Article Text


Minimally-invasive surfactant therapy in preterm infants on continuous positive airway pressure
  1. Peter A Dargaville1,2,
  2. Ajit Aiyappan1,3,
  3. Antonio G De Paoli1,
  4. Carl A Kuschel3,4,5,
  5. C Omar F Kamlin3,4,5,
  6. John B Carlin6,7,
  7. Peter G Davis3,4,5
  1. 1Department of Paediatrics, Royal Hobart Hospital and University of Tasmania, Hobart, Australia
  2. 2 Menzies Research Institute, Hobart, Australia
  3. 3 Neonatal Services, Royal Women's Hospital, Melbourne, Australia
  4. 4 Department of Obstetrics and Gynaecology, University of Melbourne, Melbourne, Australia
  5. 5 Murdoch Childrens Research Institute, Melbourne, Australia
  6. 6 Clinical Epidemiology and Biostatistics Unit, Murdoch Childrens Research Institute, Melbourne, Australia
  7. 7 Department of Paediatrics, University of Melbourne, Melbourne, Australia
  1. Correspondence to A/Prof Peter Dargaville, Department of Paediatrics, Royal Hobart Hospital, Liverpool Street, Hobart TAS 7000, Australia; peter.dargaville{at}


Objective To evaluate the applicability and potential effectiveness of a technique of minimally-invasive surfactant therapy (MIST) in preterm infants on continuous positive airway pressure (CPAP).

Methods An open feasibility study of MIST was conducted at two sites. Infants were eligible for MIST if needing CPAP pressure ≥7 cm H2O and FiO2 ≥0.3 (25–28 weeks gestation, n=38) or ≥0.35 (29–32 weeks, n=23). Without premedication, a narrow-bore catheter was inserted through the vocal cords under direct vision. Surfactant (100 or 200 mg/kg Curosurf) was then instilled, followed by reinstitution of CPAP. Outcomes were compared between surfactant-treated infants and historical controls achieving the same CPAP and FiO2 thresholds.

Results Surfactant was successfully administered via MIST in all cases, with a rapid and sustained reduction in FiO2 thereafter. For infants at 25–28 weeks gestation, need for intubation <72 h was diminished after MIST compared with controls (32% vs 68%; OR 0.21, 95% CI 0.083 to 0.55), with a similar trend at 29–32 weeks (22% vs 45%; OR 0.34, 95% CI 0.11 to 1.1). Duration of ventilation and incidence of bronchopulmonary dysplasia were similar, but infants receiving MIST had a shorter duration of oxygen therapy.

Conclusion Surfactant delivery via a narrow-bore tracheal catheter is feasible and potentially effective, and deserves further investigation in clinical trials.

Statistics from

What is already known on this topic

  • Many preterm infants managed initially on continuous positive airway pressure (CPAP) go on to require intubation because of respiratory distress related to surfactant deficiency.

  • There is emerging evidence that techniques of minimally-invasive surfactant therapy (MIST) may help to avoid intubation and its deleterious consequences in infants on CPAP.

What this study adds

  • A technique of MIST using a narrow-bore vascular catheter inserted into the trachea was successfully applied by neonatal physicians at two sites.

  • Administration of surfactant via MIST resulted in a sustained reduction in oxygen requirement, and a decrease in the need for intubation for infants 25–28 weeks gestation.


The evolution of respiratory care for the preterm infant has seen an increase in the use of nasal continuous positive airway pressure (CPAP) as the initial means of respiratory support.1 ,2 In infants <30 weeks gestation, clinical trials of early CPAP, without prior intubation and surfactant therapy, have shown this approach to be at least as effective as intubation at the outset, with trends towards reduction in the duration of ventilation and incidence of bronchopulmonary dysplasia (BPD).3,,5 Application of CPAP to an unselected population of preterm infants does, however, carry the risk of CPAP failure relating to unremitting respiratory distress syndrome (RDS). In the CPAP trials to date, up to one-half of infants commencing on CPAP have required intubation in the first week of life,3,,5 many with high-oxygen requirements in the context of RDS related to surfactant deficiency. It is thus conceivable that outcomes for infants commencing on CPAP could be further improved if those with significant RDS were to receive exogenous surfactant at an early stage.

Recognising the merits of early surfactant,6 the technique of intubation, surfactant administration and extubation (INSURE) has been investigated in preterm infants on CPAP, with the finding of a reduced need for mechanical ventilation in some studies,7 ,8 but not others, mostly attributable to inability to extubate after the procedure.9 This limitation of INSURE has prompted the pursuit of alternative and less-invasive means of giving surfactant to preterm infants on CPAP. Several techniques of minimally-invasive surfactant therapy (MIST) have been described,10,,14 of which those involving direct tracheal catheterisation have the most currency, being more readily applicable without special expertise or delivery devices. Use of a flexible feeding tube positioned in the trachea with Magill's forceps to administer surfactant has recently been reported to reduce the need for subsequent mechanical ventilation, and some evidence of a shorter duration of oxygen therapy.15

We have recently reported an alternative method of surfactant delivery via tracheal catheterisation using a semirigid vascular catheter, which we found to be practicable in a small group of preterm infants with RDS, with no significant procedural complications, and evidence of effective surfactant delivery.14 Herein, we report an extension of this work to a two-site feasibility study of this method of MIST. We aimed to (A) evaluate the applicability and apparent safety of the technique in the hands of a larger number of operators at two sites, (B) document the physiological response to surfactant administration, and (C) compare outcomes of infants receiving MIST with like-gestation historical controls with RDS of similar severity.


Study sites

The study was conducted within the Neonatal Intensive Care Units of the Royal Hobart Hospital (RHH) and the Royal Women's Hospital, Melbourne (RWH). Approval for the study was obtained from the institutional Ethics committee at each site. Care provided to preterm infants at 25–32 weeks gestation is similar in the two centres.16 Both units use CPAP where possible as the initial means of respiratory support, and neither unit practices INSURE nor uses the high flow nasal cannula system for initial respiratory support. CPAP levels range from 5 to 8 cm H2O, with pressure titrated according to oxygen requirement and work of breathing. FiO2 is adjusted to maintain oxygen saturation (SpO2) in the range of 88–92%. Infants are intubated where FiO2 ≥0.5, or where there is respiratory acidosis (pH <7.2) or unremitting apnoea. This approach to intubation was maintained throughout the timeline of this study, and was thus applicable to both the group receiving MIST, and the historical controls.

Study groups

Infants receiving MIST

The feasibility study of MIST was conducted at RHH between June 2009 and May 2011, and at RWH between May 2010 and February 2011. Preterm infants between 25 and 32 completed weeks of gestation were eligible for inclusion if they were treated with CPAP as initial respiratory support, were less than 24 h of age, and required a CPAP pressure of ≥7 cm H2O and an FiO2 ≥0.3 (25–28 weeks gestation) or ≥0.35 (29–32 weeks). Infants were not included if they had been previously intubated, or if there was a congenital anomaly affecting respiratory function. Written parental consent was obtained prior to enrolment in the study.


In both participating centres, data were collected from infants managed on CPAP in a time period immediately before beginning the MIST feasibility study (RHH: from June 2006 to June 2009, RWH: from May 2009 to April 2010).16 The chosen time period at each site reflected the differing admission rates at each centre. Inclusion and exclusion criteria identical to the MIST group, including achievement of the same CPAP and FiO2 thresholds, were applied to all like-gestation infants on CPAP in these time periods, allowing identification of a relevant control group.

MIST procedure

Surfactant instillation via tracheal catheterisation (the ‘Hobart method’) was performed as previously reported,14 with some modifications as described below. The technique was only performed on infants in a stable condition, with heart rate >120 bpm and oxygen saturation (SpO2) >85%. A 16 gauge, 130 mm vascular catheter (16G Angiocath, BD, Sandy, Utah, USA) was marked to indicate desired depth of insertion (25–26 weeks: 1 cm, 27–28 weeks: 1.5 cm, 29–32 weeks: 2 cm). Direct laryngoscopy was performed, the tracheal catheter was inserted beyond the vocal cords to the required depth, and held in position at lips. If catheterisation of the trachea was not possible within 20–30 s, CPAP was briefly reinstituted followed by a further attempt. Once the catheter was correctly positioned, surfactant (Curosurf, Chiesi Farmaceutici, Parma, Italy) was given at a dose of 100 or 200 mg/kg, with the 200 mg/kg dose opted for in 25–28-week infants in the latter part of the study to discern whether there may be any potential benefit compared with 100 mg/kg dosing. The tracheal catheter was immediately withdrawn, and CPAP recommenced. Positive pressure inflations were given by mask if the infant was apnoeic or bradycardic. Modifications made to the MIST technique during the course of the study included retaining the CPAP prongs in situ throughout the procedure, and delivery of the surfactant as 3–4 boluses over 15–30 s (rather than more rapidly). Some operators fashioned a slight curvature in the catheter prior to use.

Management after MIST was according to standard practice for infants on CPAP at each study site. Infants were intubated if FiO2 was >0.5, or if there was respiratory acidosis or apnoea. A further dose of surfactant (100 mg/kg) was given after intubation if clinically indicated.

Care throughout hospitalisation was as per routine for all infants, including monitoring for, and treatment of, patent ductus arteriosus (PDA), and screening for intraventricular haemorrhage (IVH) and retinopathy of prematurity (ROP) according to standard schedules.

Data collection

Demographic and clinical data were collected for all infants from the unit databases and the patient records. In infants receiving MIST, details of the procedure were recorded prospectively, along with paired PCO2 values from blood gas samples before, and 2 h after MIST. Changes in CPAP pressure and FiO2 in the first 4 h after MIST, and at 24 and 48 h of life, were documented.

Neonatal outcomes recorded for all infants included need for intubation and mechanical ventilation in the first 72 h (and thereafter), further surfactant therapy, PDA requiring medical and/or surgical therapy and pneumothorax. The incidence of mortality, BPD (need for supplemental oxygen at 36 weeks corrected gestational age),17 IVH grades III and IV,18 ROP greater than stage 2 and Bell stage II or III necrotising enterocolitis (NEC)19 was noted. The presence of major morbidity (at least one of BPD, grade III or IV IVH, periventricular leucomalacia, ROP>stage 2, or NEC)20 was ascertained, as were the durations of respiratory support (ventilation+CPAP), oxygen therapy and intensive care admission.

Statistical analysis

Data were analysed within two gestation ranges, 25–28 weeks and 29–32 weeks. Continuous data are summarised with the median and IQR, or mean and SD where symmetrically distributed. Continuous variables were compared between infants receiving MIST and controls using the Mann–Whitney test, and dichotomous outcomes with χ2 or Fisher's exact test, as appropriate. CPAP, FiO2 and PCO2 values after MIST were compared with pre-MIST values using paired t tests. All reported p values are two-sided.

Given the nature of the study and the limited data on the likely reduction in rate of intubation in infants receiving MIST by the Hobart method, a formal sample size calculation could not be performed. We aimed to enrol at least 20 infants at each gestation range, and to include data on an equivalent or greater number of historical controls.


A total of 61 infants were enrolled in the MIST feasibility study at the two sites during the respective study periods, including 31 at RHH and 30 at RWH. Eighteen of the infants receiving MIST at RHH have been previously reported in a more limited form.14 Their inclusion in this report allows a relatively large cohort of infants receiving MIST to be compared with relevant controls. Demographic and clinical characteristics of the infants receiving MIST were generally well matched with the historical controls (table 1).

Table 1

Demographic and clinical characteristics

Infants at 25–28 weeks gestation received surfactant via MIST relatively early (median 3.1 h), whereas the 29–32 week group were treated at a median age of 9.9 h, reflecting at least in part the more stringent FiO2 enrolment criterion in this group (table 2). The MIST technique was performed by 12 consultant neonatologists and three neonatal fellows, with successful administration of surfactant in every case. A second or third catheterisation attempt was required in 20% of infants overall, and more often for a first-time operator than those with prior experience of the technique. Surfactant dose varied in the 25–28-week group, with seven infants receiving a dose of 200 mg/kg. Surfactant reflux was observed in 30% of infants overall, and bradycardia in 36%, most usually related to direct laryngoscopy. Median time to return the infant to previous CPAP settings was less than 2 min for both gestation ranges. Resuscitative measures other than positive pressure inflations were not required in any case, and no infant was intubated within 60 min of the procedure.

Table 2

Details of MIST procedure

For infants at 25–28 weeks, paired blood gas samples (n=28) showed a slightly lower PCO2 and higher pH at 2 h after the procedure (PCO2 pre-MIST: mean 51 (SD 7.9) mm Hg; 2 h: 48 (6.6) mm Hg; p=0.021; pH 7.30 (0.06) vs 7.32 (0.06); p=0.0047). No such changes were noted in the 29–32-week group (data not shown).

A modest decrease in CPAP pressure was noted after MIST, sustained at least to 24 h of life (figure 1). Oxygenation improved rapidly after MIST in both gestation ranges, with a reduction in FiO2 at 4 h after MIST, and at 24 and 48 h of life, compared with pre-MIST values.

Figure 1

CPAP pressure and FiO2 after MIST. Plots of CPAP pressure (panels A, B) and FiO2 (panels C, D) in the first 4 h after MIST, and at 24 and 48 h of life, for infants at 25–28 weeks (panels A, C) and 29–32 weeks (panels B, D). Mean and standard error. X-axis is discontinuous. Black arrow indicates the timing of MIST, which was performed at the median time of 3.1 h (25–28 weeks) and 9.9 h (29–32 weeks). Numbers in parentheses represent number of data points. Data is not used once intubated, nor, in the case of CPAP pressure, when CPAP discontinued. CPAP, continuous positive airway pressure; MIST, minimally-invasive surfactant therapy. *Differs from pre-MIST value, *p<0.05, **p<0.01, paired t test.

Compared with historical controls, the need for intubation before 72 h was considerably reduced after MIST in 25–28-week infants (OR 0.21, 95% CI 0.083 to 0.55), with a smaller reduction in the 29–32-week group (OR 0.34, 95% CI 0.11 to 1.1) (table 3). In the 25–28-week group, none of the seven infants receiving a surfactant dose of 200 mg/kg required intubation <72 h, compared with 12 of 31 infants receiving a lesser dosage (p=0.075, Fisher's exact test). Surfactant was administered to 66% and 39% of control infants. In 25–28-week infants, MIST was associated with a greater incidence of later medical therapy for PDA. None of the study infants required ductal ligation.

Table 3

Neonatal outcomes

Rates of pneumothorax, BPD and other major morbidities were not substantially different between the MIST groups and their respective controls (table 3). Duration of mechanical respiratory support was similar between MIST and control groups, but the duration of oxygen therapy was lower in infants receiving MIST.


A deficiency in the care of preterm infants with RDS managed on CPAP is that they currently do not receive surfactant unless intubated. This has prompted a search for less-invasive means of surfactant instillation. In this study, we found MIST using a semirigid tracheal catheter could be successfully applied in infants on CPAP, with no significant procedural complications, and instillation of surfactant and return to CPAP in every case. There was an immediate physiological benefit, with a rapid and sustained reduction in FiO2 after MIST. The need for subsequent mechanical ventilation was reduced compared with controls, as was the duration of oxygen therapy.

Our feasibility study allowed the Hobart method of MIST to be evaluated in the hands of 15 different neonatologists and neonatal trainees in two study sites. Given the initial lack of experience with the method, the low proportion of cases requiring a second catheterisation attempt (20%) is evidence of the simplicity of the technique. The narrow bore and semirigid design of the catheter means that, unlike a standard endotracheal tube, it can be passed down the eyeline without obscuring the view of the glottis, and, with an external diameter less than half that of a 2.5 mm endotracheal tube, the MIST catheter passes easily through the vocal cords. Most operators were confident with the technique after two separate trials.

The MIST procedure was generally well tolerated by infants on CPAP, despite receiving no premedication. We elected not to administer narcotic agents so as to avoid the risk of prolonged need for positive pressure ventilation after the procedure. Bradycardia sustained for more than 10 s was noted in around one-third of cases overall, most usually as a view of the vocal cords was being attained by manipulation of the laryngoscope blade. In most instances, the heart rate normalised simply by halting the procedure momentarily. Where needed, positive pressure inflations were generally given after surfactant instillation as an aid to the resumption of respiratory effort. Keeping nasal prongs in situ appears to help restore spontaneous breathing after MIST.

Surfactant administration via MIST was associated with a more rapid and pronounced improvement in oxygenation than has been described after surfactant therapy in intubated infants,21 and indicates very effective and widespread reversal of atelectasis. The resultant reduction in pulmonary vascular resistance, and potential for left to right shunt through the ductus arteriosus,22 may explain the increase in the incidence of PDA in the 25–28-week gestation infants receiving MIST compared with controls.

The criteria used to identify infants eligible for MIST selected a group with significant RDS, who were at risk of failing CPAP, and of the adverse outcomes that appear to follow.16 ,23 ,24 Administration of surfactant via MIST appeared to reduce these risks, with the proportion of infants requiring intubation <72 h approximately halved compared with controls, with a significant difference noted in the 25–28-week gestation range. We also found that MIST appeared to have a lasting effect on lung function, with a shorter duration of oxygen therapy in both gestation ranges.

On the basis of our results, and those of others,15 large-scale randomised controlled trials of MIST appear warranted. Such trials need to be powered to give definitive information about the place of MIST, in particular whether the reduction in the need for mechanical ventilation that is already apparent translates into improved clinical outcomes.25 A trial comparing the Hobart method of MIST with continuation of CPAP is now underway in infants 25–28 weeks gestation (OPTIMIST-A trial, ACTRN12611000916943).26 A delivery room trial comparing surfactant delivery via flexible feeding tube and standard endotracheal tube is also proceeding (NINSAPP trial, NCT00751959). These investigations should guide recommendations and guidelines regarding the use of MIST.

In conclusion, surfactant delivery via a narrow-bore tracheal catheter is feasible and potentially effective, and deserves further investigation in clinical trials.


The authors thank the staff in the Neonatal Units at our two hospitals for their assistance in performing the MIST procedures, Richard Dalton for data collection in controls, and Francesca Orsini for assisting with statistical analysis.


View Abstract


  • Funding Supported by Clinical Grants 11-382 and 12-028 from the Royal Hobart Hospital Research Foundation, and Program Grant 1005345 from the National Health and Medical Research Council, Australia. The funding sources had no role in the study design, data collection, data analysis, preparation of the report or decision to submit the manuscript.

  • Competing interests None.

  • Ethics approval University of Tasmania Human Research Ethics Committee.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Data sharing statement Data from this manuscript will be made available for authorised and approved research activities.

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.