Table 2.
Study characteristics of the included publications, by study type
| Study characteristics | Development only (n=48) | Development and external validation (n=14) | All (n=62) |
|---|---|---|---|
| n | n | n | |
| Cancer type | |||
| Lung | 6 | 2 | 8 |
| Breast | 6 | - | 6 |
| Colon/colorectal/rectal | 3 | 3 | 6 |
| Pancreatic | 1 | 2 | 3 |
| Liver | 2 | - | 2 |
| Gastric | 3 | - | 3 |
| Head and neck | 5 | - | 5 |
| Spinal | 4 | - | 4 |
| Brain (including meningioma, glioblastoma) | 4 | 1 | 5 |
| Oral (including nasopharyngeal carcinoma) | 2 | 1 | 3 |
| Gynaecological (including cervical, ovarian, endometrial) | 5 | 1 | 6 |
| Prostate/Penile | 4 | 1 | 5 |
| Skin (including melanoma) | 1 | 1 | 2 |
| Other* | 2 | 2 | 4 |
| Outcome type | |||
| Binary | 40 | 8 | 48 |
| Continuous | - | 1 | 1 |
| Multinomial | 2 | - | 2 |
| Time to event | 6 | 5 | 11 |
| Data source/study design⁎⁎ | |||
| RCT | - | 1 | 1 |
| Prospective cohort | 9 | - | 9 |
| Retrospective cohort | 11 | 3 | 14 |
| Registry | 15 | 6 | 21 |
| Routine care database | 7 | 2 | 9 |
| Other⁎⁎⁎ | 2 | 1 | 3 |
| Unclear | 4 | 1 | 5 |
| Intended user | |||
| Health care providers | 27 | 7 | 34 |
| Public/patients | 2 | - | 2 |
| Researchers | 1 | - | 1 |
| Health care providers and patient/public | 1 | 3 | 4 |
| Health care providers and researchers | 2 | - | 2 |
| Unclear | 15 | 4 | 19 |
| Aim of model | |||
| Predict risk | 25 | 11 | 36 |
| Predict length of stay (continuous outcome) | - | 1 | 1 |
| Classify patients | 23 | 2 | 25 |
| Sufficient information to apply the model and make predictions or classifications | |||
| No | 42 | 10 | 52 |
| Yes | 6 | 4 | 10 |