Table 2.

 Detailed characteristics of surveillance programmes used in five randomised trials of intensive versus control follow up of patients after curative resection for colorectal cancer

Study	Intensive follow up	Control follow up
Makela et al, 199517	Seen in clinic 3 monthly for first 2 years, then 6 monthly: physical examination, full blood count, faecal occult blood test, carcinoembryonic antigen levels, and chest x ray. Yearly colonoscopy. Sigmoidoscopy 3 monthly for rectal and sigmoid cancers. Ultrasonography of liver 6 monthly. Computed tomography yearly. All followed up to 5 years	Seen in clinic 3 monthly for first 2 years, then 6 monthly: physical examination, full blood count, faecal occult blood test, carcinoembryonic antigen levels, and chest x ray. Yearly barium enema. Rigid sigmoidoscopy 3 monthly for rectal cancers. All followed up to 5 years
Ohlsson et al, 199518	Seen in clinic 3 monthly for the first 2 years, then 6 monthly: physical examination, rigid proctosigmoidoscopy, liver function tests, carcinoembryonic antigen levels, faecal occult blood test, chest x ray. Colonoscopy at 3, 15, 30, and 60 months, computed tomography after abdominoperineal resection at 3, 6, 12, 18, and 24 months. All followed up to 5 years	No systematic follow up. Patients were instructed to leave samples for faecal occult blood test testing every third month during the first 2 years and then every year. All accounted for to 5 years
Schoemaker et al, 199819	Seen in clinic 3 monthly for first 2 years, then 6 monthly for 5 years; physical examination, full blood count, liver function tests, and Haemoccult II. Yearly chest x ray and computed tomography of liver. Yearly colonoscopy. Carcinoembryonic antigen measurements were performed but not used to trigger further examinations. 94% followed up to 5 years	Seen in clinic 3 monthly for first 2 years, then 6 monthly for 5 years; physical examination, full blood count, liver function tests, carcinoembryonic antigen levels, and Haemoccult II. Carcinoembryonic antigen measurements were performed but not used to trigger further examinations. 95% followed up to 5 years
Pietra et al, 199820	Seen in clinic 3 monthly for first 2 years, then 6 monthly for next 3 years, thereafter yearly; physical examination, ultrasonography of liver, carcinoembryonic antigen levels. Yearly colonoscopy, chest x ray, and computed tomography. All followed up to 5 years	Seen in clinic 6 monthly for first year, then yearly; physical examination, ultrasonography of liver, carcinoembryonic antigen levels. Yearly colonoscopy and chest x ray. All followed up to 5 years
Kjeldsen et al, 199721	Physical examination, digital rectal examination, gynaecological examination, Haemoccult-II, colonoscopy, chest x ray, full blood count, erythrocyte sedimentation rate, liver function tests, at 6 monthly in first 3 years, then 12 monthly for next 2 years, then 5 yearly. 79% followed up to 5 years	Physical examination, digital rectal examination, gynaecological examination, Haemoccult-II, colonoscopy, chest x ray, full blood count, erythrocyte sedimentation rate, liver function tests, at 5 and 10 years. 73% followed up to 5 years