On-line Supplement
Autotitrating v standard non-invasive ventilation; a
randomised crossover trial.
Jay Jaye, Michelle Chatwin, Mark Dayer, Mary J Morrell
& Anita K Simonds
Materials and Methods
Participants were recruited from the Royal Brompton Hospital Ventilator clinic.
Forty-one patients fulfilled the inclusion criteria. Of these patients 25 were
recruited and 20 successfully completed the protocol (see Figure 1). Inclusion
criteria were: patients > 18 years old with congenital or acquired
neuromuscular disease or chest wall disease, who had nocturnal
hypoventilation defined as nocturnal peak transcutaneous PCO2 (TcCO2) >
6.5 kPa. All patients were experienced ventilator users having been
established long-term on nocturnal NIV for a minimum of six months, and all
had been stable and free from respiratory tract infection/acute exacerbation
for at least three months.
Patients were excluded if they had previous experience using a VPAP device
(VPAP® models, ResMed, Australia) to ensure they were unfamiliar with the
ventilators being compared; or if they had uncontrolled respiratory failure,
defined as daytime resting PaO2 < 7.5kPa and/or PaCO2 > 8.0kPa on air,
uncontrolled heart failure, arrhythmia, moderate or severe bulbar weakness.
Patients were also excluded if inspiratory positive airway pressure (IPAP) on
their current ventilator was > 25cmH2O or they were using a mask that was
incompatible with the autotitrating ventilator being tested, and they were
unable (or unwilling) to change to a mask that was compatible with the
autotitrating ventilator. The study was approved by the Royal Brompton &
Harefield NHS Trust ethics committee (London, UK) and all patients gave
informed, written consent.
The autotitrating NIV has a one-hour learn period from which it calculates
target gross alveolar ventilation (henceforth termed “target alveolar
ventilation” for brevity) and variable back-up respiratory rate (RR), taking into
account estimated anatomical deadspace. During therapy PS (IPAP minus
EPAP) was varied between pre-set minimum and maximum limits to achieve
the calculated target alveolar ventilation.
The algorithm aims to deliver a minimum target gross alveolar ventilation,
calculated during the one-hour learn period and taking into account the
estimated anatomical deadspace initially entered by the clinician using the
formula Dv = 100*(H/1.7)^3, where Dv = deadspace ventilation and H = height
(or arm span) in metres. Gross alveolar ventilation calculated by a low pass
filter is compared to target gross alveolar ventilation and pressure support is
varied within preset limits to increase or decrease ventilation in response to
the magnitude and direction of the difference. Under normal conditions the
ventilator responds to changes in ventilation in around six seconds. If
ventilation falls well below the target the ventilator responds to the changes
more quickly, within four seconds, and increases pressure support at a
greater rate.
The variable back-up respiratory rate moves between two calculated levels; a
timed back-up rate which is the most efficient respiratory rate calculated
during the one hour learn period, and a triggered back-up rate which is 2/3
that of the timed back-up rate. The triggered back-up is functioning whenever
the patient has adequate effort to trigger the ventilator. If the patient fails to
trigger the ventilator the back up rate is gradually increased on a breath-bybreath basis until equal to the timed back-up rate, usually within 5 breaths.
The rate at which it is increased depends on the magnitude of the current
difference between gross alveolar ventilation and the target.
While the patient is conscious and triggering the ventilator the pressure
support and back up rate interfere as little as possible with the patient’s
spontaneous ventilation. It has the ability to respond swiftly to marked
hypoventilation, and slowly to smaller drops in gross alveolar ventilation
making it more comfortable.
Protocol
The trial design was an interventional, randomised, crossover study
comparing NIV administered via an automatically titrated ventilator
(AutoVPAP®; ResMed, Australia) to that administered via a conventional,
expert set-up VPAPIII (VPAPIIIST-A; ResMed, Australia). Both ventilators
were used for one month in random order, patients using the ventilator each
night.
Patients attended hospital for a baseline assessment, which included
measurement of arterialised capillary blood gas tensions, TcCO2, SaO2,
spirometry, non-invasive respiratory muscle strength, and arm span.
Following these measurements patients were randomised to either
autotitrating or standard NIV. The allocation sequence was generated using
twenty sealed, opaque envelopes containing the ventilator name for the first
treatment period (10 AutoVPAP, 10 VPAPIII). At the point of randomisation
an envelope was selected by an independent researcher who did not take
part in NIV set-up or data analysis. The patient was initiated on the first
ventilator and the first one-month treatment period began.
At the end of the first treatment period all baseline measurements were
repeated with the exception of arterial blood gases (see Figure 2). Subjective
tolerance of the ventilator was assessed using visual analogue scales (VAS).
Patients underwent overnight polysomnography including TcCO2
measurement, and 24-hour Holter monitoring, carried out in the sleep clinic or
in the patient’s home, depending on patient preference. If the patient elected
to have polysomnography at home, both studies were conducted in the home,
similarly with patients having in-laboratory studies. Following the first sleep
study, the patient was set up on the second ventilator and began the second
one-month treatment period. At the end of the second treatment period all
measurements were repeated (see Figure 2).
Ventilator Settings
NIV set-ups were performed by a single, trained technician following a
standard set-up procedure.
For the standard NIV treatment period the ventilator was used in
spontaneous/timed mode with IPAP pressure, expiratory positive airway
pressure (EPAP), back-up RR matched as closely as possible to those of the
patient’s usual ventilator. These settings had been established by clinical
adjustment to reduce respiratory effort and maximise comfort at the bedside
followed by overnight monitoring of TcCO2 and SaO2 to ensure optimal
nocturnal blood gas control. For patients whose ventilators had IPAP but no
EPAP, the EPAP on the standard NIV was set to the minimum setting
available on both ventilators (4cmH2O). Appropriate mask type was selected,
IPAP min (minimum inspiratory time) was set to 1.0 second and all other
values were initially set to defaults.
For the autotitrating treatment period, EPAP was manually set as described
for standard NIV, to be matched as closely as possible to those of the
patient’s usual ventilator, minimum and maximum PS values were initially set
to maximise the ability of the ventilator to meet the target alveolar ventilation.
During therapy PS was varied within these limits to maintain a calculated
target alveolar ventilation. Estimated anatomical deadspace was entered
using the formula Dv = 100*(H/1.7)^3, where Dv = deadspace ventilation and
H = height (or arm span) in metres. Appropriate mask type was selected and
all other values were left at defaults.
Target alveolar ventilation and back-up RR were based on the patient’s
breathing over the one-hour learn period, during which the patient sat quietly
with their mask on while breathing 4 cmH2O CPAP. The patient was asked
to remain relaxed but alert and not to doze or talk during this period.
Following set-up on both ventilators patients had at least 15 minutes of
treatment to acclimatize and adjustments were made to settings if patient was
unable to tolerate treatment. During both treatment periods patients could call
if they could not tolerate the settings and adjustments were made to optimize
comfort.
Measurements
Polysomnography: Electroencephalograms (EEG: C3A2, C4A1, O1A2), right
and left electro-oculograms (EOG), and submental electromyogram (EMG),
were recorded using the International 10-20 system of electrode placement [1]
and portable polysomnography equipment (Somnoscreen; SOMNOmedics
GmbH & Co., Kist, Germany). Respiration was monitored using mask
pressure and abdominal plus thoracic effort bands to record movement.
Electrocardiography (ECG), SaO2, leg EMG, and body position were also
recorded. Overnight TcCO2 was measured (TOSCA; Linde Medical Sensors
AG, Basel, Switzerland) and daytime TcCO2 and resting SaO2 were also
measured using the combined earlobe sensor.
The sleep architecture and arousal frequency were analysed using standard
criteria [2,3] by a single investigator blinded to the intervention and the subject
identity. Sleep study data were analysed for the period the patient was in bed
with the ventilator delivering treatment. The mean overnight TcCO 2 and SaO2
were analysed over the total sleep period.
Heart Rate Variability: two-channel Holter monitor (Vista; Novacor, Rueilmalmaison, France) with 200Hz sampling rate, was used in accordance with
the European Society of Cardiology recommendations for heart rate variability
(HRV) recording [4]. Data was edited and analysed according to previously
described protocol [5] using commercial software (Holtersoft; Novacor, Rueilmalmaison, France). Data from a six-hour period at night was used to obtain
high frequency (HF; 0.15 – 0.40 Hz), low frequency (LF; 0.04 – 0.15 Hz), very
low frequency (VLF; 0.003 – 0.04 Hz) power spectral density, and VLF index.
A single investigator analysed all data, blinded to patient identity, intervention
and sleep study results.
Blood Gases: arterialised capillary blood was taken from the earlobe and
analysed using an automated blood gas analyser (Rapidlab 348; Diamond
Diagnostics, Holliston, USA).
Pulmonary Function: measurements were performed using a portable, handheld spirometer (2120; Vitalograph Ltd, Maids Moreton, UK) according to
previously described protocol [6]. Forced vital capacity (FVC) and forced
expiratory volume in one second (FEV1) were measured.
Non-Invasive Muscle Strength: maximal static inspiratory and expiratory
mouth pressure [7] (PI,max and PE,max respectively), maximal whistle mouth
pressure (Pmo,W) [8] and sniff nasal inspiratory pressure (SNIP) [9] were
measured according to a previously defined protocol [10] using a hand-held,
portable, pressure transducer (Morgan Pmax; Morgan Medical Ltd., UK). The
patient sat comfortably and performed each manoeuvre at least 6 times until
they obtained 3 values within 10% of each other, the maximum value was
taken.
Subjective Tolerance: A 10cm VAS was administered at the end of each
treatment period with four questions relating to the comfort and ease of
ventilator use. At the end of the second treatment period a fifth question was
added, asking which ventilator the patient preferred using.
REFERENCES
1. Klem GH, Luders HO, Jasper HH, Elger C. The ten-twenty electrode
system of the International Federation. The International Federation of
Clinical Neurophysiology. Electroencephalogr Clin Neurophysiol Suppl
1999;52:3-6.
2. Rechtschaffen A, Kales A. A manual of standardised terminology,
techniques, and scoring systems for sleep stages of human subjects.
Publication No. 204. Washington, National Institute of Health, 1968.
3. EEG arousals: scoring rules and examples: a preliminary report from
the Sleep Disorders Atlas Task Force of the American Sleep Disorders
Association. Sleep 1992;15(2):173-84.
4. Heart Rate Variability. Standards of measurement, physiological
interpretation, and clinical use. Task force of the European Society of
Cardiology and the North American Society of Pacing and
electrophysiology. Eur Heart J 1996;17:354-381.
5. Vazir A, Dayer M, Hastings PC, McIntyre HF, Henein MY, PooleWilson PA, Cowie MR, Morrell MJ, Simonds AK. Can heart rate
variability rule out sleep disordered breathing in heart failure? Eur Resp
J 2006;27:571 -577.
6. Miller MR, Hankinson J, Brusasco V, Burgos F, Casaburi R, Coates A,
et al. Standardisation of spirometry. Eur Respir J 2005;26(2):319-38.
7. Black LF, Hyatt RE. Maximal respiratory pressures: normal values and
relationship to age and sex. Am Rev Respir Dis 1969;99(5):696-702.
8. Chetta A, Harris ML, Lyall RA, Rafferty GF, Polkey MI, Olivieri D, et al.
Whistle mouth pressure as test of expiratory muscle strength. Eur
Respir J 2001;17(4):688-95.
9. Heritier F, Rahm F, Pasche P, Fitting JW. Sniff nasal inspiratory
pressure. A noninvasive assessment of inspiratory muscle strength.
Am J Respir Crit Care Med 1994;150(6 Pt 1):1678-83.
10. Chatwin M, Ross E, Hart N, Nickol AH, Polkey MI, Simonds AK. Cough
augmentation with mechanical insufflation/exsufflation in patients with
neuromuscular weakness. Eur Respir J 2003;21(3):502-8.
11. Wilson SH, Cooke NT, Edwards RH, Spiro SG. Predicted normal
values for maximal respiratory pressures in caucasian adults and
children. Thorax 1984;39(7):535-8.