Monitoring Long Term Noninvasive Ventilation: Benefits, Caveats and Perspectives

Abstract

Long term noninvasive ventilation (LTNIV) is a recognized treatment for chronic hypercapnic respiratory failure (CHRF). COPD, obesity-hypoventilation syndrome, neuromuscular disorders, various restrictive disorders, and patients with sleep-disordered breathing are the major groups concerned. The purpose of this narrative review is to summarize current knowledge in the field of monitoring during home ventilation. LTNIV improves symptoms related to CHRF, diurnal and nocturnal blood gases, survival, and health-related quality of life. Initially, patients with LTNIV were most often followed through elective short in-hospital stays to ensure patient comfort, correction of daytime blood gases and nocturnal oxygenation, and control of nocturnal respiratory events. Because of the widespread use of LTNIV, elective in-hospital monitoring has become logistically problematic, time consuming, and costly. LTNIV devices presently have a built-in software which records compliance, leaks, tidal volume, minute ventilation, cycles triggered and cycled by the patient and provides detailed pressure and flow curves. Although the engineering behind this information is remarkable, the quality and reliability of certain signals may vary. Interpretation of the curves provided requires a certain level of training. Coupling ventilator software with nocturnal pulse oximetry or transcutaneous capnography performed at the patient's home can however provide important information and allow adjustments of ventilator settings thus potentially avoiding hospital admissions. Strategies have been described to combine different tools for optimal detection of an inefficient ventilation. Recent devices also allow adapting certain parameters at a distance (pressure support, expiratory positive airway pressure, back-up respiratory rate), thus allowing progressive changes in these settings for increased patient comfort and tolerance, and reducing the requirement for in-hospital titration. Because we live in a connected world, analyzing large groups of patients through treatment of "big data" will probably improve our knowledge of clinical pathways of our patients, and factors associated with treatment success or failure, adherence and efficacy. This approach provides a useful add-on to randomized controlled studies and allows generating hypotheses for better management of HMV.

Keywords: chronic hypercapnic respiratory failure; home ventilation; long term mechanical ventilation; monitoring; non-invasive ventilation.

Figure 1

Adherence to NIV. Graphic transcription of ventilator use provided by ventilator software. Y axis: time of the day/night; X axis: days. Each vertical bar represents time spent using the ventilator for a given day. (A) Very regular use suggesting that patient is adherent and comfortable with his/her treatment. (B) Frequent interruptions and irregular use of ventilator suggesting discomfort and/or comorbidity.

Figure 2

Spontaneous vs. controlled cycles. (A) From top to bottom: flow (including leaks), pressure, unintentional leaks (i.e.: difference between total leak flow and estimated flow through the leak valve (small holes) of the mask), total leaks, and respiratory rate. Seventy six-year-old female subject with kyphoscoliosis, obstructive sleep apnea and obesity; ventilator settings: ST mode (spontaneous/timed); IPAP (Inspiratory positive airway pressure): 30 cmH₂O; EPAP: 10 cmH₂O (Expiratory positive airway pressure); Back-up respiratory rate (BURR): 18 cycles/min. Nasal mask (facial mask not tolerated). Flow tracing shows intermittent flattening of inspiratory curve suggesting persisting flow limitation (upper airway collapse to be compensated by increasing EPAP). Blue arrow: pressure tracing shows vertical marks associated with each cycle indicating controlled cycle (i.e., delivered by ventilator). On this segment, patient is continuously on back-up respiratory rate. Low level of intentional leaks. (B) From top to bottom: flow (including leaks), pressure, total leaks, and respiratory rate. Fifty four-year-old male subject with restrictive disorder. Ventilator settings: ST mode (spontaneous/timed); IPAP (Inspiratory positive airway pressure): 16 cmH₂O; EPAP: 5 cmH₂O (Expiratory positive airway pressure); Back-up respiratory rate (BURR): 16 cyc/min; nasal mask. Blue arrow: as opposed to (A), all cycles are triggered by the patient (i.e., spontaneous). Normal aspect of flow and pressure tracings. DirectView software, Philips Respironics.

Figure 3

Upper airway obstruction under NIV. 65-year-old male subject, obesity-hypoventilation with obstructive sleep apnea syndrome (OSAS). Bi-level pressure support ventilator, ResScan software, ResMed. Facial mask. Ventilator settings: ST Mode (spontaneous/timed), IPAP (Inspiratory positive airway pressure): 20 cmH₂O; EPAP (Expiratory positive airway pressure): 10 cmH₂O; BURR (Back-up respiratory rate): 16 cycles/min. From top to bottom: flow, unintentional leaks (i.e.: difference between total leak flow and estimated flow through the leak valve of the mask), tidal volume, pressure. 5 min window. Red line on leaks tracing (24L/min) is a threshold value suggested by manufacturer for upper limit of acceptable leaks (see text for comments). Event A: marked decrease in flow with tracing suggesting increase in upper airway resistance (leading to intermittent complete obstruction); simultaneous decrease in tidal volume without increase in leaks. Event B: sudden transient resumption of flow with a simultaneous increase in tidal volume. Increase in upper airway resistance could be related to an insufficient “pneumatic splint” effect in spite of rather high insufflation pressures, or to glottic closure (further characterization would require respiratory polygraphy). In a patient with a known OSAS, a pragmatic trial of increasing EPAP is an option. Value of IPAP should be increased accordingly to maintain same level of pressure support (if tolerated). Use of a facial mask may also contribute to these events, and may be replaced by nasal mask with chin strap if tolerated.

Figure 4

Patient ventilator asynchrony (PVA): illustrative ventilator tracings. One-minute windows. From top to bottom: pressure, flow, tidal volume, respiratory rate and unintentional leaks (i.e., without intentional leak through exhalation valve of mask). Rescan software, ResMed. (A) 74-year-old male subject with severe COPD (FEV1: 20% of predicted). Bi-level pressure support ventilator, Rescan software, Resmed. Ventilator settings: ST (spontaneous/timed) mode; IPAP (Inspiratory positive airway pressure): 30 cmH₂O; EPAP (Expiratory positive airway pressure): 7 cmH₂O; BURR (back-up respiratory rate) 20 cycles/min. Red arrows show intermittent double-triggering. In this case, leaks are not in cause. Possible causes are dysfunction of ventilator, too high inspiratory trigger sensitivity, too short minimal inspiratory time (TI_MIN) with prolonged inspiratory efforts. Adjustments of settings (if required) to be considered are: to increase TI_MIN; to adapt cycling sensitivity (at a lower percentage of peak inspiratory flow, which delays cycling); to increase IPAP and reduce rise time; or to decrease sensitivity of inspiratory trigger. (B) 67-year-old male subject with obesity hypoventilation and severe OSAS. Bi-level pressure support ventilator, Rescan software, Resmed. Ventilator settings: ST (spontaneous/timed) mode; IPAP (Inspiratory positive airway pressure): 21 cmH₂O; EPAP (Expiratory positive airway pressure): 14 cmH₂O; BURR (back-up respiratory rate) 16 cycles/min. Red arrows show low amplitude repeated increases in flow and pressure which represent unrewarded efforts (i.e., inspiratory efforts by the patient which do not trigger the ventilator). Among possible causes are: inappropriate setting of inspiratory trigger sensitivity, increase in upper airway resistance, leaks, intrinsic PEEP (Positive end expiratory pressure), decrease in inspiratory muscle function. For all PVA, control of leaks is mandatory before adjusting other settings.

Figure 5

Impact of major leaks. Seventy five-year old male subject with severe COPD (GOLD D; FEV1: 17% of predicted). Bi-level pressure support ventilator, S/T mode (spontaneous/timed); IPAP (Inspiratory positive airway pressure): 24 cmH₂O; EPAP (Expiratory positive airway pressure): 4 cmH₂O; Back-up respiratory rate (BURR): 18 cycles/min. Facial mask. Five-minute window, Rescan software, ResMed. (A) normal tracing albeit for a few cycles with decrease in flow; (B) Vertical arrow marks appearance of major leaks (could be related to transient displacement of interface). Pressure tracing shows episodes of auto-triggering, and double triggering. Marked drop in pressure, flow and V_T (explained by magnitude of leaks). (C) As leaks progressively decrease, breathing pattern becomes more regular; pressure and flow increase progressively.

Figure 6

Proposed algorithm for a systematic approach to monitoring of patients under LNIV and analysis of ventilator software. *: the reader is referred to references 4 and 8 for further details.