Harris 2013.

Study characteristics
Methods	Randomised controlled trial
Participants	514 infants ≥ 35 weeks' gestation, < 48 postnatal hours old and at risk for hypoglycaemia. Risk factors included mother with diabetes, small (birthweight < 10th centile, or < 2500 grams) or large (birthweight > 90th centile or > 4500 grams) size, preterm (35 or 36 weeks' gestation) birth and other reasons such as poor feeding. Of these, 237 became hypoglycaemic and were randomised (5 infants randomised in error). The majority of infants were not admitted to NICU.  Setting: New Zealand Timing   Trial: 1 December 2008 to 31 November 2010 Follow‐up at two years: 21 July 2010 to 30 January 2013 Follow‐up at 4.5 years: September 2011 to June 2015.
Interventions	Infants who became hypoglycaemic (< 2.6 mmol/L) were encouraged to feed (determined by maternal choice) and were randomised to receive 40% oral dextrose gel (0.5 mL/kg) (n = 118) or placebo gel (n = 119) massaged into the buccal membrane. Blood glucose concentration was measured 30 minutes following gel treatment. Gel was repeated if hypoglycaemia persisted. A maximum of six doses of gel could be given within a 48‐hour period.
Outcomes	Primary outcomes Treatment failure defined as blood glucose concentration ≤ 2.6 mmol/L, 30 minutes after the second of two treatment doses of gel Neurosensory impairment at 4.5 years' corrected age (any of: cerebral palsy; visual impairment; deafness; IQ < 85; Beery visual‐motor integration score < 85; Movement ABC score < 15th centile; low executive function or motion coherence threshold (worse than 1.5 SD from the mean))* Secondary outcomes Admission to the newborn intensive care unit Frequency of breastfeeding Total volume and frequency of expressed breast milk Total volume and frequency of infant formula Total volume and frequency of intravenous dextrose Total volume and frequency of oral dextrose gel Method of feeding 2 weeks after discharge Incidence of rebound and recurrent hypoglycaemia after successful treatment (defined as blood glucose > 2.6 mmol/L 30 minutes after treatment) Total duration of interstitial glucose concentrations < 2.6 mmol/L up to 48 hours after birth Visual impairment at 4.5 years' corrected age* Hearing impairment at 4.5 years' corrected age* Cerebral palsy at 4.5 years' corrected age* WPPSI‐3 full and index scale scores at 4.5 years' corrected age* Executive function composite score at 4.5 years' corrected age* BRIEF‐P Global Executive Composite t‐score at 4.5 years' corrected age* CBCL t‐score at 4.5 years' corrected age* SDQ total difficulties score at 4.5 years' corrected age*
Notes	*Data received from study authors. Blood glucose measured using glucose oxidase method. Funding sources included Waikato Medical Research Foundation, Auckland Medical Research Foundation, the Maurice and Phyllis Paykel Trust, the Health Research Council of New Zealand and the Rebecca Roberts Scholarship. Disclosure: 3 of the authors of this review (PJW, DLH and JE Harding) were involved in the design and conduct of this study.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Used computer‐generated blocked randomisation, with variable block sizes.
Allocation concealment (selection bias)	Low risk	Central allocation by computer which provided a randomisation number corresponding to a numbered treatment pack.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Dextrose and placebo gels were identical in appearance. Clinicians, families and study investigators were masked to treatment allocation until completion of data analysis.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Outcome assessors were masked to treatment allocation.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Performed an intention‐to‐treat analysis. Five infants were randomised in error and were excluded from the analysis, leaving 118 infants in the oral dextrose gel group and 119 infants in the placebo gel group. Of the 237 randomised, 78% were followed up and assessed at 4.5 years' corrected age (96/118 oral dextrose gel group and 89/119 in the placebo gel group). Children assessed at 4.5 years were more likely to be exposed to alcohol and smoking during pregnancy, be of Māori ethnicity, have a lower minimum interstitial blood glucose concentration and be admitted to NICU compared with children not assessed. Those assessed were also less likely to be a singleton and of an ethnicity other than Māori or New Zealand European. All other maternal and infant characteristics were similar in those who were and were not assessed at 4.5 years.
Selective reporting (reporting bias)	Low risk	The only outcome not listed a priori was acceptability of the intervention. Otherwise, all prespecified outcomes were reported.
Other bias	Low risk	A slightly greater number of mothers in the group allocated to oral dextrose gel than to placebo gel intended to breastfeed (114 of 115 vs 109 of 115). Fewer boys were allocated to the oral dextrose gel group than to the placebo gel group (48 of 118 vs 65 of 119). Groups were otherwise balanced.