Table 2

Long-term colorectal cancer (CRC) incidence in the intermediate-risk group by baseline characteristics and number of surveillance visits

N (%)Person-yearsCRC casesIncidence per 100 000 person-years (95% CI)Univariable HR (95% CI)P valueMultivariable HR (95% CI)P value
Total11 852 (100)111 270246221 (195 to 251)
Number of surveillance visits*0.00040.0009
 04683 (39.5)53 927135250 (211 to 296)11
 13343 (28.2)33 28462186 (145 to 239)0.58 (0.42 to 0.79)0.59 (0.43 to 0.81)
 22279 (19.2)15 47731200 (141 to 285)0.53 (0.35 to 0.81)0.56 (0.36 to 0.85)
 ≥31547 (13.1)858218210 (132 to 333)0.44 (0.26 to 0.77)0.45 (0.26 to 0.77)
Sex0.280.0549
 Women5271 (44.5)51 049105206 (170 to 249)11
 Men6581 (55.5)60 221141234 (199 to 276)1.15 (0.89 to 1.48)1.28 (0.99 to 1.66)
Age, years<0.0001<0.0001
 <552097 (17.7)24 99528112 (77 to 162)11
 55–643158 (26.7)33 53052155 (118 to 204)1.44 (0.91 to 2.28)1.41 (0.89 to 2.24)
 65–743915 (33.0)35 39198277 (227 to 338)2.74 (1.80 to 4.19)2.66 (1.74 to 4.06)
 ≥752682 (22.6)17 35468392 (309 to 497)4.25 (2.71 to 6.65)3.64 (2.31 to 5.74)
Number of adenomas0.370.20
 17793 (65.8)74 791168225 (193 to 261)11
 23053 (25.8)27 50264233 (182 to 297)1.06 (0.79 to 1.41)0.92 (0.68 to 1.25)
 3 or 41006 (8.5)897714156 (92 to 263)0.71 (0.41 to 1.23)0.61 (0.34 to 1.08)
Adenoma size, mm†0.08660.18
 <101006 (8.5)897714156 (92 to 263)11
 10–196802 (57.4)64 716134207 (175 to 245)1.30 (0.75 to 2.26)1.53 (0.87 to 2.70)
 ≥204044 (34.1)37 57798261 (214 to 318)1.64 (0.94 to 2.88)1.69 (0.94 to 3.04)
Adenoma histology‡<0.00010.0025
 Tubular4694 (39.6)44 36971160 (127 to 202)11
 Tubulovillous5537 (46.7)51 211114223 (185 to 267)1.40 (1.04 to 1.88)1.29 (0.95 to 1.75)
 Villous1134 (9.6)10 10831307 (216 to 436)1.93 (1.27 to 2.95)1.44 (0.93 to 2.24)
 Unknown487 (4.1)558130538 (376 to 769)3.23 (2.10 to 4.98)2.76 (1.64 to 4.64)
Adenoma dysplasia‡0.00020.0038
 Low-grade9399 (79.3)87 581166190 (163 to 221)11
 High-grade1979 (16.7)17 40253305 (233 to 399)1.62 (1.19 to 2.21)1.47 (1.07 to 2.02)
 Unknown474 (4.0)628727429 (295 to 626)2.11 (1.39 to 3.20)1.86 (1.21 to 2.86)
Proximal polyps0.02840.0025
 No8254 (69.6)79 798162203 (174 to 237)11
 Yes3598 (30.4)31 47184267 (216 to 331)1.35 (1.04 to 1.76)1.54 (1.17 to 2.02)
Completeness of colonoscopy0.00070.0022
 Complete8967 (75.7)80 572150186 (159 to 218)11
 Incomplete1321 (11.2)11 54549424 (321 to 562)1.58 (1.22 to 2.06)1.55 (1.18 to 2.06)
 Unknown1564 (13.2)19 15247245 (184 to 327)
Bowel preparation quality0.130.14
 Excellent or good3974 (33.5)37 49371189 (150 to 239)11
 Satisfactory1903 (16.1)15 45136233 (168 to 323)1.28 (0.86 to 1.92)1.47 (0.98 to 2.22)
 Poor660 (5.6)484017351 (218 to 565)1.92 (1.13 to 3.25)1.67 (0.98 to 2.85)
 Unknown5315 (44.8)53 485122228 (191 to 272)1.17 (0.88 to 1.57)1.13 (0.84 to 1.53)
Year of baseline visit0.00440.0078
 1984–19991870 (15.8)25 32983328 (264 to 406)11
 2000–20044222 (35.6)42 95792214 (175 to 263)0.66 (0.48 to 0.90)0.63 (0.46 to 0.87)
 2005–20105760 (48.6)42 98371165 (131 to 208)0.57 (0.40 to 0.80)0.59 (0.40 to 0.85)
Length of baseline visit0.01810.0082
 1 day6697 (56.5)63 453117184 (154 to 221)11
 2 days–3 months2343 (19.8)21 66960277 (215 to 357)1.53 (1.12 to 2.08)1.65 (1.20 to 2.26)
 3–6 months1403 (11.8)13 27732241 (170 to 341)1.32 (0.89 to 1.95)1.34 (0.90 to 1.99)
 ≥6 months1409 (11.9)12 87137287 (208 to 397)1.57 (1.09 to 2.27)1.58 (1.08 to 2.30)
  • The final multivariable model contained number of surveillance visits, age, adenoma dysplasia, proximal polyps, completeness of colonoscopy, year of baseline visit and length of baseline visit. For these variables, the multivariable HRs were from the final multivariable model and the p values were for inclusion of the variable in the model, calculated with the likelihood ratio test. For the remaining variables, the multivariable HRs were for if the variable was added as an additional variable to the final multivariable model.

  • *The number of surveillance visits was included as a time-varying covariate; patients who attended any surveillance contributed person-years to more than one category of number of surveillance visits.

  • †Adenoma size was defined according to the largest adenoma seen at baseline.

  • ‡Adenoma histology and dysplasia were defined according to the worst histology (greatest degree of villousness) and worst dysplasia (highest grade) seen at baseline, respectively.