Kugelman 2007.

Methods	Study design: randomized controlled prospective clinical trial Setting: single‐center study at Bnai Zion Medical Center in Israel Duration of study: September 2004‐April 2006
Participants	Inclusion criteria: infants between 24 and 34 6/7 weeks' gestational age with RDS, which study authors defined as clinical features such as tachypnea, grunting, nostril flaring, and retractions, as well as a positive chest x‐ray Exclusion criteria: cardiac disease, congenital malformation, sepsis, anemia, severe IVH, refusal of consent, ventilatory unavailability Number randomized: 84 infants total (53 males, 31 females)
Interventions	Both modes of respiratory support were administered via nasal prongs (INCA, Ackrad Laboratories, Berlin, Germany) by the SLE 2000 Ventilator (Specialized Laboratory Equipment, Croydon, UK). NCPAP (n = 41) was administered at 6‐7 cmH2O, and NIPPV (n = 43) was given with PIP 14‐22 cmH2O (adjusted according to chest excursion and birth weight), positive end expiratory pressure 6‐7 cmH2O, and 12‐30 breaths per minute. FiO2 was adjusted to keep oxygen saturation (measured by pulse oximetry) between 88% and 92%. NIPPV was synchronized.
Outcomes	Primary outcome: failure of nasal respiratory support (i.e. need for endotracheal intubation). Criteria for failure: worsened RDS in conjunction with at least 1 of the following: pH < 7.2, PCO2 > 60 mmHg, PO2 < 50 mmHg, arterial oxygen saturation SpO2 < 88% on FiO2 > 50%, recurrent significant apnea needing stimulation or bag and mask ventilation Secondary outcomes: blood pressure, heart rate, respiratory rate, pulse oximetry saturation, respiratory status before mechanical ventilation, time to stop nasal support (stopped if FiO2 < 30% and no clinical signs of respiratory distress), incidence of IVH, duration of mechanical ventilation, incidence of BPD, time until full feeds, and length of hospital stay
Notes	Two infants were switched from NCPAP to NIPPV treatment group after randomization. Upon request, study authors provided raw data for outcomes of need for intubation and chronic lung disease for performance of a sensitivity analysis excluding those 2 infants. Results remained significant upon sensitivity analysis.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Study authors stated that they generated a random sequence order and performed block randomization to provide similar numbers of infants in each treatment group for the subgroup of birth weights above and below 1500 grams.
Allocation concealment (selection bias)	Low risk	Study authors placed individual cards with sequence designations in sealed opaque envelopes.
Blinding (performance bias and detection bias) All outcomes	High risk	Blinding was impossible owing to the nature of the interventions.
Incomplete outcome data (attrition bias) All outcomes	Low risk	No missing data
Selective reporting (reporting bias)	Low risk	Primary and secondary outcomes listed in the protocol were published in this manuscript or in a subsequent manuscript in the journal Acta Paediatrica.
Other bias	High risk	Two infants were switched from NCPAP to NIPPV in violation of the study protocol.