close approximation to the alveolar gas concentration. Capnography is useful in the assessment of successful tracheostomy or endotracheal tube placement, to monitor weaning from mechanical ventilation, and as a monitor of resuscitation. The ability to detect hypercarbia during ventilator weaning can diminish the need for serial blood gas measurements. In conjunction with pulse oximetry, many patients can be weaned successfully from mechanical ventilation, without reliance upon arterial blood gases or invasive hemodynamic monitoring.

Other information is acquired from capnography as well. Prognostically, an ETCO₂-PaCO₂ gradient of 13 mm Hg or more after resuscitation has been associated with increased mortality in trauma patients. A sudden decrease or even disappearance of ETCO₂ can be correlated with potentially serious pathology or events, such as a low cardiac output state, disconnection from the ventilator, or pulmonary thromboembolism. A gradual increase of ETCO₂ can be seen with hypoventilation; the converse is also true. Another cause of gradually decreasing ETCO₂ is hypovolemia.

INVASIVE HEMODYNAMIC MONITORING

Arterial Catheterization

Measurement of arterial blood pressure is one of the simplest, most reproducible methods of evaluating hemodynamics. Automated noninvasive blood pressure cuff devices are accurate (error, ±2%), but take measurements only periodically. If fluctuations require more frequent monitoring, continuous monitoring is available via an indwelling arterial catheter. Indications for invasive arterial monitoring include prolonged operations or prolonged mechanical ventilation (>24 hours), unstable hemodynamics, substantial blood loss, a need for frequent blood sampling, or a need for precise blood pressure control, (e.g., neurosurgical patients, patients on cardiopulmonary bypass). Although there is morbidity from insertion and from indwelling catheters, there is also morbidity from repetitive arterial punctures; the risk:benefit analysis is a matter of clinical judgment for “less-unstable” patients.

Arterial catheters may be placed in any of several locations. The catheter should be a special-purpose thin-walled catheter to maintain fidelity of the waveform and to avoid obstructing the vessel lumen; a standard intravenous cannula should not be used. The radial artery at the wrist is the most commonly used site; although the ulnar artery is usually of larger diameter, it is relatively inaccessible percutaneously. Careful confirmation of a patent collateral circulation to the hand is mandatory before cannulation of an artery at the wrist, to minimize potential tissue loss from arterial occlusion or embolization. In neonates, the umbilical artery may be catheterized; intestinal ischemia is a rare complication. The axillary artery is relatively spared by atheromata, supported by good collaterals at the shoulder, and easy to cannulate percutaneously, making it a suitable choice. The superficial femoral artery may also be used, but is not a location of choice because the burden of plaque (and therefore the risk of distal embolization) is higher, as is the infection rate. The superficial temporal artery is difficult to cannulate because of small caliber and tortuosity. The dorsalis pedis artery is accessible, but should be avoided in patients with peripheral vascular disease. The brachial artery should be strictly avoided, because the collateral circulation around the elbow is poor and the risk of ischemia of the hand or forearm is high. Severe peripheral vasoconstriction due to vasopressor therapy may necessitate a longer catheter at a more central location (e.g., axillary, femoral) in order to place the catheter tip into an artery in the torso that would be less affected. Nosocomial infection of arterial catheters is unusual, provided that basic tenets of infection control are honored and femoral artery catheterization is avoided. Other complications from arterial catheterization include bleeding, hematoma, and pseudoaneurysm.

Central Venous Pressure Monitoring

The central venous pressure (CVP) is an interplay of the circulating blood volume, venous tone, and right ventricular function. The CVP measures the filling pressure of the right ventricle, providing an estimate of intravascular volume status. Central venous access can be obtained via the basilic, femoral, external jugular, internal jugular, or subclavian vein. In the ICU, the internal jugular, subclavian, and femoral veins are used in decreasing frequency. The internal jugular site is the most popular because of ease of accessibility, a high technical success rate of cannulation, and a low rate of complications. However, it is difficult to keep an adherent dressing in place, and the infection rate is higher than for subclavian catheters. Subclavian insertion is technically demanding, and has the highest rate of pneumothorax (1.5%–3%), but the lowest rates of infection. The femoral vein site is least preferred, despite the relative ease of catheter placement. It is accessible during cardiopulmonary resuscitation or emergency intubation, so procedures can occur concurrently. However, the site is particularly prone to infection, and the risks of arterial puncture (9%–15%) and venous thromboembolic complications are higher than for jugular or subclavian venipuncture. Overall complications are comparable for internal jugular and subclavian vein cannulation (6%–12%), and higher for femoral vein cannulation (13%–19%). The incidence of carotid puncture during internal jugular cannulation (6%–9%) is higher than the incidence of puncture of the subclavian artery during subclavian vein catheterization (3%–5%).

PULMONARY ARTERY CATHETERIZATION

A pulmonary artery catheter (PAC) is a balloon-tipped, flowdirected catheter that is usually inserted percutaneously via a central vein and transits the right heart into the PA. Data from PACs are used mainly to determine cardiac output (Q) and preload, which is most commonly estimated in the clinical setting by the PA occlusion pressure (PAOP). Pulmonary artery diastolic pressure corresponds well to the PAOP. Diastolic pressure can exceed the PAOP when pulmonary vascular resistance is high (e.g., pulmonary fibrosis, pulmonary hypertension).

Normally, PAOP approximates left atrial pressure, which in turn approximates left ventricular end-diastolic pressure (LVEDP), itself a reflection of left ventricular end-diastolic volume (LVEDV). The LVEDV represents preload, which is the actual target parameter. Factors that may cause PAOP to reflect LVEDV inaccurately include mitral stenosis, high levels of PEEP (>10 cm H₂O), and changes in left ventricular compliance (e.g., due to myocardial infarction, pericardial effusion, or increased afterload). Inaccurate readings may result from balloon overinflation, improper catheter position, alveolar pressure exceeding pulmonary venous pressure (as with ventilation with PEEP), or severe pulmonary hypertension (which may make PAOP measurement hazardous). Elevated PAOP occurs in left-sided heart failure. Decreased PAOP occurs with hypovolemia or decreased preload.

A desirable feature of PA catheterization is the ability to measure mixed venous oxygen saturation (S_mvO₂), although controversially, sampling from the superior vena cava via a central venous catheter may provide data of comparable utility. True mixed venous blood is blood from both the superior and inferior vena cava admixed in the right atrium, which may be sampled for blood gas analysis from the distal port of the PAC. Some catheters have embedded fiberoptic sensors that measure S_mvO₂ directly. Causes of low S_mvO₂ include anemia, pulmonary disease, carboxyhemoglobinemia, low Q, and increased tissue oxygen demand. The S_aO₂:(S_aO₂ – S_mvO₂) ratio determines the adequacy of O₂ delivery (DO₂). Ideally the P_mvO₂ should be 35–40 mm Hg, with a S_mvO₂ of about 70%. Values of P_mvO₂ < 30 mm Hg are critically low.

CLINICAL USE OF THE PULMONARY ARTERY CATHETER

Many indications have been championed for the PAC, but no studies have demonstrated unequivocally that PAC use decreases morbidity or mortality. Pulmonary artery catheters may still be useful in cardiomyopathy, shock of various etiologies, suspected pulmonary hypertension, or an unpredicted or poor response to conventional fluid therapy. Transesophageal echocardiography is an alternative, but it is not widely available outside cardiac surgery operating rooms. Critically ill patients receiving inotropic agents despite resuscitation with large volumes of fluid may also benefit from monitoring by PAC. However, a recent ARDSnet trial demonstrated no differences in outcome when ALI/ARDS was managed by PAC versus CVP monitoring.

LIBERATION FROM MECHANICAL VENTILATION

Objective measures and proactive strategies are available to hasten the moment when mechanically ventilated patients can be liberated from the ventilator. The stakes are high, because each day of mechanical ventilation via artificial airway (e.g., endotracheal or tracheostomy tube) increases the need for sedation, which may postpone “liberation day.” Moreover, each day of mechanical ventilation increases the risk of VAP, which may prolong further the need for mechanical ventilation.

Some patients do not separate readily from the ventilator, which may be due to disease- or therapy-related reasons. Most clinical cases of failed liberation from the ventilator are multifactorial, but respiratory muscle fatigue is a common factor, in that the load on the respiratory system exceeds the capacity to breathe (Table 6). The increased load may take the form of a demand for increased V_E, or increased work of breathing. Increased V_E may result from increased CO₂ production, increased dead space (V_D) ventilation, or increased ventilatory drive. Increased CO₂ production may be caused by a catabolic state, or excess carbohydrate administered during nutritional support. Increased V_D (ventilation of unperfused or under-perfused lung) may be caused by decreased Q, pulmonary embolism, pulmonary hypertension, severe ALI, or iatrogenically from positive-pressure ventilation. Increased ventilatory drive may occur from muscle fatigue or failure, stimulation of pulmonary J receptors (usually by lung inflammation or parenchymal hemorrhage), or lesions of the central nervous system. Psychological stress is also an important factor that may manifest itself as tachypnea, hypoxemia, or agitation or delirium. Stress may be caused by inadequate analgesia or sedation, or untreated delirium. Acute alcohol or drug withdrawal is a major factor in some patients.

Table 6 Load on Respiratory System

Increased work of breathing results from either increased airflow resistance or decreased thoracic compliance. Airway obstruction can result from reversible small airways disease (e.g., bronchospasm), tracheal stenosis, tracheomalacia, glottic edema or dysfunction, mucus plugging, or muscle weakness or fatigue. Muscle dysfunction may be caused by nutritional or metabolic causes (including hypocalcemia, hypokalemia, or hypophosphatemia). The critical illness polyneuropathy syndrome has poorly understood pathophysiology, but is associated with sepsis and multiple organ dysfunction syndrome, and is often diagnosed when sought specifically by electromyography. Other potential causes of muscular failure or weakness include hypoxemia, hypercarbia, and possible anemia.

Patients who “fight” the ventilator technically have the syndrome of patient–ventilator dyssynchrony. The cause can usually be found and must be sought; to sedate the patient more deeply (or administer neuromuscular blockade) before correctable causes are identified and remedied is incorrect and may be catastrophic if an unstable airway is the cause. A systematic approach to evaluation is advocated; recognizing that the patient and the ventilator are supposed to be working in concert facilitates an understanding that the problem may be the patient or the ventilator. The cause may be found anywhere on the continuum from the alveolus to the power outlet or the source of respiratory gases, and must be sought systematically (Table 7). The first step is always to ensure that the patient has a patent airway that is positioned properly.

Table 7 Therapies to Reverse Ventilatory Failure

Improve muscular function
Treat sepsis—avoid aminoglycosides Nutritional support without overfeeding (follow indirect calorimetry) Replete electrolytes to normal Assure periods of rest—do not exhaust the patient Limit neuromuscular blockade Avoid oversedation Identify/correct hypothyroidism
Reduce respiratory load
Airway resistance	Ensure airway patency/adequate caliber
Compliance (elastance)	Treat pneumonia
	Treat pulmonary edema
	Identify/reduce intrinsic PEEP (auto-PEEP)
	Drain large pleural effusions
	Evacuate pneumothorax
	Treat ileus (promotility agents)
	Decompress abdominal distention/treat abdominal compartment syndrome
	Position patient 30 degrees head up
Minute ventilation	Treat sepsis
	Antipyresis (temperature > 40° C
	Avoid overfeeding
	Correct metabolic acidosis
	Identify/reduce intrinsic PEEP (auto-PEEP)
	Bronchodilators
	Maintain least possible PEEP
	Resuscitate shock/correct hypovolemia
	Identify and treat pulmonary embolism

PEEP, Positive end-expiratory pressure.

Liberation from mechanical ventilation can be easy in patients requiring short-term support. However, as many as 25% of patients will experience respiratory distress such that ventilation has to be reinstituted; patients recovering from acute respiratory failure, necrotizing pneumonia, or major torso trauma can be especially challenging. Patients who cannot be weaned have a characteristic response to trials of spontaneous breathing: there is an almost immediate increase in respiratory rate and decrease in V_T. As the trial of spontaneous breathing continues over 30–60 minutes, work of breathing increases substantially by four- to seven-fold. Increased oxygen demand is met by increased oxygen extraction, which eventually causes decreased DO₂ and arterial hypoxemia. Pulmonary compliance decreases, and gas trapping from lengthened I:E doubles measured auto-PEEP. The rapid, shallow breathing pattern causes CO₂ retention because of increased dead space ventilation despite increased V_E. There is considerable cardiovascular stress also, with pulmonary and systemic hypertension and increased afterload on both ventricles, likely from the extreme changes in intrathoracic pressure generated by the struggling patient.

Timing is important; if weaning is delayed unnecessarily, the patient remains at risk for a host of ventilator-associated complications. If weaning is performed prematurely, failure may lead to cardiopulmonary decompensation and further prolonged mechanical ventilation. In general, discontinuation of mechanical ventilation is not attempted in the setting of cardiopulmonary instability or PaO₂ < 60 mm Hg with an F_IO₂ of 0.60 or higher. However, satisfactory oxygenation does not predict successful weaning reliably; rather, a more important determinant is the ability of respiratory muscles to perform increased respiratory work. Decisions based solely on clinical judgment are frequently erroneous. Parameters gathered traditionally, including maximal negative inspiratory pressure, vital capacity, and V_E, have limited predictive accuracy. Respiratory frequency (f)/V_T during 1 minute of spontaneous breathing (the Rapid Shallow Breathing Index) is a more accurate predictor (95% probability of success) if f/V_T is less than 80 after a 30-minute trial of spontaneous breathing. Calculation of f/V_T during PSV is considerably less accurate.

The process of weaning begins by determining patient readiness (Figure 1). Patients should be screened carefully for hemodynamic stability, cooperative mental status, respiratory muscle strength, consistent and adequate wakefulness, ability to manage secretions, nutritional repletion and normalization of acid–base and electrolyte status, and an artificial airway of adequate size. Particular attention should be given to acceptance of hypercapnia if chronically present and avoidance of new metabolic alkalosis. Finally, ensure normality of electrolytes affecting muscle function (e.g., calcium, phosphate, and potassium). If the aforementioned conditions are addressed, weaning may be attempted.