Henry 2001.

Methods	RCT Number of centres: 1 (The Ohio State University College of Dentistry, USA) Recruitment period: not stated Design: parallel group 3‐arm RCT
Participants	Adults presenting for emergency treatment Group 1 (penicillin) Mean age 37 years (SD 16.5 years). Gender: 10 women, 9 men. Median baseline pain (SD): 2.00 (2.00). Median baseline percussion pain (SD): 2.00 (2.00). Median baseline swelling (SD): 1.00 (2.00) Group 2 (placebo) Mean age 38 years (SD 18.8 years). Gender: 10 women, 12 men. Median baseline pain (SD): 2.00 (1.00). Median baseline percussion pain (SD): 2.00 (2.00). Median baseline swelling (SD): 0 (1.00) Included participants had a symptomatic necrotic tooth and actively had spontaneous pain. To be eligible the affected tooth had to test negative to an electric pulp test (Analytic Technology Corp, Redmond, WA) and ice; have a periapical radiolucency and not have had previous endodontic treatment. Included participants were in good health (as determined by written and verbal history), had not received antibiotics in the 30 days prior to enrolment to the trial and did not have a probable or actively draining sinus tract Number of participants at randomisation: not stated in paper, approximately 51 (from personal communication) Number of participants included in the analysis: group 1 = 19; group 2 = 22
Interventions	Endodontic treatment: all participants underwent total pulpectomy of the affected tooth on day 0. Canals were prepared using a step‐back preparation and K‐type files (LD Caulk, Inc, Milford, DE) and irrigated with 2.62% hypochlorite. Following instrumentation, canals were dried and a temporary restoration placed (CavitTM (a light‐cured temporary sealing compound for temporary restoration of cavities)) Participants were then assigned to a trial arm Group 1: oral penicillin (phenoxymethyl) VK 500 mg, 6‐hourly for 7 days Group 2: oral matched placebo taken according to the same regimen Analgesics: all participants received a supply of ibuprofen and were advised to take 400 mg (2 x 200 mg tablets) every 4‐6 hours, as required. Each participant also received a labelled bottle of paracetamol (acetaminophen) with codeine (30 mg), which they were instructed to take 1 or 2 tablets every 4 hours only if 2 ibuprofen tablets did not relieve their discomfort
Outcomes	Primary outcomes Participant‐reported pain, percussion pain and swelling experience at the baseline visit and upon rising for 7 days after treatment on categorical scales. Participants received a 7‐day diary to record postoperative symptoms upon rising each day. This was returned at the obturation appointment (typically the end of root canal treatment). Pain was assessed using a short ordinal numerical scale from 0 to 3: 0 = no pain; 1 = mild pain; 2 = moderate pain; 3 = severe pain. Participants used the same scale to rate pain to percussion (achieve by tapping the affected tooth with a finger). Swelling was assessed on a similar ordinal numerical scale from 0 to 3: 0 = no swelling; 1 = mild swelling, a mild puffiness that was not bothersome; 2 = moderate swelling that caused facial distortion and was bothersome; 3 = a severe swelling that caused serious facial distortion and was very bothersome Secondary outcomes The number and type of pain medication taken
Notes	Funding source: Graduate Endodontic Student Research Fund and Goldberg Memorial Fund, Graduate Endodontics, College of Dentistry, The Ohio State University Sample size calculation: not stated
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "Each patient was assigned a 5‐digit random number from a random number table before the experiment" (email from author)
Allocation concealment (selection bias)	Low risk	Quote: "The study investigator only gave the labelled bottle to the subject without knowing the content because he only saw the random numbers not the assignment of the drug" (email from author)
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Quote: "Each 500 mg gelatin capsule of either penicillin or placebo was identical in form. The 500 mg tablets of penicillin VK were ground into a powder and placed into clear gelatin capsules. The white powder of the lactose placebo was indistinguishable from the white powder of the penicillin tablets when viewed through the capsule"
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Primary outcome measures were participant‐assessed and it was highly unlikely blinding was broken
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Insufficient reporting of relative attrition rates and reasons for withdrawal precludes judgement of the risk of bias
Selective reporting (reporting bias)	Low risk	All outcomes were reported
Other bias	Low risk	No other sources of bias were identified

RCT: randomised controlled trial; SD: standard deviation; VAS: visual analogue scale.