Table 5.
Risk of bias assessment using the PROBAST
| Signaling question | Yes or probably yes | No information | No or probably no |
|---|---|---|---|
| Domain 1: participants | |||
| Were appropriate data sources used? | 21 (100%) | 0 (0%) | 0 (0%) |
| Were all inclusions and exclusions of participants appropriate? | 19 (90.48%) | 2 (9.52%) | 0 (0%) |
| Domain 2: predictors | |||
| Were predictors defined and assessed in a similar way for all participants? | 18 (85.71%) | 1 (4.76%) | 2 (9.52%) |
| Were predictor assessments made without knowledge of outcome data? | 7 (33.33%) | 12 (57.14%) | 2 (9.52%) |
| Are all predictors available at the time the model is intended to be used? | 21 (100%) | 0 (0%) | 0 (0%) |
| Domain 3: outcome | |||
| Was the outcome determined appropriately? | 17 (80.95%) | 4 (19.05%) | 0 (0%) |
| Was a prespecified or standard outcome definition used? | 1 (4.76%) | 20 (95.24%) | 0 (0%) |
| Were predictors excluded from the outcome definition? | 21 (100%) | 0 (0%) | 0 (0%) |
| Was the outcome defined and determined in a similar way for all participants? | 1 (4.76%) | 20 (95.24%) | 0 (0%) |
| Was the outcome determined without knowledge of predictor information? | 1 (4.76%) | 20 (95.24%) | 0 (0%) |
| Was the time interval between predictor assessment and outcome determination appropriate? | 17 (80.95%) | 4 (19.05%) | 0 (0%) |
| Domain 4: analysis | |||
| Were there a reasonable number of participants with the outcome? | 1 (4.76%) | 1 (4.76%) | 19 (90.48%) |
| Were continuous and categorical predictors handled appropriately? | 21 (100%) | 0 (0%) | 0 (0%) |
| Were all enrolled participants included in the analysis? | 20 (95.24%) | 0 (0%) | 1 (4.76%) |
| Were participants with missing data handled appropriately? | 2 (9.52%) | 19 (90.48%) | 0 (0%) |
| Was selection of predictors based on univariable analysis avoided? | 16 (76.19%) | 0 (0%) | 5 (23.81%) |
| Were complexities in the data accounted for appropriately? | 21 (100%) | 0 (0%) | 0 (0%) |
| Were relevant model performance measures evaluated appropriately? | 8 (38.1%) | 0 (0%) | 13 (61.9%) |
| Were model overfitting, underfitting, and optimism in model performance accounted for? | 7 (33.33%) | 0 (0%) | 14 (66.67%) |
| Do predictors and their assigned weights in the final model correspond to the results from the reported multivariable analysis? | 21 (100%) | 0 (0%) | 0 (0%) |