Maor 1986.

Methods	Location of trial: US. Number of centres: 4. Funding: National Cancer Institute (core grant CA23113). Trial ID: RTOG 7808.
Participants	Inclusion criteria: untreated squamous cell carcinoma, T2‐T4, with any N but M0. Patients referred to trial following unsatisfactory response to initial radiotherapy. Exclusion criteria: more than 1 primary tumour or Karnofsky status < 50. Recruitment period: October 1978 to August 1982. OC: 30/115 (26%). OP: 59/115 (51%). OC+OP: 89/115 (77%). Number randomised: 118. Number analysed: 115.
Interventions	Neutron boost versus photon boost Neutron boost (n = 57): to include only areas involved by gross tumour, primary, or nodes plus a margin of 2 cm. Boost given in 4‐6 fractions in 2‐3 weeks. Neutron dose depended on radiobiological effectiveness, equivalent to 25‐30 Gy photons. Photon boost (n = 58): to include areas with gross tumour plus a 1 cm margin (2 cm with cobalt). 25‐30 Gy over 2‐3 weeks in 5 daily fractions per week. All patients received 45‐50 Gy photons in daily fractions of 1.8‐2 Gy.
Outcomes	Primary outcome unclear. Tumour clearance, locoregional control, overall survival.
Notes	Data taken from Kaplan‐Meier graphs (no numbers at risk).
Risk of bias
Bias	Authors' judgement	Support for judgement
Adequate sequence generation?	Low risk	Randomisation stratified by institution, T stage and region and performed by central office. No details given but assumed adequate as for other RTOG trials.
Allocation concealment?	Unclear risk	Insufficient information.
Blinding ‐ Outcome Assessors	High risk	Not mentioned.
Incomplete outcome data addressed?	Low risk	3/118 excluded from the analysis. Unlikely to affect results.
Free of selective reporting?	Low risk	Important outcomes of overall survival, local clearance, locoregional control planned and reported.
Free of other bias?	High risk	1 patient in neutron boost group and 5 in photon boost group received interstitial implants to deliver boost.