A current review of airflow monitoring devices used during polysomnography.

When faced with the task of identifying sleep-disordered breathing during polysomnography, most sleep clinicians will agree that the airflow parameter is clearly the most critical piece of data available. In fact, the standard definitions of the most common sleep-disordered breathing events include apnea and hypopnea and referring to amplitude criteria of the airflow signal.

Since the advent of sleep-disordered medicine, airflow monitoring methodology has consisted of four devices including thermocouple, thermister, and, for the last 3 to 4 years, nasal pressure and polyvinylidene fluoride film (PVDF).

Thermocouple
Thermocouples are composed of two wires made of dissimilar metals, fused together at the tip. The fusion of the dissimilar metals creates a small voltage. When the tip of the two metals is placed under a patient’s nose or in front of the mouth, the temperature changes associated with the patient’s breathing cause the diameter of the wires to expand and contract, thereby increasing and decreasing the voltage being produced by the device. These changes in voltage are recorded by the polysomnograph (PSG).

A clear advantage of the thermocouple is that it is very easy to combine multiple “fusion points” in a circuit allowing for a combined oral/nasal signal. Additionally, the low cost of the materials required to construct a thermocouple, as well as the absence of an external power source requirement, results in a very inexpensive monitoring device.

A significant disadvantage of the thermocouple is its slow response to changes in temperature. This slow response, which can be more than 1 second in some instances, is due to the thermal mass of the thermocouple itself. The material used to construct the thermocouple and the insulating material covering the thermocouple must change temperature in order to generate a voltage change. It is the mass of the material that impacts the time to respond to the temperature change. As a practical example, consider that a piece of aluminum foil in a campfire will heat up and cool down substantially faster than the cast iron oven in the same campfire.

When attempting to monitor airflow during nasal CPAP (nCPAP) titration in the sleep laboratory by placing a thermocouple device under a nasal mask, many commercially available combination oral/nasal thermocouple designs cause uncorrectable mask leaks. Additionally, higher pressures of nCPAP (14 cm H2O and higher) result in a reduced amount of intramask temperature variance due to the nCPAP pressure flow, thereby causing deterioration of the airflow signal.

Thermister
A thermister produces a very similar signal to that of a thermocouple. Thermistors are composed of a material that changes electrical resistance when exposed to temperature changes. As with the thermocouple, the temperature changes are sampled under a patient’s nose or in front of the mouth. Thermistors require that a current be generated and conducted through the material in order to measure a voltage across the resistance change of the material.

(For a list of thermister advantages and disadvantages, see Table 2.)

The general voltage of the thermister output can often be higher than that of a conventional thermocouple. This can be advantageous with regard to reducing or eliminating extraneous electrical noise from a tracing, as lower polygraph amplifier gain settings can be used.

A disadvantage of the thermister is the increased cost of the device, which is largely due to the external power source and circuitry. Most PSGs designed for thermistor input provide the current source inside the machine. Thermistors used with PSGs that were not specifically designed for them must have an external current source that is usually supplied by a battery and circuitry in the cable between the sensor and the PSG.

Additionally, as with the thermocouple, the slow response time due to the device’s thermal mass is a clear disadvantage. The slow response time of thermistors and thermocouples is the reason that waveforms produced by these technologies are a smoothened average of the actual changes in airflow temperature that occur as the patient inhales and exhales. Subtle sleep-disordered breathing, such as the flow limitations associated with respiratory effort-related arousals (RERAs), cannot be detected by thermistors or thermocouples. In a study comparing the performance of thermistors and nasal pressure monitoring, researchers showed that at a respiratory disturbance index (RDI) of five events per hour, 18% of patients would have been missed using thermistors alone and 10% of patients would have been missed with thermistors at an RDI of 15 per hour.

Nasal Pressure Monitoring
A nasal pressure monitoring (NPM) device incorporates an air pressure transducer connected to a nasal cannula or oral/nasal cannula. This device responds very quickly to changes in pressure and hence does not have the slow response time associated with the thermal mass of thermistors and thermocouples. The waveforms produced when using NPM are not average representations of the patients’ actual airflow. The waveform produced may often be described as “bumpy or noisy,” when in fact the waveform displayed is a reasonably accurate representations of the air pressure changes associated with breathing... most of the time. Similar to a thermister, the NPM device provides a relatively strong signal requiring less polygraph amplifier gain, and can easily be connected to a standard supplemental oxygen port found on a conventional nCPAP mask; this allows for an easy method of “in-line” airflow monitoring during nCPAP titration in the laboratory, without the patient discomfort and mask leaks experienced with the thermocouple and thermister.

(For a list of NPM advantages and disadvantages, see Table 3 on page 45.)

Most NPM devices are not linear. This disadvantage results in the device exaggerating both apnea and hypopnea waveform changes. A 50% reduction in airflow pressure will be displayed as a 75% reduction in waveform amplitude on the PSG. This results in many more apneas being scored than would have been with a thermal monitoring device (see Figure 1 on page 45).

Figure 1. Linearity comparison of a thermal sensor and nasal pressure.

Nasal pressure monitoring devices have also been shown to miss respiratory events during mouth breathing, even when using a combination nasal/oral cannula. A nasal cannula itself, when inserted into the nares, increases flow resistance. Additionally, increased cost can be experienced due to disposable cannula requirements, as well as the external power source requirement resulting in periodic battery replacement.

Polyvinylidene Fluoride
PVDF is a treated plastic film that is polarized with an electric charge. The film is sensitive to both temperature changes and vibration, and can output as much as 25 volts. It has been in use for many years in other applications including LED lighted tennis shoes, automobile motion detectors, traffic light sensors, and coin-operated vending machines.

In order to document the difference in linearity between NPM and PVDF, the author compared the results of a PSG study while simultaneously monitoring nasal pressure and the PVDF sensor on 10 patients with suspected OSA and undergoing a scheduled polysomnography.

The results indicated that in 10 out of 10 patients, the PVDF airflow sensor revealed many respiratory events consistent with hypopnea, which were presented by the NPM to be apnea. In other words, the nasal pressure device exaggerated both apnea and hypopnea events as compared to the PVDF, thus confirming the nonlinearity of the NPM. Additionally, PVDF and the NPM provided nearly identical results for the sum of apnea and hypopnea events, thereby confirming the comparable response time of the two devices (see Figure 2, above).

Figure 2. A comparison of the apnea scoring ability of PVDF and NPM.

Regarding the response time of PVDF film to changes in temperature, the output remains constant up to approximately 200 Hz, which is 0.005 seconds response time. This compares to a response time of about 2.5 seconds for thermocouple and thermister based sensors. As with NPM, this fast response time of PVDF may also produce a waveform that can be described as “bumpy or noisy,” resulting in the need for more aggressive high-pass filtering to achieve the smoother waveform clinicians are most used to viewing. Conversely, this fast response time allows PVDF as well as NPM the ability to display subtle changes in airflow, such as those associated with RERAs and flow limitation, which thermocouples and thermistors cannot detect.

(For a list of PVDF advantages and disadvantages, see Table 4.)

A RERA comparison example (see Figure 3) demonstrates that the PVDF sensor’s response time is comparable to the NPM as measured by the ability to display waveform changes resulting from RERAs. During a number of events, the PVDF sensor actually demonstrated a higher sensitivity than the NPM, as indicated by the presence of cardiac oscillations in the PVDF signal, which was absent in the nasal pressure signal (see Figure 4).

Figure 3. RERA waveform display comparison of PVDF and nasal pressure device.	Figure 4. Cardiac oscillations present in the PVDF signal.

In light of the wide-ranging variables associated with currently available airflow monitoring devices, sleep-disorder technologists and clinicians should not only become familiar with the differences in the devices, but also determine which device will best meet the needs of the patients studied within their respected testing program.

Todd Eiken, RPSGT, is director of the Metropolitan Sleep Disorders Center, St Paul, Minn; he is also a member of Sleep Review’s Editorial Advisory Board.