When breathing becomes difficult or impossible, it may be necessary to use a ventilator to sustain life. Usually the need for a ventilator is temporary, such as during a surgical procedure. However, if breathing difficulty is chronic, ventilatory support may be required for an extended period, sometimes a lifetime. The main indications for ventilatory support are respiratory insufficiency resulting in hypoventilation (not enough gas moving into and out of the lungs), hypoxemia (not enough oxygen in the arterial blood), or hypercapnea (too much carbon dioxide in the blood). Many medical conditions can cause severe respiratory insufficiency requiring ventilatory support. Examples include cervical spinal cord injury (rostral enough to impair diaphragm function), muscular dystrophy, amyotrophic lateral sclerosis, and chronic obstructive pulmonary disease.

Several types of ventilator systems are available for individuals with respiratory insufficiency, including positive-pressure ventilators, negative-pressure ventilators, phrenic nerve pacers, abdominal pneumobelts, and rocking beds (Banner, Blanch, and Desautels, 1990; Hill, 1994; Levine and Henson, 1994). Positive-pressure ventilators operate by “pushing” air into the pulmonary system for inspiration, whereas negative-pressure ventilators work to lower the pressure around the respiratory system and expand it for inspiration. Phrenic nerve pacers stimulate the phrenic nerves and cause the diaphragm to contract to generate inspiration. Abdominal pneumobelts displace the abdomen inward (by inflation of a bladder) to push air out of the pulmonary system for expiration, and then allow the abdomen to return to its resting position (by deflation of the bladder) for inspiration. Rocking beds are designed to move an individual upward toward standing and downward toward supine to drive inspiration and expiration, respectively, using gravitational force to displace the abdomen and diaphragm. All of these systems are currently used (Make et al., 1998); however, the most commonly used one today and the one that the speech-language pathologist is most likely to encounter in clinical practice is the positive-pressure ventilator (Spearman and Sanders, 1990).

The positive-pressure ventilator uses a positive-pressure pump to drive air through a tube into the pulmonary system. The tube can be routed through (1) the larynx (in this case, it is called an endotracheal tube), such as during surgery or acute respiratory failure; (2) the upper airway, via a nose mask, face mask, or mouthpiece (this is called noninvasive ventilation); or (3) a tracheostoma (a surgically fashioned entry through the anterior neck to the tracheal airway). The latter two modes of delivery are used in individuals who need long-term ventilatory support. With noninvasive positive-pressure ventilation, speech is produced in a relatively normal manner. That is, after inspiratory air from the ventilator flows into the nose and/or mouth, expiration begins and speech can be produced until the next inspiration is delivered. The situation is quite different, however, when inspiratory air is delivered via a tracheostoma. In some cases it is not possible to produce speech with the ventilator-delivered air because the air is not allowed to reach the larynx. This occurs when the tracheostomy tube, which is secured in the tracheostoma and provides a connection to the ventilator's tubing, is configured so as to block airflow to the larynx. This is done by inflating a small cuff that surrounds the tube where it lies within the trachea. However, if the cuff is deflated (or if there is no cuff), it is possible to speak using the ventilator-delivered air. Because the air from the ventilator enters below the larynx, speech can be produced during both the inspiratory and the expiratory phase of the ventilator cycle (Fig. 1). During the inspiratory phase, speech production competes with ventilation because the ventilator-delivered air that flows through the larynx to produce speech is routed away from the pulmonary system, where gas exchange takes place (i.e., oxygen is exchanged for carbon dioxide). For this and other reasons, the act of speaking with a tracheostomy and positive-pressure ventilator is challenging, and the resultant speech often is quite abnormal.

Figure 1..  

Inspiration and expiration during positive-pressure ventilation with a deflated tracheostomy tube cuff. (From Hoit, J. D., and Banzett, R. B. [1997]. Simple adjustments can improve ventilator-supported speech. American Journal of Speech-Language Pathology, 6, 87–96, adapted from Tippett, D. C., and Siebens, A. A. [1995]. Preserving oral communication in individuals with tracheostomy and ventilator dependency. American Journal of Speech-Language Pathology, 4, 55–61, Fig. 7. Reproduced with permission.)

Positive-pressure ventilators can be adjusted to meet each individual's ventilatory needs. These adjustments typically are determined by the pulmonologist and executed by the respiratory therapist. The most basic adjustments involve setting the tidal volume and breathing rate, the product of which is the minute ventilation (the amount of air moved into or out of the pulmonary system per minute). These parameters are adjusted primarily according to the client's body size and breathing comfort, and their appropriateness is confirmed by blood gas measurements. Most ventilators allow adjustment of other parameters, such as inspiratory duration, magnitude and pattern of inspiratory flow, fraction of inspired oxygen, and pressure at end-expiration (called positive end-expiratory pressure, or PEEP), among others. How these parameters are adjusted influences ventilation and can also have a substantial influence on speech production.

The speech produced with a tracheostomy and positive-pressure ventilator is usually abnormal. Some of its common features are short utterances, long pauses, and variable loudness and voice quality (Hoit, Shea, and Banzett, 1994). The mechanisms underlying these speech features are most easily explained by relating them to the tracheal pressure waveform associated with ventilator-supported speech. This waveform is shown schematically in Figure 2, along with a waveform associated with normal speech production. Whereas tracheal pressure during normal speech production is positive (i.e., above atmospheric pressure), generally low in amplitude (i.e., in the range of 5–10 cm H₂O), and relatively unchanging throughout the expiratory phase of the breathing cycle, tracheal pressure during ventilator-supported speech production is generally fast-changing (i.e., rapidly rising during the inspiratory phase of the ventilator cycle and rapidly falling during the expiratory phase of the cycle), high-peaked (approximately 35 cm H₂O in the figure), and not always above atmospheric pressure (i.e., during the latter 2 s of the cycle in the figure). These waveforms can also be examined relative to the minimum pressure required to maintain vibration of the vocal folds for phonation (labeled Threshold Pressure in the figure). From this comparison, it is clear that the tracheal pressure associated with normal speech production exceeds this threshold pressure throughout the cycle (expiratory phase), whereas the pressure associated with the ventilator-supported speech production is below the threshold pressure for nearly half the cycle. This latter observation largely explains why ventilator-supported speech is characterized by short utterances and long pauses. The periods during which the pressure is above the voicing threshold pressure are relatively short (compared with normal speech-related expirations) and the periods during which the pressure is below that threshold are relatively long (compared with normal speech-related inspirations). The reason why ventilator-supported speech is variable in loudness and voice quality has to do with the fast-changing nature of the tracheal pressure waveform. The rapid rate at which the pressure rises and falls makes it impossible for the larynx to make the adjustments necessary to produce a steady voice loudness and quality.

Figure 2..  

Schematic representation of tracheal pressure during normal speech production and ventilator-supported speech production. The dashed line indicates the minimum pressure required to vibrate the vocal folds. (From Hoit, J. D. [1998]. Speak to me. International Ventilator Users Network News, 12, 6. Reproduced with permission.)

There are several strategies for improving ventilator-supported speech. One set of strategies is mechanical in nature and involves modifying the tracheal pressure waveform. Specifically, speech can be improved if the tracheal pressure stays above the voicing threshold pressure for a longer portion of the ventilator cycle (to increase utterance duration and decrease pause duration) and if it changes less rapidly and does not peak as highly (to decrease variability of loudness and voice quality). The tracheal pressure waveform can be modified by adjusting certain parameters on the ventilator (such as those mentioned earlier) or by adding external valves to the ventilator system (e.g., Dikeman and Kazandjian, 1995; Hoit and Banzett, 1997). Ventilator-supported speech can also be improved using behavioral strategies. Such strategies include the use of linguistic manipulations designed to hold the floor during conversation (e.g., breaking for obligatory pauses at linguistically inappropriate junctures) and the incorporation of another sound source to supplement the laryngeal voicing source (e.g., buccal or pharyngeal speech).

Evaluation and management of the speech of a client with a tracheostomy and positive-pressure ventilator involves a team approach, with the team usually consisting of a speech-language pathologist working in collaboration with a pulmonologist and a respiratory therapist. Such collaboration is critical because speech production and ventilation are highly interdependent in a client who uses a ventilator. An intervention designed to improve speech will almost certainly influence ventilation, and an adjustment to ventilation will most likely alter the quality of the speech. As an example, a speech-language pathologist might request that a client be allowed to deflate his cuff so that he can speak. Cuff deflation should not compromise ventilation as long as tidal volume is increased appropriately (Bach and Alba, 1990). By understanding the interactions between speech production and ventilation, clinicians can implement interventions that optimize spoken communication without compromising ventilation, thereby improving the overall quality of life in clients who use ventilators.