Marcial 1993.

Methods	Location of trial: US. Number of centres: approximately 7. Funding: National Cancer Institute (grants CA12258 and CA32115). Trial ID: RTOG trial (number unclear).
Participants	Inclusion criteria: patients with all stages of untreated cancer of the tonsillar fossa. Exclusion criteria: patients aged > 80 years, with adenocarcinoma, other cancers (previous or present except for skin cancers), presence of distant metastases, medical conditions which made treatment completion unlikely, or patient deemed unlikely to complete follow‐up. Recruitment period: 1971 to 1976. OP: 147/147 (100%). Number randomised: 147. Number analysed: 137.
Interventions	Accelerated split course radiotherapy versus continuous radiotherapy Split course (n = 63): Phase 1: 3 Gy per fraction, 10 fractions over 2 weeks total of 30 Gy. 3 weeks rest. Phase 2: 3 Gy per fraction and further 10 fractions over 2 weeks. Continuous (n = 74): 2.0‐2.2 Gy per fraction, with 30‐33 fractions over 6 weeks to a total dose of 66 Gy. Original protocol was modified to allow 2 Gy fractions to a total of 60‐66 Gy. Spinal cord protection was required after 50 Gy in Phase 2. Source was teletherapy energy at 1 MeV or higher with minimal source skin distance of 75 cm. Surgical salvage was permitted at least 2 months after completion of radiotherapy.
Outcomes	Primary: tumour clearance. Secondary: locoregional control, overall survival, acute and late toxicity. Duration of follow‐up: minimum of 7 years (4% of patients lost to follow‐up prior to 7 years).
Notes	Dichotomous data only; unable to calculate HR. Locoregional control data presented as percentage but unclear of denominator.
Risk of bias
Bias	Authors' judgement	Support for judgement
Adequate sequence generation?	Low risk	Randomisation done at RTOG headquarters in Philadelphia and was stratified by institution, T stage, N stage and gender. No details on method of sequence generation provided but it is assumed to be adequate.
Allocation concealment?	Low risk	Treatment allocation made by telephone call to RTOG headquarters.
Blinding ‐ Outcome Assessors	High risk	Not mentioned.
Incomplete outcome data addressed?	Unclear risk	10 patients (7%) excluded. 17 reasons for exclusions (3 cancellation, 3 ineligible, 11 no data) for 10 people ‐ unclear how many and which treatment group they were from.
Free of selective reporting?	Low risk	Important outcomes of tumour response, locoregional control, overall survival and toxicity reported.
Free of other bias?	Low risk	No significant differences between the groups at baseline. No other apparent bias.