Kramer 2008

Methods	Study design: prospective, randomized double‐blind placebo control study. Study dates and duration: January 2004 to January 2006. Method of temperature measurement: children > 2 years oral, Children < 2 years rectal. Parents given thermometers for home use. Time points measured in study:temperature measurements at hours 0, 3, 4, 5, 6.
Participants	Number: 40 randomized. Number of patients in each intervention: paracetamol n=19, paracetamol alternating with ibuprofen n=19. Inclusion criteria: healthy children presenting to the out patient clinic with chief complaint of fever. Fever in clinic > 38 °C Exclusion criteria: history of any antipyretic use in the preceding 4 hours or if they had an allergy or other medical contraindication to the medications. Baseline characteristics: Sex, N(%): Aal (n=38), males = 18 (47.4), females = 20 (52.6) Diagnoses, N(%) bacterial illness 13 (34.2), viral illness 25 (65.8).
Interventions	Group A: paracetamol (15 mg/kg) alternated with placebo Group B: paracetamol (15 mg/kg) alternating with Ibuprofen (10 mg/kg) Administration regime: Time                     Group A                             Group B 0                           APAP                                 APAP 3                           placebo                              ibuprofen 4                           APAP                                 placebo
Outcomes	Primary: Temperature at enrolment and hours 3, 4, 5, 6 Secondary: Symptom checklist at hours 3 and 4 Parental perception of efficacy at hours 3 and 4
Notes	Location: Washington, USA Setting: single centre: pediatric clinic at Madigan Army Medical centre in Tacoma, Washington. Funding: Resident Research Grant from the American Academy of Pediatrics.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Children were assigned to treatment group A or B using previously generated computer based randomization blocks performed by the Department of Clinical investigation.
Allocation concealment (selection bias)	Unclear risk	Each caretaker received a sealed envelope containing their randomization sequence. No mention of the envelope being opaque.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Parents and investigators remained blinded to the regimen each child had received. Pharmacist was unblinded, but did not assess patients.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Parents and investigators who measured temperature remained blinded to the regimen each child received.
Incomplete outcome data (attrition bias) All outcomes	Low risk	All subjects were accounted for. Loss to follow up: alternating group 0.5%, paracetamol 0.5%.
Selective reporting (reporting bias)	Low risk	All assessed outcomes were reported.
Other bias	Low risk