Parameter values used within the base case model analysis
| Model parameter | Parameter estimate | Sources used to inform model parameter |
|---|---|---|
| *Personal communication, Dr Julietta Patnick, Director of NHS Cancer Screening Programmes; †personal communication, Professor Neil Shepherd, Consultant Pathologist, Royal Gloucestershire Hospital. | ||
| COL, colonoscopy; CRC, colorectal cancer; FOBT, faecal occult blood test; FSIG, flexible sigmoidoscopy. | ||
| Test characteristics | ||
| FOBT sensitivity for polyps | 5.00% | Assumption |
| FOBT sensitivity for CRC | 40.58% | Allison et al42,43 |
| FOBT specificity | 98.50% | Allison et al41,43 |
| FSIG and COL sensitivity for low-risk distal polyps | 76.00% | Bressler et al,44 Hixson et al,45 Rex et al46 |
| COL sensitivity for low-risk proximal polyps | 76.00% | Bressler et al,44 Hixson et al,45 Rex et al46 |
| FSIG and COL sensitivity for high-risk distal adenomas and CRC | 97.00% | Bressler et al,44 Hixson et al,45 Rex et al46 |
| COL sensitivity for high-risk proximal polyps and CRC | 94.00% | Bressler et al,44 Hixson et al,45 Rex et al46 |
| COL and FSIG specificity | 100.00% | Assumption |
| Natural history parameters | ||
| Probability of distal polyp given proximal cancer | 28.00% | Dinning et al47 |
| Normal epithelium to low-risk polyp (men and women) | 1.60% | Eide et al,26 Rickert et al,27 Williams et al,28 |
| Blatt,29 Vatn et al,30 | ||
| Arminski et al,31 model calibration | ||
| Low-risk polyp to high-risk polyp | 2.12% | Knoernschild,25 model calibration |
| High-risk polyp to Dukes’ A | 3.26% | Stryker et al,24 model calibration |
| Dukes’ A to Dukes’ B | 58.29% | Model calibration |
| Dukes’ B to Dukes’ C | 65.55% | Model calibration |
| Dukes’ C to stage D | 86.48% | Model calibration |
| Probability of recurrence given history of low-risk polyp (year 1) | 18.00% | Winawer et al33 |
| Probability of recurrence given history of low-risk polyp (year 2+) | 5.00% | Winawer et al33 |
| Probability of recurrence given history of high-risk polyp (year 1) | 25.00% | Winawer et al33 |
| Probability of recurrence given history of high-risk polyp (year 2+) | 6.00% | Winawer et al33 |
| Probability of presenting symptomatically with Dukes’ A | 7.00% | Model calibration |
| Probability of presenting symptomatically with Dukes’ B | 32.00% | Model calibration |
| Probability of presenting symptomatically with Dukes’ C | 49.00% | Model calibration |
| Probability of presenting symptomatically with stage D | 85.40% | Model calibration |
| Annual CRC-specific mortality rate (Dukes’ A) | 0.00% | Model calibration |
| Annual CRC-specific mortality rate (Dukes’ B) | 1.00% | Model calibration |
| Annual CRC-specific mortality rate (Dukes’ C) | 6.02% | Model calibration |
| Annual CRC-specific mortality rate (stage D) | 38.67% | Model calibration |
| Harm parameters | ||
| COL probability of perforation (without polypectomy) | 0.08% | Atkin et al11 |
| COL probability of perforation (with polypectomy) | 0.17% | Atkin et al11 |
| COL probability of death following perforation | 5.82% | Gatto et al35 |
| FSIG probability of perforation (without polypectomy) | 0.0025% | Atkin et al11 |
| FSIG probability of perforation (with polypectomy) | 0.0025% | Atkin et al11 |
| FSIG probability of death following perforation | 5.82% | Gatto et al35 |
| Probability of bleeding following FSIG | 0.0295% | Atkin et al11 |
| Probability of bleeding following COL | 0.439% | Atkin et al11 |
| Screening participation parameters | ||
| FOBT participation rate | 60.00% | Hardcastle,3 UK Colorectal Cancer Screening Pilot Group |
| FSIG compliance | 60.00% | Assumption based on Atkin et al11 |
| COL compliance | 80.00% | Lund et al36 |
| Health-related quality of life parameters | ||
| Utility cancer free | 0.91 | Ness et al37 |
| Utility Dukes’ A | 0.74 | Ness et al37 |
| Utility Dukes’ B | 0.70 | Ness et al37 |
| Utility Dukes’ C | 0.50 | Ness et al37 |
| Utility stage D | 0.25 | Ness et al37 |
| Resource use parameters | ||
| FSIG probability of inadequate bowel preparation | 5.26% | Atkin et al11 |
| COL probability of inadequate bowel preparation | 10.00% | Assumption |
| Cost of FSIG (with/without polypectomy) | £51.60 | Whynes et al38 |
| Cost of FOBT (2 tests) | £11.74 | Personal communication* |
| Cost of COL | £188.40 | NHS Reference Costs39 |
| Cost of treating bowel perforation (major surgery) | £5407.74 | NHS Reference Costs39 |
| Cost of admittance for bleeding | £250.21 | NHS Reference Costs39 |
| Pathology cost for adenoma | £30.00 | Personal communication† |
| Pathology cost for cancer | £250.00 | Personal communication† |
| Lifetime cost of Dukes’ A | £8299.24 | Tappenden et al18 |
| Lifetime cost of Dukes’ B | £12 441.41 | Tappenden et al18 |
| Lifetime cost of Dukes’ C | £19 076.90 | Tappenden et al18 |
| Lifetime cost of stage D | £11 945.78 | Tappenden et al18 |