Assessment of Nocturnal Hypoventilation by Different Methods and Definitions in Children with Neuromuscular Disease: Oxycapnography and Blood Gas Analysis.

OBJECTIVE: To investigate whether partial arterial carbon dioxide pressure (PaCO 2 ) level in arterial blood gas analysis that was used to predict nocturnal hypoventilation (NH) is concordant with nocturnal end-tidal CO 2 (PetCO 2 ) measurement obtained by a noninvasive method of oxycapnography in children with neuromuscular disease (NMD).
METHODS: Twenty-one patients aged 6-18 years with a confirmed diagnosis of NMD were enrolled. Each patient underwent a nocturnal oxycapnography study using an orinasal probe and a pulse oximetry finger probe to record PetCO 2 , oxygen saturation (SpO 2 ), pulse rate, and respiratory rate. Arterial blood gas analysis was performed to record PaCO 2 levels on three occasions at night (23:00 pm, 03:00 am, 07:00 am).
RESULTS: The mean overnight PaCO 2 level of the three blood gas analyses (mean PaCO 2 noct) was 41.78±4.69 mmHg. A significant change was observed between mean PaCO 2 23:00 and PaCO 2 07:00 levels (p=0.032). There was no significant difference between PaCO 2 , PetCO 2 , and SpO 2 levels in the NMD group. The interclass correlation coefficient between PaCO 2 07:00 and PetCO 2 levels was 0.791 (95% CI: 0.533-0.923); the interclass correlation coefficient between overnight mean PaCO 2 and PetCO 2 levels was 0.811 (95% CI:0.533-0.923).
CONCLUSION: Our study indicates that nocturnal PetCO 2 and PaCO 2 levels were statistically comparable but the use of PaCO 2 alone is not adequate to make an early diagnosis of NH in NMD. There is a need for making more restrictive definitions for NH, and conducting studies with larger study populations to reach an agreement on the best definition of hypoventilation, and updating consensus guidelines. © Copyright Istanbul Medeniyet University Faculty of Medicine.

Introduction

Patients with neuromuscular disease (NMD) will not only experience parenchymal lung diseases such as recurrent pneumonia and atelectasis but also nocturnal sleep-disordered breathing disorders as their respiratory muscle weakness progresses[1]. In this group of patients, respiratory system muscles are also affected in addition to skeletal muscles. Patients can avoid hypoxemia and hypercarbia via controlled hyperventilation when awake. In sleep, however, they become vulnerable to hypoventilation and hypercarbia associated with shallow ventilation and reduced tonus of upper respiratory tract and intercostal muscles. Neuromuscular weakness will worsen nocturnal respiratory problems in sleep, and thus will lead to hypercapnia and hypoxemia that frequently emerge at night before daytime ventilation disorders become evident[2].

In children with NMD, the prevalence of sleep-disordered breathing (SDB) disorders (hypoventilation, hypoxemia, central or obstructive apnea-hypopnea) was reported as high as 30-70%[3,4]. As there is no large-scale studies, the incidence varies by different study centers, children’s age range, NMD type, the diagnostic method used, and even differing criteria used in the diagnostic methods.

As symptoms and signs alone are not sufficiently reliable for diagnosing sleep-disordered breathing (SDB) and as polysomnography (PSG) has limited practicality and availability, various researchers have thought various alternative diurnal and nocturnal screening tests to predict SDB. Furthermore, daytime respiratory function tests do not reliably identify patients with SDB in the absence of hypercapnia[5,6]. Thus, screening tests are needed to diagnose SDB before the development of signs of daytime hypercapnia among children with neuromuscular weakness.

Although in the literature PSG is considered the gold standard test for diagnosing and classifying SDB and determining its severity, there is an ongoing debate as to whether or not routine monitoring of the CO2 level is performed in most sleep laboratories. Therefore, CO2 measurements have been recommended as a marker of airway obstruction and quantitative measurement of alveolar hypoventilation in all children[7,8]. However, there has been no consensus yet on a validated definition of nocturnal hypoventilation (NH) in children. A study where PSG, transcutaneous CO2 (PtcCO2), and end-tidal CO2 (PetCO2) levels were compared in 609 patients, mean PtcCO2 was found above 50 mmHg in 11.9% of 318 patients who had an apnea-hypopnea index (AHI) below 59. Additionally, nocturnal PtcCO2 level was found to increase independently of the severity of the obstructive sleep apnea (OSA) due to respiratory muscle hypotonicity among children with Down syndrome[10]. Hence, as polysomnography is not readily available for children in our country and as increased PCO2 levels are detectable independently of AHI, NH screening using a simpler method will be more practical and feasible.

Capnography is a noninvasive method that enables continuous monitoring of a patient’s ventilatory status[11]. We hope that it will guide us in early diagnosis and management. Hence, we performed an overnight recording and analysis of end-tidal carbon dioxide (PetCO2) level and measured oxygen saturation (SpO2) in combination with respiratory and pulse rates using an oxycapnography device in order to screen nocturnal breathing disorders (hypercarbia, hypoxia). So far, we used to check the PaCO2 level to predict nocturnal hypoventilation by performing blood gas analyses at our clinic (on three occasions: 23:00 pm, 03:00 am, and 07:00 am). Our objective was to investigate whether PaCO2 levels in blood gas analysis we use to predict hypoventilation and nocturnal PetCO2 levels measured using the noninvasive method of oxycapnography would show concordance in children with NMD. The literature review revealed that these two methods were compared in healthy subjects and different patient groups but they were not used to screen sleep-disordered breathing in children with NMD. As there is no confirmed NH level associated with end-organ damage in children, we made the comparison in accordance with the American Academy of Sleep Medicine (AASM) 2019 criteria and at lower CO2 levels. We assessed hypoventilation by different “hypercarbia” levels and “nocturnal hypercarbia indexes”.

Material and Methods

Patients: This study enrolled 21 patients aged 6-18 years who were followed up at Istanbul Medeniyet University Pediatric Neurology outpatient clinic with genetically confirmed diagnosis of pediatric-onset NMD. Oxycapnography was performed on patients between April 2018 and December 2018. Parents of all patients gave their written informed consent for participation in the study by signing the “informed consent form for study participation”. The local ethics committee of our hospital approved the study with a report number of 2018/0277.

Patients with signs of acute respiratory failure, acute pulmonary parenchymal disease (pneumonia, bronchitis, etc.), primary congenital heart disease, and chronic lung disease were excluded. Patients in the early onset muscle disease group like spinal muscular atrophy were also excluded. Patients with any of the following conditions including “adenoid-tonsil hypertrophy, micrognathia, macroglossia, septal deviation, or nasal deformity” that could cause elevated CO2 levels via “mechanical upper airway obstruction” apart from the primary disease detected in physical examination were also excluded. Patients were questioned about signs or symptoms likely indicating hypercarbia and hypoxia, such as snoring, restless sleep, waking up suddenly, morning dizziness, and morning headache, and those with any of these conditions were excluded. Thus, by recruiting patients without primary pulmonary parenchymal disease, mechanical causes of OSA, and clinical signs of nocturnal hypoventilation, we had the opportunity to assess nocturnal hypercarbia and hypoxia using two different methods at the earliest stage of the disease.

Method

A spirometric assessment was performed using the Spirolab III Color LCD device in patients who could cooperate with the research team. The study data were recorded on a computer using the WinspiroPRO 6,8 software that was compatible with the device. Spirometry was performed by a trained and experienced a nurse working at the respiratory function test (RFT) unit and in accordance with the recommendations made by the European Respiratory Society and American Thoracic Society[12,13]. Forced expiratory volume measured in the first second of exhalation after maximal inspiration (FEV1), forced vital capacity (FVC), and 25-75% volume capacity of exhalation (FEF25-75) were recorded. FEV1/FVC ratio was documented. The spirometric measurements were analyzed and interpreted in accordance with the guidelines of the above-mentioned societies[12,13]. Then, the patients were monitored during an overnight stay at the Pediatric Chest Diseases Department. Blood samples were taken via radial artery puncture for blood gas analysis to record the PaCO2 level at 11:00 pm, 03:00 am, and 07:00 am. The patients were connected to an oxycapnography device with a pediatric CO2Oral/ Nasal sampling set between 11:00 pm and 07:00 am, and pulse oximetry monitoring was performed using fingertip type pediatric probes. The capnography device was calibrated every 24 hours as recommended by the manufacturer. PetCO2, SpO2, pulse rate, and respiratory rate were measured every second by the oxycapnography device and saved in the device memory. The recorded data were transferred to the Excel (TM) software in the morning and were then analyzed in the computer. The mean PetCO2 level, the mean level and duration (hypercarbia time) of PetCO2 levels of ≥50 mmHg, and the mean level and duration (hypoxia time) of SpO2 readings below 90% were recorded. The “nocturnal hypercarbia index” and “nocturnal hypoxia index” were calculated by dividing the total recording time by hypercarbia time and hypoxemia time, respectively. Then, the correlation between the PaCO2 level measured in three blood gas analyses and the nocturnal PetCO2 levels were analyzed.

Definitions

Nocturnal Hypercarbia: Defined as the mean PetCO2 level ≥ 50 mmHg or mean PaCO2 level ≥ 45 mmHg for greater than 2% of the total recording time[14,15].

Nocturnal Hypoxia: Defined as the mean SpO2 level being below 90% for greater than 2% of the total recording time[16].

Nocturnal Hypoventilation: Defined as PetCO2 level ≥50mmHg for greater than 25% of the total recording time or SpO2 level below 90% for greater than 10% of the total recording time[14-16].

Nocturnal Hypercarbia Index: Defined as the ratio of the recording time showing hypercarbia on oxycapnography to the total recording time.

Nocturnal Hypoxia Index: Defined as the ratio of the recording time showing hypoxia on oxycapnography to the total recording time.

Night Arterial Blood Gases: According to routine clinical practice in the unit, nocturnal blood gas levels were measured at 11:00 pm as well as at 03:00 am and 07:00 am. Blood samples were drawn at rest and immediately carried in an ice bag to the central hospital laboratory for analysis. Oxycapnograph: The Capnostream™ 35 Portable Respiratory Monitor PM35MN, with Microstream™ EtCO2 and Nellcor™ SPO2 Technologies, is a portable bedside monitor that continuously monitors:

End-tidal carbon dioxide (PetCO2)-level of carbon dioxide in exhaled breath (CO2 Oral/Nasal sampling set)

Respiratory rate (RR)

Oxygen saturation (SpO2)

Pulse rate (PR)

Statistical Analysis

The statistical analyses of this study were performed using NCSS (Number Cruncher Statistical System) 2007 Statistical Software Package (Utah, USA). In addition to descriptive statistical methods (mean and standard deviation), temporal comparison of normally distributed variables was performed with simple repeated measures analysis of variance; subgroup comparisons with Newman-Keuls multiple comparison tests; independent group comparisons with one-way analysis of variance (ANOVA); dependent group comparisons with paired samples t-test; independent group comparisons of non-normally distributed variables with Kruskal Wallis test; independent two group comparisons with Mann Whitney-U test; the relations between qualitative variables with Chi-square test; and inter-variable correlations with Pearson’s correlation test. Bland &Altman test and interclass correlation coefficient were used to test agreement between mean end-tidal CO2, mean PaCO2 23:00, and mean PaCO2 07:00. The results were analyzed at the significance level of p<0.05.

Results

After patients with NMD were evaluated, Twenty-nine out of 50 NDM patients who did not meet the inclusion criteria were excluded from the study (Figure 1). The mean age of the study population including 17 (81%) and 4 (19%) female patients was 10.55±5 (6-16) years. The mean Body Mass Index (BMI) was 15.95±3.76 kg/m2. BMI percentile was<5 in 11 (52.38%), 25-50 in 5 (23.81%), and higher than 50 in 5 patients.

Figure 1.

After patients with NMD were evaluated, 29 patients of a total of 50 patients were excluded due to not having met the inclusion criteria, and the remaining 21 patients were enrolled.

The mean duration of NMD follow-up was 4.38±3.32 years. The distribution of NMDs was as follows: congenital muscular disease (n:16 including 10 patients with Duchenne or Becker muscular dystrophy (DMD-BMD), one patient with facioscapulohumeral muscular dystrophy (FSHMD), one patient with limb-girdle muscular dystrophy (LGMD), four patients with congenital myopathy), and five patients had hereditary polyneuropathy (hereditary motor-sensory neuropathy (HMSN)).

Spirometry test results of 14 (66.67%) patients were as follows: mean FEV1%, 78.09±27.72); mean FVC%, 73.09±23; mean FEV1/FVC , 101.18±12.87. As comorbidities gastroesophageal reflux disease was seen in one (4.76%) and scoliosis in six (28.57%) patients. The patients with thoracic scoliosis had a significantly lower mean SpO2 level compared to those without (p=0.019).

Mean values of parameters determined in blood gas analysis were as follows: PaCO2 at 11:00 pm (PaCO2 23:00): 40.28±5.37 mmHg; PaCO2 at 03:00: 42.13±4.39 mmHg; PaCO2 07:00: 42.93±6.67 mmHg, and finally PaCO2 noc: 41.78±4.69 mmHg which is the mean value of the three nocturnal blood gas PaCO2 levels. A significant change was observed between the mean PaCO2 23:00, PaCO2 03:00, and PaCO2 07:00 (p=0.032). The mean PaCO2 23:00 level was significantly lower than mean PaCO2 03:00 and mean PaCO2 07:00 levels (p=0.042, p=0.032); the mean PaCO2 03:00 and mean PaCO2 07:00 levels were statistically comparable (p=0.550). Oxycapnography recording data were presented in Table I.

Table 1.

Oxycapnography data of patients with NMD.

Oxycapnography Data	n	Mean±SD	Median (IQR)
Total recording time (hours)	21	7.46±1.37	7.8 (7.4-8.2)
Mean PetCO2	21	37.75±3.45	38.3 (36.05-40.25)
Maximum PetCO2	21	47.67±3.04	47 (45.5-49.5)
Mean PetCO>45 mmHg 2	17	45.82±0.55	45.9 (45.4-46.1)
Nocturnal Hypercarbia	17	5.02±12.69	0.7 (0.26-3.2)
index (%)
Mean SpO2	21	97.45±1.12	97.6 (97.05-98.3)
Minimum SpO2	21	84.62±6.98	86 (79.5-91)
Mean SpO2<90	13	85.27±3.12	86.1 (85.4-87)
Nocturnal hypoxia index (%)	13	0.84±1.97	0.1 (0.06-0.415)

PetCO2: end-tidal CO2 partial pressure, SpO2: oxygen saturation

There were significant differences between PaCO2 results and PetCO2 and SpO2 recordings obtained in muscular dystrophy (MD) (n:12), myopathy (n:4) and polyneuropathy (n:5) groups (p>0.05). Tables 2 and 3 shows the duration of the primary disease, BMI, and pulmonary function test results, and the correlation between blood gas analysis and oxycapnography recordings of the patients in the whole study population and the muscular dystrophy group. The correlation between PaCO2 level in blood gas analysis and PetCO2 level measured by oxycapnography was shown in Tables 4 and 5. We may state that there is a concordance between PaCO2 and PetCO2 levels. However, the intraclass correlation coefficient is slightly above the upper limit of normal, and 95% CI was slightly wider than we expected.

Table 2.

Hypercarbia and hypoxemia status and the relationship between pulmonary function tests and other clinical conditions of the patients with NMD.

		Disease Duration	BMI	FEV1	FVC	FEV1/FVC
Mean PetCO2	r	-0.189	0.473	0.441	0.311	0.657
	p	0.412	0.03	0.175	0.352	0.028
Maximum PetCO2	r	0.018	-0.022	-0.205	-0.374	0.348
	p	0.938	0.926	0.546	0.257	0.295
Mean PetCO2>45 mmHg	r	-0.069	0.099	-0.024	-0.164	0.005
	p	0.792	0.705	0.951	0.673	0.991
Nocturnal Hypercarbia index (%)	r	-0.106	0.190	0.213	0.124	0.582
	p	0.686	0.466	0.583	0.751	0.100
Minimum SpO2	r	-0.247	0.206	0.396	0.285	0.781
	p	0.281	0.372	0.228	0.395	0.005
Mean SpO2<90	r	0.127	0.028	-0.233	-0.111	0.06
	p	0.68	0.927	0.615	0.813	0.898
Overall mean SpO2	r	-0.315	0.292	0.391	0.304	0.553
	p	0.165	0.198	0.235	0.364	0.078
Overnight Blood Gas mean PaCO2	r	-0.338	0.220	0.025	-0.044	-0.066
	p	0.134	0.339	0.943	0.899	0.847

Pearson Correlation

PetCO: end-tidal COpartial pressure, SpO: oxygen saturation, PaCO: partial carbon dioxide pressure, BMI: Body mass index, 22 22FEV1: Forced expiratory volume in one second, FVC: forced vital capacity.

Table 3.

Hypercarbia and hypoxemia status and the relationship between pulmonary function tests and other clinical conditions of the patients with Muscular Dystrophy.

		Disease Duration	BMI	FEV1	FVC	FEV1/FVC
Mean PetCO2	r	-0.147	0.417	0.521	0.391	0.758
	p	0.648	0.178	0.368	0.515	0.137
Maximum PetCO2	r	0.053	-0.201	-0.298	-0.384	-0.06
	p	0.871	0.531	0.626	0.524	0.924
Mean PetCO2>45 mmHg	r	0.055	-0.176	-0.673	-0.689	-0.560
	p	0.888	0.650	0.027	0.011	0.440
Nocturnal Hypercarbia index (%)	r	-0.345	0.139	0.814	0.851	0.323
	p	0.363	0.722	0.186	0.149	0.677
Minimum SpO2	r	-0.636	0.507	0.815	0.719	0.639
	p	0.026	0.092	0.093	0.171	0.018
Mean SpO2< 90	r	0.106	0.407	0.400	0.212	0.973
	p	0.842	0.423	0.738	0.864	0.147
Overall mean SpO2	r	-0.325	0.245	0.285	0.179	0.841
	p	0.303	0.443	0.643	0.773	0.074
Overnight Blood Gas mean PaCO2	r	-0.068	0.041	-0.068	-0.143	0.002
	p	0.833	0.900	0.913	0.819	0.998

Pearson Correlation

PetCO: end-tidal COpartial pressure, SpO: oxygen saturation, PaCO: partial carbon dioxide pressure, BMI: Body mass index, 22 22FEV1: Forced expiratory volume in one second, FVC: forced vital capacity.

Table 4.

Oxycapnography data of patients with NMD.

		PaCO207:00	Overnight Blood Gas mean PaCO2
Mean PetCO2	r	0.713	0.678
	p	0.003	0.007
Maximum PetCO2	r	0.409	0.418
	p	0.066	0.059
Mean PetCO2>45 mmHg	r	0.02	0.097
	p	0.939	0.710
Pearson Correlation
Mean PetCO2		Intra-class correlation Coefficient	95% Confidence Interval
PaCO207:00		0.791	0.517-0.843
Overnight Blood Gas mean		0.811	0.533-0.923
PaCO2

PetCO2: end-tidal CO2 partial pressure, PaCO2: partial carbon dioxide pressure.

Table 5.

Prevalence of nocturnal respiratory disorder assessed by two different methods.

Patient No	Age (years)	Thoracic scoliosis	Diagnostic Group	PaCO2 23:00	PetCO2 23:00*	SpO2 23:00	PaCO2 03:00	PetCO2 03:00*	SpO2 03:00	PaCO2 07:00	PetCO2 07:00*	SpO2 07:00	Mean PetCO2**	Nocturnal Hypercarbia Index	min Sp02	Nocturnal Hypoxia index
1	6.0	-	Muscular Dystrophy	35.2	41	97	42.7	41	97	38.0	41	97	44	-	84	0.08%
2	14.5	-	Muscular Dystrophy	41.6	35	98	43.3	46	97	44.8	45	92	46	0.90%	83	0.1%
3	18.0	-	Congenital Myopathy	37.2	38	95	26.1	41	94	38.1	41	93	44	-	80	0.1%
4	9.5	-	Congenital Myopathy	53.2	41	89	47.6	41	89	53.3	40	89	50	2.6%	91	-
5	8.5	-	Muscular Dystrophy	30.5	38	100	40.7	46	98	30.8	44	100	43	-	84	0.02%
6	17.0	Yes	Polyneuropathy	41.6	38	98	44	14	96	45.7	21	97	48	0.8%	87	0.1%
7	8.0	-	Congenital Myopathy	38.9	40	97	38.5	38	100	39.4	38	99	47	0.2%	86	0.04%
8	14.0	Yes	Congenital Myopathy	37.1	34	97	39.9	39	97	37.5	38	99	50	3.8%	74	0.2%
9	7.5	-	Muscular Dystrophy	39.3	38	100	41.1	36	97	37.1	37	98	49	0.5%	92	-
10	15.0	Yes	Muscular Dystrophy	40.2	39	96	40.6	42	94	38.4	39	96	49	0.7%	86	0.04%
11	6.0	-	Muscular Dystrophy	39.3	32	98	39.9	38	98	58.8	41	96	45	0.04%	92	-
12	6.0	Yes	Muscular Dystrophy	40.0	45	98	42.5	46	98	47.5	45	98	50	15.9%	90	-
13	7.5	-	Muscular Dystrophy	41.6	20	97	39.6	41	98	41.6	43	96	46	0.32%	79	23.8%
14	6.0	-	Polyneuropathy	48.6	40	97	46.3	38	97	50.9	15	98	47	0.97%	80	0.63%
15	17.0	-	Polyneuropathy	45.0	42	91	49.5	51	94	46.3	48	94	49	0.7%	75	0.1%
16	10.5	-	Muscular Dystrophy	48.0	40	95	49.5	48	96	48.3	52	92	53	4.5%8	92	-
17	18.0	-	Polyneuropathy	35.8	43	96	49.1	39	97	41.6	42	98	55	52.0%	95	-
18	17.0	Yes	Muscular Dystrophy	40.5	42	95	46.1	47	96	37.1	28	98	50	0.05%	74	7.05%
19	9.0	-	Muscular Dystrophy	33.6	35	93	40.9	42	96	46.6	48	96	49	1.6%	91	-
20	6.0	-	Muscular Dystrophy	34.8	39	95	30.7	38	96	35.7	41	95	44	-	90	-
	6.0	Yes	Polyneuropathy	43.9	44	98	45.1	41	98	44.1	46	97	46	0.5%	72	0.1%
The number and percentage of nocturnal hypercarbia cases (According to blood gas analysis)										8 (38%)			6(28,5%)		2 (9.5%)
The number and percentage of cases with nocturnal hypercarbia (According to PetCO2)															2 (9.5%)
The number and percentage of cases with nocturnal hypoxia
The number of cases with nocturnal hypoventilation according to AASM criteria

*PetCO2: end-tidal CO2 partial pressure at a specific time, **PetCO2: All night average end-tidal CO2 partial pressure, PaCO2: partial carbon dioxide pressure, SpO2: oxygen saturation, AASM: American Academy of Sleep Medicine (2019).

We had no patient with a PaCO2 level above 45 mmHg at 23:00 pm. PaCO2 levels were above 45 mmHg at 03:00 am in 6; and at 07:00 am in 5 patients; and three patients had a PaCO2 level above 50 mmHg. Two of our patients although their PaCO2 levels were below 45 mmHg at 07:00 am (Patient No.17 and 18) met the oxycapnographic criteria of nocturnal hypoventilation ie. one of them had hypercarbia and the other hypoxia. In contrast, one of the two patients with a PaCO2 level above 50 mmHg at 07:00 am had normal oxycapnography levels while the other had nocturnal hypercarbia. Two of the five patients with a PaCO2 level above 45 mmHg at 07:00 am had nocturnal hypercarbia; only one of the three patients with a PaCO2 level above 50 mmHg had nocturnal hypercarbia.

We detected nocturnal gas exchange problems in a total of seven (33%) patients including nocturnal hypercarbia in 5 and nocturnal hypoxia in 2 patients. According to the AASM criteria, only two (9.5%) of our patients met the criteria for nocturnal hypoventilation.

Discussion

We performed screening tests to reflect strictly early signs of SDB in a patient group free of any signs or symptoms of nocturnal hypoventilation in which we eliminated the likelihood of any mechanical upper airway obstruction contributing to a respiratory disorder. We observed nocturnal hypercarbia in oxycapnography monitoring in about one third of our patients. PaCO2 and PetCO2 levels shown by oxycapnography and blood gas analysis were statistically concordant. However, our intraclass correlation coefficient was slightly above the upper limit of normal. The 95% CI and SD values, which we expected to be closer to each other, were also slightly wider. We believe that we would have attained more realistic results when we had separately analyzed the diagnostic groups within themselves and run the study with a larger number of patients.

Whereas NH was detected in only 9.5% of our patients according to the internationally recognized AASM criteria. While the proportion of patients with signs suggesting a nocturnal respiratory disorder rose to 33% when other data were also taken into account. Although our “nocturnal hypercarbia and hypoxia” definitions based on capnography results did not indicate a need for noninvasive mechanical ventilation treatment in the face of current treatment guidelines, they nevertheless indicated that one third of our patients were in a high-risk group necessitating close follow-up.

Arterial blood gas analysis is the gold standard test for assessing PaCO2 levels. However, it is not regarded as an appropriate screening test considering its painful and invasive nature, as well as the need for continuous catheterization[17]. Various techniques for noninvasive CO2 level monitoring have been tested in different patient populations for diagnostic efficacy since their introduction into clinical practice. Takano et al.[18] reported that PetCO2 and PaCO2 were correlated in spontaneously breathing adults with the chronic parenchymal pulmonary disease; however, they still recommended the combined use of both methods. A study of 129 patients connected to a mechanical ventilator for any cause at a pediatric intensive care unit showed that PetCO2 monitoring may be recommended for the assessment of ventilation in patients with the mild-to-moderate pulmonary parenchymal disease. Nevertheless, the authors of the same study also stressed that the concordance between PetCO2 and PaCO2 was not strong enough among patients with severe parenchymal problems, for whom they argued blood gas analysis should continue to be the primary determinant of ventilatory efficiency[19].

Another study that enrolled patients admitted to the intensive care unit for neurological problems suggested that PetCO2 monitorization could be used to monitor ventilation and hemodynamic status of critically ill patients with neurological problems although taking at least one blood gas sample at the beginning of monitorization would be also appropriate[20]. We solely enrolled asymptomatic children with NMD. We screened sleep-disordered breathing using both PaCO2 and PetCO2 levels. We found a significant correlation between PaCO2 levels measured at 03:00 am or 07:00 am and the PetCO2 levels. But our intraclass correlation coefficient was slightly above the upper limit of normal. Additionally, we noted a slightly wider 95% CI and SD ranges than we expected.

We believe that we would have attained more realistic results when we had separately analyzed the diagnostic groups within themselves and run the study with a larger number of patients. Although SDB is common among children with NMD, it is dificult to predict its time of onset. Thus, SDB can be screened with capnography, a noninvasive method, when longer follow-up periods are typically needed.

Among children with NMD, PFTs are particularly known for being the best determinant of nocturnal hypoventilation. In a study of 19 patients with DMD, FEV1 level below 40% was found to be a sensitive (91%) but a nonspecific predictor of SDB[21]. Another prospective study reported that a drop in inspiratory vital capacity (IVC) to 60% may be a predictive criterion for the onset of nocturnal hypoventilation[22]. A drop of IVC ratio to 40%, on the other hand, was proposed as a better indicator of nocturnal hypoventilation[23]. However, several other studies contradicted these findings. In a study of 52 children with NMD in whom overnight capnography monitoring, daytime blood gas analysis, and PFT was performed, 22 patients were found to have nocturnal hypercarbia (PtcCO2>50 for greater than 2% of the recording) and 20 patients nocturnal hypoxemia (SpO2<90 for greater than 2% of the recording); additionally, there was a correlation between FVC and nocturnal SpO2. However, that study failed to show any correlation between daytime pulmonary function tests and nocturnal PetCO2[24]. On the other hand, there was a correlation between PetCO2 and SpO2 and FEV1/FVC when the entire group of patients with NMD was analyzed in our study. However, when we analyzed the muscular dystrophy patients separately, we noted a negative correlation between FEV1, FVC, and PetCO2, and a correlation between FEV1/FVC and min SpO2. We had only three patients with FEV1 and FVC levels below 60%, of whom only one had nocturnal hypercarbia. All of the remaining patients had FEV1 and FVC levels as expected. In agreement with the preceding study, we showed a statistically significant correlation between FEV1/ FVC and min SpO2.

Although we statistically showed that PetCO2 increased as FEV1 and FVC levels were reduced in the muscular dystrophy group, we could not ascertain the statistical relevance of this finding since basal pulmonary function levels were not much lower than expected. Spirometric pulmonary function test results such as FVC and FEV1 are highly correlated to nocturnal hypoventilation in adults with NMD, but it can not show such a correlation in children who have limited ability to complete pulmonary function tests[25].

Kyphoscoliosis is a common comorbidity in patients with NMD. It contributes to an ineffective respiratory pump during expansion of the chest by creating a mechanical restraint around costovertebral joints[26]. It is particularly known to cause reductions in vital capacity as high as 20%, which is further increased by the use of corsets to manage scoliosis[27]. We detected nocturnal hypercarbia in only two of our six patients with thoracic scoliosis, and we found no difference between patients with and without thoracic scoliosis with respect to spirometric pulmonary function test results. Our small sample size may indicate that our data on thoraco-scoliosis may not be reliable.

When sleep-disordered breathing in children is categorized into two main groups as OSA and nocturnal hypoventilation, polysomnography, and screening tests that can reveal nocturnal hypercarbia and hypoxemia come to one’s mind. As polysomnography has remained hardly accessible in clinical practice for years, and even episodes of hypercarbia discordant with the AHI index are detectable, other screening tests have gradually begun to be prioritized[4]. Kirk et al.[9] measured PSG and PetCO2 levels of 609 pediatric patients and found clinically significant increase, PetCO2 in 12% of patients with a low apnea-hypopnea index. A weak correlation between hypoventilation and AHI suggests that monitoring CO2 levels may prove predictive for SDB.

Depending on the screening test performed and the accepted NH definition, the prevalence of nocturnal hypoventilation ranges between 10.3% and 61.2%[28]. A previous study enrolled 232 adult patients with NMD based on eight different NH definitions (according to blood gas analysis and oxycapnography); daytime hypercapnia was detected in 38% of 232 patients and NH in 28% of those patients who were normocapnic during the daytime. The prevalence of SDB was found 36.8% in the muscular dystrophy group[14]. All of our patients were normocapnic during the daytime, and NH was detected in 33% (7/21) of them. Of those patients, 23.8% (5/21) showed nocturnal hypercarbia without concomitant nocturnal hypoxemia. We diagnosed NH in 25% of the muscular dystrophy group. Among those patients, two had isolated nocturnal hypercarbia and a patient had isolated nocturnal hypoxemia.

In a study on 221 pediatric patients, the NH prevalence was found 16% when the AASM criteria were used; moreover, its prevalence increased as the nocturnal hypercarbia index dropped. The prevalence of NH was found 31% when PetCO2 was greater than 50% and the nocturnal hypercarbia index was greater than 2%[15]. We likewise detected NH in 9.5% of our patients when we based it on the AASM criteria and 33% when we accepted a Nocturnal Hypercarbia Index of 2%.

We admit that our study has some limitations. There is no unanimously accepted criterion for the definitions of nocturnal hypoxemia and hypercapnia. Without having any chance to assess the level of prognostic relevance, we arbitrarily chose three of the previously used criteria[14-16]. Although our study’s prospective design is a favorable aspect, it has certain limitations such as the inclusion of a heterogeneous NMD population and a relatively small sample size. Furthermore, we lacked data on muscle strength (maximum inspiratory pressure and maximum expiratory pressure), cough efficiency (peak cough flow) and lung volumes in supine position due to lack of technical equipment and inadequate patient compliance, which resulted in spirometric pulmonary function testing being applied only in a certain proportion of the study population. As our sample size was small, the correlation between spirometric pulmonary tests and signs of abnormal nocturnal gas exchange may have been far from the actual figure. We could not compare PetCO2 /SpO2/PaCO2 monitoring with respiratory polygraphy or PSG to detect NH in patients with NMD. Instead, we aimed to determine if nocturnal hypercapnia existed in clinically asymptomatic NMD. We considered such approach would be an advantage for our study as signs of nocturnal hypoventilation are already more frequent than apneic events in patients with NMD.

Conclusion

We report for the first time the results of nocturnal PetCO2/SpO2/PaCO2 monitoring in an NMD population of the youngest age group reported so far. Our study indicated that nocturnal PetCO2 and PaCO2 levels were statistically concordant with each other although the use of PaCO2 alone is not sufficient for early detection of NH in NMD. We believe that nocturnal capno-oximetry needs to be considered among scanning tools for the diagnosis of hypoventilation.

NH is a condition that is difficult to diagnose since it onsets with subtle symptoms. Considering its major impact on morbidity and mortality as well as its amenability to therapy, it is of importance to diagnose NH. Our data indicate variations in the prevalence of NH depending on the definition and the methods used. As there is no verified nocturnal hypoventilation level associated with end-organ damage, it is warranted to consider more limited definitions for NH, to conduct larger studies to reach an agreement on the best definition of hypoventilation and to update the existing consensus guidelines.