(Associate Editor: Mary Morell).
Polysomnographic evaluation of non-invasive ventilation in children with neuromuscular disease
Article first published online: 23 DEC 2013
© 2013 The Authors. Respirology © 2013 Asian Pacific Society of Respirology
Volume 19, Issue 1, pages 80–84, January 2014
How to Cite
Falsaperla, R., Wenzel, A., Pavone, P., Di Mauro, C. and Vitaliti, G. (2014), Polysomnographic evaluation of non-invasive ventilation in children with neuromuscular disease. Respirology, 19: 80–84. doi: 10.1111/resp.12194
- Issue published online: 23 DEC 2013
- Article first published online: 23 DEC 2013
- Accepted manuscript online: 13 SEP 2013 03:22AM EST
- Manuscript Accepted: 1 JUL 2013
- Manuscript Revised: 18 JUN 2013
- Manuscript Revised: 17 JUN 2013
- Manuscript Revised: 19 MAY 2013
- Manuscript Revised: 19 MAY 2013
- Manuscript Revised: 21 MAR 2013
- Manuscript Revised: 19 MAR 2013
- Manuscript Received: 7 JAN 2013
- blood gas analysis;
- childhood polysomnography;
- neuromuscular disease;
- non-invasive ventilation efficacy
Background and objective
In the last 20 years, research efforts have been focused on the use of non-invasive ventilation (NIV) as a mean of avoiding tracheostomy in patients affected by neuromuscular diseases (NMD). Nocturnal NIV has been a particular focus as sleep is a risk factor for respiratory failure in NMD patients. The objective of our study was to evaluate the efficacy of nocturnal NIV in improving the respiratory function of NMD patients evaluated by polysomnography (PSG) and arterial blood gas (ABG) analysis parameters.
Ten children affected by NMD underwent PSG and ABG analysis evaluation at the onset of their respiratory failure and during nocturnal NIV therapy.
We found a statistically significant improvement of the lowest oxygen desaturation (nadir SaO2), apnoea–hypopnoea index (AHI) and oxygen desaturation index (ODI) after NIV treatment in all patients. Mean SaO2 also improved, although this result was not statistical significant, while the percentage of episodes of desaturation with a SaO2 <90% and <80% decreased with a statistical significance (P < 0.0001).
After NIV, only one patient showed an episode of desaturation lasting more than 5 min (10.6 min length), and we also found an improvement of daytime blood gas parameters with a normalization of these indexes.
NIV was effective in improving respiratory parameters at night in patients affected by respiratory muscular weakness, as evaluated by PSG and ABG analysis.
arterial blood gas
oxygen desaturation index
In the last 20 years, research efforts have been focused on the use of non-invasive ventilation (NIV) as a means of avoiding tracheostomy in patients affected by neuromuscular diseases (NMD).
Sleep-disordered breathing, linked to muscle weakness, commonly affects patients with NMD. Mechanisms leading to these respiratory disorders involve reduction of alveolar ventilation (as normally occurs during sleep due to a fall in central respiratory drive), reduced activity of respiratory muscles and blunted arousal thresholds.[2-5] Moreover, during rapid eye movement sleep (when a maximal muscle hypotonia is present), a disproportionate loss of upper airway muscle tone and diaphragm weakness that add additional loads have been demonstrated. All these events and the respiratory muscle weakness that characterizes NMD predispose these patients to sleep-disordered breathing with desaturations, hypopnoeas, obstructive and central apnoea, and hypercapnic hypoventilation, mostly occurring during rapid eye movement sleep.[4-7]
Usually, SBD manifests with hypercapnia as a signal of a critical imbalance between ventilatory demand and respiratory capacity to compensate. Hypercapnic hypoventilation, as a consequence of nocturnal respiratory failure, contributes to the development of daytime respiratory failure. In order to monitor the onset of this daytime respiratory failure, continuous follow up of arterial blood gas (ABG) analysis at daytime should be considered[8, 9] because, in this patients group, an altered activity of respiratory muscles with increased upper airway resistance and reduced central sensitivity to changes in blood gases[11-13] have been observed.
Considering these characterizing features in patients with NMD, NIV for nocturnal use has proved to be efficient in relieving symptoms and improving daytime blood gases.[9-11] Bye et al. found significant correlations between the lowest oxygen saturation (nadir SaO2) during rapid eye movement sleep and daytime respiratory parameters, such as vital capacity, decrease in vital capacity from the erect to the supine position, oxygen tension (PaO2) and carbon dioxide tension (PaCO2). However, a limitation of their study was the low case number that led to the conclusion that a sleep study with respiratory monitoring is required.
Moreover, several studies have shown that sleep fragmentation causes daytime sleepiness and cognitive impairment,[12, 13] necessitating the definition of parameters to demonstrate the efficacy of NIV in improving NMD respiratory function.
In this context, polysomnography (PSG) and arterial ABG analysis appeared the most appropriate measurements to evaluate NIV efficacy in improving respiratory failure in NMD patients, more so if we consider that there is no consensus on the inclusion of sleep parameters in diagnostic recordings of NMD.
The objective of our study was to evaluate the efficacy of 12 h of nocturnal NIV in improving NMD respiratory function as measured by PSG and ABG analysis before and after treatment.
Patients with progressive NMD, followed up in our outpatient clinic for evaluation of respiratory function, were considered for inclusion. We evaluated all patients with the functional disability scale (with values ranging between 0, when functional abilities were normal, to 8 in case of children confined to bed, without any possibility to move with a wheelchair), and the results of this evaluation showed a mean score of 7. The Review Board of University of Catania gave to all authors the consent to access patients' data.
Ten patients, mean (standard deviation) age 3.11 (2.08) years, were evaluated on clinical grounds, that is, symptoms and signs of restrictive respiratory dysfunction, for eventual initiation of NIV therapy. The reported symptoms were impaired consciousness and/or developmental delay, daytime fatigue, sleep fragmentation, daytime dyspnoea and impairment of concentration. These patients, presenting with respiratory failure, cyanosis, cough, nasal breath, multiple nocturnal awakenings and evidence of respiratory muscles fatigue, were considered for NIV and were admitted for a 5-day hospital stay. On admission, they underwent a PSG evaluation and ABG analysis examination, which were repeated after the 12-h NIV therapy at night. The respiratory parameters were recorded by specialized physicians.
We considered nadir SaO2, apnoea–hypopnoea index (AHI) and oxygen desaturation index (ODI), mean SaO2, and the percentage of episodes of SaO2 <90% and <80% in order to evaluate the efficacy of NIV in improving patients' respiratory outcome at night. Moreover, we evaluated BGA before and after the therapy, comparing these parameters with PSG indexes.
A four ExG channel with a paper speed of 1 cm/s was used for the polysomnographic recordings (Embletta X100, Natus Medical Inc., Frankfurt, Germany), which included an airflow (three-port thermistor), respiratory movements of chest and abdomen (Siemens Sensor 230, Siemens Elema Inc., Stockholm, Sweden) and body movements (BioMatt, Duorec Inc., Turku, Finland) sensors.
Sleep was scored according to Rechtschaffen and Kales guidelines. Apnoea was scored according to convention. A hypopnoea was scored when there was a decrease of airflow ≥50% followed by an arousal independent of oxygen levels. The AHI was defined as the number of events per hour of sleep.
The ODI was calculated as the mean number of oxygen desaturations ≥4%/h of sleep. A nadir SaO2 below 90% was considered pathological. Hypoventilation was defined as partial pressure of end tidal (pet) CO2 levels exceeding 6.5 kPa (49 mm Hg). Continuous CO2 monitoring was obtained by a transcutaneous CO2 sensor. In those patients when the paCO2 recording was not technically acceptable, a smooth decrease in SaO2 below 90% for a period of 3 min or more was regarded as hypoventilation. Motion artefacts were detected in the oximetry recordings by mean of body movement recording.
The children in the study group were started on NIV at admission and after a first PSG exam. NIV was performed according the standard procedure in children with daytime PCO2 > 45 mm Hg and symptoms of night hypoventilation (daytime sleepiness and sleep disturbance with frequent arousals). The backrest was lifted to an angle of 45°, or the patient was put in a lateral decubitus if unable to sustain a semirecumbent position. A facial mask of adequate size was put in place, and bi-level NIV started. Initially, the facial mask was held gently in place by hand and then tightened with tight-fitting securing system when tolerated. Hydrocolloid dressing was used to prevent pressure sores at the bridge of the nose and other points of pressure. The alarm was set monitoring O2 saturation. NIV was applied using the following ventilator: Airox Legend Air (Linde Healthcare, Emden, Germany). The parameters we choose to start NIV were IPAP 12 and EPAP 4 cmH2O, FiO 2% 1 L/min, Tv: 5–7 mL/kg, RR 25 b.p.m. These parameters were then changed for each patient according to his or her respiratory symptoms at night, PSG and ABG analysis results after the first night on NIV. The child starting NIV was hospitalized for 5 days, and for the first three nights, we performed NIV gradually, starting with 3-h cycles, in order to allow the child to get accustomed to the method. On the fourth night, we performed NIV for the whole night, and the following night, we evaluated PSG parameters.
Values are expressed as mean ± standard deviation. Changes observed following NIV administration were compared using paired Student's t-test and chi-square test when appropriate. The Spearman's rank correlation coefficient was used for correlation calculations between ODI and AHI, and nadir SaO2 and PCO2 and HCO3−. For statistical evaluation, a normal distribution of data was considered.
P < 0.05 was considered as statistically significant.
We enrolled 10 consecutive patients, followed for NMD in our Paediatric Department, AOU Policlinico-OVE, University of Catania, Italy. Demographic data are summarized in Table 1. All patients were affected by respiratory impairment and apnoeic/hypopnoea episodes diagnosed by a PSG exam upon admission.
|Number of patients||10|
|Gender||F = 7; M = 3|
|Age (mean ± SD)||3.11 ± 2.08|
Type VI collagenopathy (Bethlem syndrome) = 1
Steinert myotonic muscular dystrophy = 2
Malformation syndrome with severe mental retardation and hypotonia = 2
ROHHADNET syndrome = 1
Spinal muscular atrophy 1 = 2
Central core myopathy = 2
PSG at admission showed that all patients had at least one episode of desaturation lasting more than 5 min (mean length: 25.80 min; maximum length: 43 min), and BGA showed the presence of hypercapnia and hypoxia (Table 2). After NIV, we found an improvement of daytime blood gas parameters, with a normalization of these indexes (Table 2). We found a significant improvement of the nadir SaO2, AHI and ODI after NIV treatment in all patients (Table 3).
|ABG pre-NIV (mean ± SD)||ABG post-NIV (mean ± SD)||P < 0.05a|
|pH||7.3 ± 0.06||7.4 ± 0.04||0.006|
|PCO2 (mm Hg)||53.7 ± 8.6||42.4 ± 4.4||0.002|
|HCO3− (mm Hg)||30.8 ± 2||25.3 ± 7.8||0.05|
|PO2 (mm Hg)||38.2 ± 6.8||46.3 ± 3.5||0.05|
|Polysomnography at diagnosis||Polysomnography in NIV||P < 0.05|
|Nadir SaO2 (%) (mean)||72.9||89.5||P < 0.05a|
|AHI (index/h) (mean ± SD)||9.4 ± 5.8||2.3 ± 3.7||P < 0.004b|
|ODI (index/h) (mean ± SD)||20.7 ± 13.9||2.1 ± 2.3||P < 0.0001b|
|Medium SaO2 (%) (mean)||93.1||94.2||NS|
|SaO2<90% (percentage of episodes) (mean)||17.4||0.6||P < 0.0001b|
|SaO2<80% (percentage of episodes) (mean)||3.3||None||P < 0.0001b|
|SaO2<90% > 5 min (number of episodes) (mean)||One in all patients (mean length: 25.8 min; max length: 43 min)||Only in one patient (maximum length: 10.6 min)|
Hypoventilation was found in all enrolled patients as expression of a mean petCO2 of 57 mm Hg that considerably decreased after NIV to a mean of 40 mm Hg. Mean SaO2 also improved in all patients, although not significantly, while the percentage of episodes of desaturation with a SaO2 <90% and <80% decreased with a high statistical significance (P < 0.0001; Table 3). Before NIV, episodes of SaO2 <90% showed a mean length of 25.80 min, with a minimum length of 5.5 min and a maximum length of 43 min. After NIV, we found that only one patient showed an episode of desaturation lasting more than 5 min (10.6-min length), while the other patients did not show any episodes of SaO2 <90% lasting more than 5 min (Table 3).
ODI was compared with AHI simultaneously measured with PSG records. The Spearman's correlation between variables from PSG with ODI and AHI was analysed. We did not find any statistically significant difference in sensitivity between ODI and AHI.
The main results of the present study show that PSG is an effective method for demonstrating the efficacy of NIV in the treatment of respiratory failure in patients affected by NMD.
In NMD patients, PSG is necessary to recognize altered respiratory function, phenotypically expressed as sleep-disordered breathing.[18-24] The particular PSG pattern in NMD is characterized by a slow and progressive decrease of SaO2, followed by weak attempts of thorax and abdomen muscles to re-establish a normal oxygen saturation.[23-25]
In this context, nocturnal ventilatory support is important to establish an efficient respiratory outcome in NMD patients, even if until now there is few literature data on the evaluation of the success of NIV in improving respiratory outcomes in a short time. For this reason, as PSG is considered the ‘gold standard’ to diagnose SBD in NMD patients, it could represent also the main technique to evaluate NIV efficacy in improving night breathing in these patients.
By monitoring sleep breathing in ventilated NMD, in our study we found statistically significant differences among nadir SaO2, AHI and ODI when comparing the same parameters before commencing NIV and after NIV therapy (Table 3). Moreover, the main respiratory disturbance was a low nadir SaO2 and the frequent occurrence of hypoventilation.
Recently, Weinberg et al., in contrast to Labanowski et al., who performed a home study PSG on NMD patients, found obstructive apnoea and hypopnoeas to be far more common than central apnoea and hypopnoeas. However, there is a risk for underestimating the occurrence of obstructive events in favour of central events when using non-invasive monitoring of respiratory movements; this risk is particularly high in patients with weak respiratory muscles.
In our study, there was no statistically significant difference between the sensitivity of ODI and AHI in diagnosing desaturation events, and both increased with a statistical significance after NIV therapy. A different result was found in the study of Weinberg et al. where the authors concluded at the possibility to underestimate the occurrence of hypopnoeas because the definition of a hypopnoea was based on the reduction of airflow in combination with a following arousal irrespective of the oxygen desaturation.[29, 30] This might also have favoured, in Weinberg's study, the more frequent recognition of obstructive hypopnoeas compared with central hypopnoeas, as obstructive hypopnoeas may cause arousals more frequently. In our study, this comparison was not statistically significant, but this result could be debatable as we enrolled a relative small number of cases. A further study with a higher number of patients should be performed in order to evaluate the different sensitivity of ODI and AHI, and the singular importance of these parameters in detecting apnoeas in NMD patients.
Similar to what was found by Weinberg et al. and in contrast to the results of Bye et al., we found neither a significant correlation between nadir SaO2 and PCO2 nor between HCO3− and nadir SaO2. Probably because ABG analysis evaluation was limited to the particular moment when the sample was taken, while an extended evaluation of these parameters would have given different results.
To our knowledge, in literature, there are few data quantifying the efficiency of NIV in NMD, and no data are published in such young children. PSG and ABG analysis seem to be valid methods, even if not the only methods, available to monitor the efficacy of NIV in NMD patients. In our study, NIV was effective in improving respiratory parameters at night in NMD patients with nocturnal apnoea/hypoponea. PSG and ABG analysis were efficient in evaluating the improvement in the respiratory function. Thus, we suggest PSG monitoring in all patients affected by respiratory muscular weakness and receiving NIV in order to assess the efficacy of NIV.
- 17British Thoracic Society guideline for respiratory management of children with neuromuscular weakness. Thorax 2012; 67(Suppl. 1): 11–40., , et al.
- 21A Manual of Standardized Terminology, Techniques and Scoring System for Sleep Stages of Human Subjects. BIS/BRI, UCLA, Los Angeles, CA, 1968., .