Sensitivity and scenario analyses
| Cost per 1000 cases (million, US$) | Surgeries avoided per 1000 cases | Incremental cost (US$) per surgery avoided | |
| Base case | |||
| S-ESD | 4.22 | 925 | – |
| U-ESD | 6.91 | 938 | 210 112 |
| WF-EMR | 4.33 | 906 | $ than S-ESD <surgeries avoided than S-ESD |
| Selective ESD using Eastern criteria (lesion assessment sensitivity 66.0%; specificity 60.8%), with base case lesion assessment test performance for WF-EMR | |||
| WF-EMR | 4.33 | 906 | – |
| S-ESD | 5.24 | 932 | 34 881 |
| U-ESD | 6.91 | 938 | 274 271 |
| Higher lesion assessment sensitivity (sensitivity 83%, specificity 70%)30 | |||
| S-ESD | 4.96 | 937 | – |
| U-ESD | 6.91 | 938 | 3 601 967 |
| WF-EMR | 6.76 | 692 | $ than S-ESD <surgeries avoided than S-ESD |
| Lower EMR effectiveness—higher risk of recurrence and recurrence requiring surgery (recurrence colon 25% vs 14.4%, rectum 35% vs 18.9%; requiring surgery—colon and rectum—10% vs 1%) | |||
| S-ESD | 4.55 | 902 | – |
| U-ESD | 6.91 | 938 | 66 768 |
| WF-EMR | 4.66 | 883 | $ than S-ESD <surgeries avoided than S-ESD |
| Lower ESD effectiveness—higher risk of recurrence and recurrence requiring surgery (recurrence colon 10% vs 2%, rectum 15% vs 3%; requiring surgery—colon 10% vs 1.5%, rectum 10% vs 2%) | |||
| S-ESD | 4.23 | 925 | – |
| U-ESD | 7.06 | 928 | 822 607 |
| WF-EMR | 4.33 | 906 | $ than S-ESD <surgeries avoided than S-ESD |
| Lower R0 resection rate (71%)31 | |||
| S-ESD | 4.26 | 922 | – |
| U-ESD | 7.02 | 929 | 379 204 |
| WF-EMR | 4.33 | 906 | $ than S-ESD <surgeries avoided than S-ESD |
| Increased risk (20%) of perforation following ESD requiring surgery | |||
| S-ESD | 4.32 | 920 | – |
| WF-EMR | 4.33 | 906 | $ than S-ESD <surgeries avoided than S-ESD |
| U-ESD | 10.15 | 758 | $ than WF-EMR and S-ESD <surgeries avoided than WF-EMR and S-ESD |
| Colonic lesions only | |||
| S-ESD | 3.93 | 936 | – |
| U-ESD | 6.70 | 946 | 277 897 |
| WF-EMR | 4.05 | 913 | $ than S-ESD <surgeries avoided than S-ESD |
| Rectal lesions only | |||
| S-ESD | 5.63 | 871 | – |
| U-ESD | 7.92 | 897 | 87 066 |
| WF-EMR | 5.68 | 871 | $ than S-ESD no additional surgeries avoided than S-ESD |
| Non-granular rectal lesions only—prevalence SMIC=0.25, % LR-SMIC=0.375 | |||
| S-ESD | 7.40 | 779 | – |
| U-ESD | 9.10 | 832 | 32 132 |
| WF-EMR | 7.48 | 779 | $ than S-ESD no additional surgeries avoided than S-ESD |
| Non-granular Paris 0-Is rectal lesions only—prevalence SMIC=0.57, % LR-SMIC=0.5 | |||
| U-ESD | 11.53 | 699 | – |
| S-ESD | 12.37 | 523 | $ than U-ESD <surgeries avoided than U-ESD |
| WF-EMR | 12.52 | 523 | $ than S-ESD and U-ESD no additional surgeries avoided than S-ESD and U-ESD |
| Varying cost of ESD—lower teshold <$1330 | |||
| U-ESD | <S-ESD | 938 | – |
| S-ESD | U-ESD | 925 | $ than U-ESD <surgeries avoided than S-ESD |
| WF-EMR | 4.33 | 906 | $ than S-ESD and U-ESD <surgeries avoided than S-ESD and U-ESD |
| Varying cost of ESD—upper threshold $6670 | |||
| WF-EMR | 4.33 | 906 | – |
| S-ESD | WF-EMR | 925 | 2.69 |
| U-ESD | S-ESD | 938 | 400 406 |
| Elimination of first surveillance colonoscopy for ESD only | |||
| S-ESD | 4.20 | 925 | – |
| U-ESD | 6.10 | 948 | 148 630 |
| WF-EMR | 4.33 | 906 | $ than S-ESD <surgeries avoided than S-ESD |
| Including non-medical costs | |||
| S-ESD | 5.24 | 925 | – |
| U-ESD | 8.26 | 938 | 236 445 |
| WF-EMR | 5.32 | 906 | $ than S-ESD <surgeries avoided than S-ESD |
ESD, endoscopic submucosal dissection; LR-SMIC, low-risk submucosal invasive cancer ; S-ESD, selective ESD; U-ESD, universal ESD; WF-EMR, wide-field endoscopic mucosal resection.