Jain 2000a.

Methods	Single‐centre randomised controlled trial performed in the UK.
Participants	40 newborn infants. Gestation 27 to 41 weeks (median 33) at 2 to 17 postnatal age (median 7). Inclusion criteria: infants admitted to the postnatal wards or NICU undergoing routine venepuncture in a hospital. Exclusion criteria: infants who were unwell, ventilated or sedated.
Interventions	Intervention: amethocaine 1.5 g portion (4%). Placebo: 1.5 g gel. Both gels were applied to the skin under an occlusive dressing for 60 minutes, then wiped off and venepuncture performed 5 minutes later.
Outcomes	Pain: pain response to needle insertion was assessed using a validated adaptation of the NFCS (NFCS‐short) and by the presence and length of crying. The NFCS‐short assesses each of the following facial characteristics as present (1 point) or absent (0 points): eye squeeze, brow bulge, open mouth, deepened nasolabial folds and cry. An NFCS score was assigned for each second, starting 5 seconds before the needle insertion and ending 5 seconds after needle insertion (giving a maximum cumulative score of 25 for each 5‐second period). The authors defined a cumulative NFCS score of 10 or less in the 5 seconds after insertion of the needle as indicating no or minimal pain. Incidence of crying. Duration of crying. Successful needle prick at first attempt. Skin reactions.
Notes	Power calculation not reported.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Infants were randomised to receive amethocaine or placebo gel. Randomisation was stratified within 3 gestation age groups. No other information provided.
Allocation concealment (selection bias)	Low risk	Pharmacist allocated treatment groups. The gels were packed in identical tubes by the hospital pharmacy who randomised and coded them. No other information provided.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Blinded to treatment staff and researchers. Gels in identical tubes.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	The code was only broken at the end of the study after the videotapes had been scored and when the method of defining a painful or non‐painful response had been agreed on.
Incomplete outcome data (attrition bias) All outcomes	Low risk	1 infant excluded before analysis before the code was broken because of restlessness before the venepuncture (< 10% attrition).
Selective reporting (reporting bias)	Unclear risk	Protocol not obtained.
Other bias	Low risk