Impact of oxygen concentration on time to resolution of spontaneous pneumothorax in term infants: a population based cohort study
© Shaireen et al.; licensee BioMed Central Ltd. 2014
Received: 26 March 2014
Accepted: 18 August 2014
Published: 23 August 2014
Little evidence exists regarding the optimal concentration of oxygen to use in the treatment of term neonates with spontaneous pneumothorax (SP). The practice of using high oxygen concentrations to promote “nitrogen washout” still exists at many centers. The aim of this study was to identify the time to clinical resolution of SP in term neonates treated with high oxygen concentrations (HO: FiO2 ≥ 60%), moderate oxygen concentrations (MO: FiO2 < 60%) or room air (RA: FiO2 = 21%).
A population based cohort study that included all term neonates with radiologically confirmed spontaneous pneumothorax admitted to all neonatal intensive care units in Calgary, Alberta, Canada, within 72 hours of birth between 2006 and 2010. Newborns with congenital and chromosomal anomalies, meconium aspiration, respiratory distress syndrome, and transient tachypnea of newborn, pneumonia, tension pneumothorax requiring thoracocentesis or chest tube drainage or mechanical ventilation before the diagnosis of pneumothorax were excluded. The primary outcome was time to clinical resolution (hours) of SP. A Cox proportional hazards model was developed to assess differences in time to resolution of SP between treatment groups.
Neonates were classified into three groups based on the treatment received: HO (n = 27), MO (n = 35) and RA (n = 30). There was no significant difference in time to resolution of SP between the three groups, median (range 25th-75th percentile) for HO = 12 hr (8–27), MO = 12 hr (5–24) and RA = 11 hr (4–24) (p = 0.50). A significant difference in time to resolution of SP was also not observed after adjusting for inhaled oxygen concentration [MO (a HR = 1.13, 95% CI 0.54-2.37); RA (a HR = 1.19, 95% CI 0.69-2.05)], gender (a HR = 0.87, 95% CI 0.53-1.43) and ACoRN respiratory score (a HR = 0.7, 95% CI 0.41-1.34).
Supplemental oxygen use or nitrogen washout was not associated with faster resolution of SP. Infants treated with room air remained stable and did not require supplemental oxygen at any point of their admission.
KeywordsOxygen Pneumothorax Newborn and nitrogen wash out
Pneumothorax is one of the most common air leak syndromes that occurs in the newborn period . It is classified into primary pneumothorax (without any obvious lung diseases) and secondary pneumothorax (due to underlying lung pathology, or associated with precipitating factors such as transient tachypnea of newborn, meconium aspiration, continuous positive pressure ventilation (CPAP), mechanical ventilation, pneumonia, respiratory distress syndrome or post surfactant treatment) [1–5]. Spontaneous pneumothorax (SP) is a form of primary pneumothorax in neonates. It usually occurs in the absence of inciting risk factors at birth . The mechanism is related to maladaptive transition after birth. The presence of persistently high or unequal transpulmonary inflating pressure in the alveoli during the transition period results in rupture of alveoli into the pleural space and produces a spontaneous pneumothorax [6, 7]. The incidence of radiologic SP is 1% to 2% and symptomatic SP is 0.05% to 1% in all live births [1, 4, 8]. Pneumothorax increases morbidity, prolongs hospital stay, causes parental anxiety and, in some cases, can also result in death [1, 8, 9].
The adult literature suggests that inhaling higher concentrations of oxygen between 60% to 100% (nitrogen washout) compared to room air improves the rate of resolution of symptomatic SP [10–12]. The theory of nitrogen washout proposes that the inhalation of 100% oxygen reduces the partial pressure of nitrogen in the alveolus compared to the pleural space. This gradient difference causes the nitrogen to diffuse from the pleural space into the alveoli, resulting in resorption of air from the pleural space into the alveoli and faster resolution of the pneumothorax [6, 10]. The treatment of SP in term neonates is based on theoretical and historical hypotheses, which favor the use of higher oxygen concentrations/nitrogen washout for rapid resolution of SP [13, 14]. However, there is minimal published evidence available for the optimal inspired oxygen concentration requirement to treat clinically significant SP in term neonates [15–17]. Empirically, SP is treated with variable concentrations of oxygen .
Treatment of SP with higher oxygen concentrations may lead to hyperoxic injury in neonates . Hyperoxia produces free oxygen radicals and enhances cellular apoptosis. It can alter the genetic expression of the cell and may be a cause of potential childhood cancers . Unrestricted oxygen use poses a financial burden to the health care system due to prolonged neonatal intensive care unit (NICU) admission ; this burden is more profound in resource poor countries .
To the best of our knowledge, in the term neonatal population, no standardized management guidelines or concrete evidence are available in the literature that show an advantage or disadvantage of using higher concentrations of oxygen for resolution of spontaneous pneumothorax. In our NICU there are three practice patterns. Some neonatologists treat SP with nitrogen washout (60 to 100% inspired O2 concentration); some titrate the O2 concentration targeting a pulse oximeter O2 saturation (SpO2) ≥ 95%; and some treat with room air alone. We hypothesized that term neonates who inhaled higher oxygen concentrations/nitrogen washout would have a quicker resolution of spontaneous pneumothorax as compared to infants treated with room air or lower oxygen concentrations. The aim of this study was to identify the time to clinical resolution of spontaneous pneumothorax in term neonates treated with high fraction of inspired oxygen (FiO2)/nitrogen washout (HO: FiO2 ≥ 60%), moderate oxygen (MO: FiO2 < 60%) or room air (RA: FiO2 = 21%).
Data source and settings
This was a population based cohort study of spontaneous pneumothorax in term neonates in Calgary, Alberta, Canada. We reviewed the medical records (both electronic data and patients’ charts) of all term infants admitted with a diagnosis of pneumothorax to all NICUs in Calgary between 1st January 2006 to 31st December 2010. The list of patients was retrieved from the NICU database and the medical records departments by identifying infants with an ICD-9-CM (512 and 512.81) and an ICD-10-CM (J93.1 and P25.1) code for pneumothorax. Ethics approval was obtained from the Conjoint Health Research Ethics Board at the University of Calgary.
All term newborns (gestational age ≥37 weeks) admitted to the NICU from birth to 72 hours of life with signs of respiratory distress and a diagnosis of spontaneous pneumothorax confirmed by chest X-ray (CXR), were included in the study. Newborns with congenital/chromosomal anomalies, history of meconium stained liquor/meconium aspiration, respiratory distress syndrome, transient tachypnea of newborn, pneumonia, tension pneumothorax requiring thoracocentesis or chest tube drainage, those who received positive pressure ventilation (PPV) or mechanical ventilation (MV) before the diagnosis of spontaneous pneumothorax were excluded.
Based on initial fraction of inspired oxygen concentration used at admission for the treatment of SP, neonates were divided into three groups: high fraction of inspired oxygen (FiO2)/nitrogen washout (HO: FiO2 ≥ 60%), moderate oxygen (MO: FiO2 < 60%) or room air (RA: FiO2 = 21%). The decision to treat a patient with room air, high or moderate oxygen concentration was based on individual attending neonatologist preference.
The initial concentration of oxygen used for treatment of spontaneous pneumothorax was the primary exposure variable. Duration of oxygen therapy was calculated from the initiation of oxygen therapy until clinical resolution of pneumothorax or till treatment failure (development of tension pneumothorax within 72 hours of diagnosis and treatment of SP). Data on maximum O2 concentration inspired during the treatment period was based on the highest recorded dose documented in the chart. The method of oxygen delivery at the time of admission was either via oxyhood, or nasal cannula. The concentration of inspired oxygen via oxyhood was analyzed through an oxygen analyzer. The concentration of inspired oxygen through nasal prongs was calculated according to the “STOP- ROP effective FiO2 conversion table” , a modified equation described by Benaron and Benitz . The severity of respiratory distress was classified according to the validated “Acute care of at-risk newborn (ACoRN)” respiratory score [22–24]. The estimation of pneumothorax size was abstracted from the radiologist report; radiologists were blinded to the treatment group. There was no validated method available to measure the size of pneumothorax on CXR in neonates so an estimation was made according to adult guidelines as mentioned in the British Thoracic Society (BTS) guidelines , and from the neonatal literature [8, 25–27].
Clinical resolution of pneumothorax was defined as cessation of respiratory distress and discontinuation of oxygen treatment with maintenance of SpO2 ≥ 95%. Nitrogen washout was defined in our NICU as the use of 60 to 100% of inspired O2 continuously for at least 6 hours.
Time to clinical resolution of spontaneous pneumothorax, measured in hours was the primary outcome variable. Neonates were followed from admission until clinical resolution of pneumothorax as documented in the patient chart. The secondary outcome variables were length of hospital stay for the treatment of pneumothorax and treatment failure (development of tension pneumothorax within 72 hours of diagnosis and treatment of SP). Tension pneumothorax was characterized by rapid instability of vital signs and shifting of the mediastinum on CXR, requiring thoracocentesis, chest tube insertion or mechanical ventilation .
Descriptive statistics were used to describe the study population. Chi-square tests, analysis of variance (ANOVA) and Kruskal-Wallis tests were used to assess differences in categorical and continuous variables stratified by the concentration of oxygen received. A Cox proportional hazards model was used to assess differences in time to clinical resolution of SP. A crude model and a model adjusted for infant gender and ACoRN respiratory score were derived. An adjusted hazard ratio (aHR) >1 would indicate a benefit of oxygen treatment. P value < 0.05 was taken as significant. All analyses were conducted in Stata SE version 11.
Baseline maternal and neonatal characteristics
RA (n = 30)
MO (n = 35)
HO (n = 27)
Maternal age (yrs), mean (SD)
Antenatal corticosteroids, n (%)
Chorioamnionitis, n (%)
Cesarean section, n (%)
Male, n (%)
GA, wks, mean (SD)
BW, g, mean (SD)
Resuscitation at birth, n (%) (free flow oxygen only)
Apgar score at 5 min, median (range 25th -75th percentile)
Age of admission, hours, median (range 25th -75th percentile)
RA (n = 30)
MO (n = 35)
HO (n = 27)
Median (range 25th-75th percentile)
SpO2 (%) at admission
Concentration of oxygen (%) inspired at initiation of treatment
Maximum oxygen concentration (%) inspired during the treatment period
Total duration of supplemental oxygen therapy, hours
Respiratory rate at admission
ACoRN score at admission
ACoRN score according to severity
Length of hospital stay, hours
Site of pneumothorax on chest x-ray
Size of pneumothorax based on CXR
Failure of treatment
Symptomatic spontaneous pneumothorax is one of the main reasons for admission of term newborn infants to the NICU. As most cases of SP present with mild respiratory distress, the management of SP with or without supplemental oxygen is invariably different between physicians [16, 17]. No consensus exists regarding the treatment of symptomatic spontaneous pneumothorax in stable term neonates. Questions such as whether or not spontaneous pneumothorax in term neonates should be treated with supplemental oxygen and whether higher oxygen concentrations, based on the historical hypothesis of nitrogen washout is helpful in the speedy resolution of spontaneous pneumothorax are still unanswered. To the best of our knowledge, this is the first study which focuses on the time to clinical resolution of SP in neonates treated with room air (21% FiO2) and with different supplemental oxygen concentrations.
No difference was observed in the time to clinical resolution of SP between infants treated with room air or different concentrations of oxygen in our study. Even after adjustment for respiratory morbidity and gender, neonates in the HO group (≥60% FiO2) did not show a rapid resolution of the SP.
There are several studies on neonatal pneumothorax and its risk factors; however there is a scarcity of studies on the effect of oxygen dose dependant resolution of SP in the neonatal population. An older study by Yu et al. in 1975 noted speedy resolution of spontaneous pneumothorax (within 48 hours) in neonates treated with higher oxygen concentrations . The drawback of this study is that it included both term and preterm neonates with all types of pneumothorax; and there was no comparison done with respect to different oxygen concentrations on the resolution rate of SP. A more recent study examined the time to resolution of spontaneous pneumothorax with 100% O2 inhalation in 45 term and near term neonates (>35 weeks gestation age) . In this retrospective study, Clark et al. compared neonates receiving inhaled 100% O2 (nitrogen washout group) (n = 26) and neonates receiving different concentrations of O2 targeting the SpO2 between 92 to 95% (n = 19), conventional therapy group . Their findings corroborate our results. They did not find a significant difference in the mean time to resolution of tachypnea (20 hours, standard deviation (SD) ± 26 vs. 37 hours, SD ± 27, p = 0.181), mean length of hospital stay (3.53 days, SD ± 1.68 vs. 4.35 days, SD ±1.96, p = 0.168) and mean time to first oral feed between the conventional therapy and nitrogen washout groups . Our study differs in that we studied the impact of different O2 concentrations as well as the effect of room air (21% FiO2) on the resolution of SP. We observed that the time to resolution of SP was not significantly longer in neonates who were just in room air (21% FiO2) vs. neonates receiving different O2 concentrations higher than the room air. An interesting finding was that neonates in room air never required supplemental oxygen, or experienced treatment failure (pneumothorax, need for thoracocentesis, chest tube insertion or mechanical ventilation) at any time during their hospitalization. This could mean that room air (21% FiO2) may be as effective as any higher inhaled O2 concentrations for resolution of SP. Future trials will possibly find the room air as the optimal O2 concentration requirement for resolution of symptomatic pneumothorax and prevention of O2 toxicity.
All other studies which focused on the resolution rate of spontaneous pneumothorax in association with inhaled O2 concentration were conducted on animal subjects. A randomized clinical trial on 23 rabbits, in whom unilateral pneumothorax was induced, Ronald et al. observed  that the resolution rate of pneumothorax was shorter (36 hours) in the group that received a higher concentration of oxygen (FiO2 ≥ 60%), compared to the group that was treated with room air (48 hours) . In two other studies on rabbits by England  and Zierold , the rate of resolution was dependent on the concentration of oxygen. The higher the concentration of inspired oxygen, the faster the resolution of pneumothorax. Two studies were identified in adults for the treatment of pneumothorax with oxygen inhalation [11, 31]. A prospective study in adults by Chadha  observed an increased rate of resolution of pneumothorax when treated with higher concentrations of inspired oxygen. The other study in 22 adult patients, divided into two groups (1-treated with room air and 2- intermittently treated with room air and oxygen from 9–36 hours, at 16 L/min flow) also showed a positive correlation on the rate of absorption of pneumothorax with oxygen treatment . The resolution rate of pneumothorax was 4.8 cm2/day with room air in both groups and increased to 17.9 cm2/day with oxygen treatment in the second group . No side effects of oxygen therapy were observed in this small study.
This retrospective study has few limitations due to the selection and information bias (detection bias). It was difficult to control bias and confounders due to the absence of randomization and blinding in this study. We had no control over some variables, therefore for homogeneity and as a reference we recorded these variables when these were observed at the time of admission. Although the use of higher oxygen concentration in sick neonates in HO and MO groups seemed justifiable at the admission, there was no targeted SpO2 goal observed to wean the inspired O2 concentrations despite of the decrement in respiratory severity scores. The oxygen concentration used at admission and the duration of oxygen treatment was widely variable and depended on the discretion of the admitting physician. The prescription for continuous “nitrogen washout” for minimum of 6 hours was clearly documented in the nursing notes, however there was no indication mentioned for titration of O2 concentration on achievement of SpO2 > 95%. This reflects the cultural and historical beliefs of physicians and nurses with regard to nitrogen washout treatment. We observed a decreasing trend in the use of oxygen treatment over time; while this did not achieve statistical significance (p = 0.822), this may be related to the small sample size when stratified by oxygen concentration and year. This decreasing trend may be the influence of emerging new literature on oxygen treatment as a drug and its pros and cons, especially toxicity of oxygen free radicals.
We did not have a CXR on every patient at the time of discharge from the NICU to determine the radiologic resolution of pneumothorax. Therefore we decided to record the time to clinical resolution of SP from clinical notes in the patient charts. According to the BTS guidelines , chest computed tomography (CT) scan is the best modality for accurate estimation of the size of pneumothorax. However, the size of pneumothorax does not always correlate well with the severity and resolution of symptoms associated with the pneumothorax . Therefore, the management should be tailored according to the clinical severity . The same clinical judgment for pneumothorax treatment is also applied to neonatal management. In newborns, there is no validated method available for correct estimation of size of pneumothorax on CXR. Portable CT scans are very costly, not available for sick babies and carry the risk of radiation. Due to the above mentioned reasons and inconsistency of the radiological reporting for the size of pneumothorax in neonates, the significance of size with the time to resolution of pneumothorax is not very reliable. Therefore, for the purpose of our study, we determined that the best option for calculating the time to resolution of pneumothorax was the clinical resolution of signs. The results of our study are applicable in neonates at other centres.
This study has been the first to determine the time to clinical resolution of spontaneous pneumothorax in term neonates treated with room air and different concentrations of inhaled oxygen. We found that supplemental oxygen use was not associated with faster resolution of SP in HO and MO groups. Neonates who were treated with room air rather than supplemental oxygen did not have longer recovery times. Oxygen treatment for pneumothorax should be viewed as a prescribed drug, with documentation for its indication, target saturation and titration goals. Future prospective trials for the treatment of symptomatic spontaneous pneumothorax using 100% inspired oxygen concentration vs. room air in sick neonates with targeted SpO2 goal, will be beneficial in developing an acceptable treatment guideline for term neonates.
This was a retrospective study based on the chart review. Personal information of patients was not disclosed. A waiver of the consent was obtained from the Conjoint Health Research Ethics Board at the University of Calgary.
British Thoracic Society
Continuous positive pressure ventilation
Saturation of peripheral oxygen
Fractional inspired oxygen
The authors wish to acknowledge the Section of Neonatology, Department of Paediatrics, and University of Calgary for their support in conducting this study. We would also like to thank Michelle Matthews, a data management clerk for her help in generating a patient list according to ICD codes.
The authors have no financial relationships relevant to this article to disclose.
No funding was secured for this study.
- Smith J, Schumacher RE, Donn SM, Sarkar S: Clinical course of symptomatic spontaneous pneumothorax in term and late preterm newborns: Report from a large cohort. Am J Perinatol. 2011, 28 (2): 163-168.PubMedGoogle Scholar
- Meberg A, Greve-Isdahl M, Heier CA: Pulmonary air-leakage in newborn infants. Tidsskr Nor Laegeforen. 2007, 127 (18): 2371-2373.PubMedGoogle Scholar
- Navaei F, Aliabadi B, Moghtaderi M, Kelishadi R: Predisposing factors, incidence and mortality of pneumothorax in a neonatal intensive care unit in Isfahan, Iran. Zhongguo Dang Dai Er Ke Za Zhi. 2010, 12 (6): 417-420.PubMedGoogle Scholar
- Apiliogullari B, Sunam GS, Ceran S, Koc H: Evaluation of neonatal pneumothorax. J Int Med Res. 2011, 39 (6): 2436-2440.PubMedGoogle Scholar
- Katar S, Devecioglu C, Kervancioglu M, Ulku R: Symptomatic spontaneous pneumothorax in term newborns. Pediatr Surg Int. 2006, 22 (9): 755-758.PubMedGoogle Scholar
- Chernick V, Avery ME: Spontaneous Alveolar Rupture at Birth. Pediatrics. 1963, 32: 816-824.PubMedGoogle Scholar
- Sistoza LC: Pneumothorax. Neonatology: Management, Procedures, On-Call Problems, Diseases and Drugs. Edited by: Gomella T, Cunningham MD, Eyal FG. 2009, New York: McGraw-Hill, 344-347. 6Google Scholar
- Abu-Shaweesh JM: Respiratory disorders in preterm and term infants. Neonatal-Perinatal Medicine: Diseases of the Fetus and Infant. Edited by: Martin RJ, Fanaroff AA, Walsh MC. 2010, St.Louis: Elsevier Health Sciences, 1164-1166. 9Google Scholar
- Aly H, Massaro A, Acun C, Ozen M: Pneumothorax in the newborn: clinical presentation, risk factors and outcomes. J Matern Fetal Neonatal Med. 2013, 27 (4): 402-406.PubMedGoogle Scholar
- MacDuff A, Arnold A, Harvey J: Management of spontaneous pneumothorax: British Thoracic Society Pleural Disease Guideline 2010. Thorax. 2010, 65 (2): ii18-ii31.PubMedGoogle Scholar
- Northfield TC: Oxygen therapy for spontaneous pneumothorax. Br Med J. 1971, 4 (5779): 86-88.PubMedPubMed CentralGoogle Scholar
- Al-Qudah A: Treatment options of spontaneous pneumothorax. Indian J Chest Dis Allied Sci. 2006, 48 (3): 191-200.PubMedGoogle Scholar
- Ashmore PG: Spontaneous Pneumothorax in the Newborn. Can Med Assoc J. 1965, 92: 309-311.PubMedPubMed CentralGoogle Scholar
- Yu VY, Liew SW, Robertson NR: Pneumothorax in the newborn. Changing pattern. Arch Dis Child. 1975, 50 (6): 449-453.PubMedPubMed CentralGoogle Scholar
- Esme H, Dogru O, Eren S, Korkmaz M, Solak O: The factors affecting persistent pneumothorax and mortality in neonatal pneumothorax. Turk J Pediatr. 2008, 50 (3): 242-246.PubMedGoogle Scholar
- Roberts D, Wacogne I: Question 3. In patients with spontaneous pneumothorax, does treatment with oxygen increase resolution rate?. Arch Dis Child. 2010, 95 (5): 397-398.PubMedGoogle Scholar
- Clark SD, Saker F, Schneeberger MT, Park E, Sutton DW, Littner Y: Administration of 100% oxygen does not hasten resolution of symptomatic spontaneous pneumothorax in neonates. J Perinatol. 2014, 34 (7): 528-531.PubMedGoogle Scholar
- Saugstad OD: Hyperoxia in the term newborn: more evidence is still needed for optimal oxygen therapy. Acta Paediatr Suppl. 2012, 101 (464): 34-38.Google Scholar
- Dennery PA: Oxygen administration in the care of neonates: a double-edged sword. Chin Med J (Engl). 2010, 123 (20): 2938-2942.Google Scholar
- Phelps DL, STOP-ROP Multicenter Study Group: Supplemental Therapeutic Oxygen for Prethreshold Retinopathy Of Prematurity (STOP-ROP), a randomized, controlled trial. I: primary outcomes. Pediatrics. 2000, 105 (2): 295-310.Google Scholar
- Benaron DA, Benitz WE: Maximizing the stability of oxygen delivered via nasal cannula. Arch Pediatr Adolesc Med. 1994, 148 (3): 294-300.PubMedGoogle Scholar
- Singhal N, Lockyer J, Fidler H, Aziz K, McMillan D, Qiu X, Ma X, Du L, Lee SK: Acute Care of At-Risk Newborns (ACoRN): quantitative and qualitative educational evaluation of the program in a region of China. BMC Med Educ. 2012, 12: 44-PubMedPubMed CentralGoogle Scholar
- Ma XL, Xu XF, Chen C, Yan CY, Liu YM, Liu L, Xiong H, Sun HQ, Lai JP, Yi B, Shi JY, Du LZ: Epidemiology of respiratory distress and the illness severity in late preterm or term infants: a prospective multi-center study. Chin Med J (Engl). 2010, 123 (20): 2776-2780.Google Scholar
- ACoRN Acute Care of at Risk Newborns, ACORN Neonatal Society. http://www.acornprogram.net,
- Carey B: Neonatal air leaks: pneumothorax, pneumomediastinum, pulmonary interstitial emphysema, pneumopericardium. Neonatal Netw. 1999, 18 (8): 81-84.PubMedGoogle Scholar
- Ponte C, Remy J, Bonte C, Lequien P, Lacombe A: Pneumothorax and pneumomediastinum in newborn infants. Clinical and radiological study of 80 cases. Arch Fr Pediatr. 1971, 28 (4): 450-PubMedGoogle Scholar
- Hoffer FA, Ablow RC: The cross-table lateral view in neonatal pneumothorax. AJR Am J Roentgenol. 1984, 142 (6): 1283-1286.PubMedGoogle Scholar
- Hill RC, DeCarlo DP, Hill JF, Beamer KC, Hill ML, Timberlake GA: Resolution of experimental pneumothorax in rabbits by oxygen therapy. Ann Thorac Surg. 1995, 59 (4): 825-827. discussion 827–828PubMedGoogle Scholar
- England GJ, Hill RC, Timberlake GA, Harrah JD, Hill JF, Shahan YA, Billie M: Resolution of experimental pneumothorax in rabbits by graded oxygen therapy. J Trauma. 1998, 45 (2): 333-334.PubMedGoogle Scholar
- Zierold D, Lee SL, Subramanian S, DuBois JJ: Supplemental oxygen improves resolution of injury-induced pneumothorax. J Pediatr Surg. 2000, 35 (6): 998-1001.PubMedGoogle Scholar
- Chadha TS, Cohn MA: Noninvasive treatment of pneumothorax with oxygen inhalation. Respiration. 1983, 44 (2): 147-152.PubMedGoogle Scholar
- The pre-publication history for this paper can be accessed here:http://www.biomedcentral.com/1471-2431/14/208/prepub
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.