Quality and transparency of reporting derivation and validation prognostic studies of recurrent stroke in patients with TIA and minor stroke: a systematic review

Table 4 Reporting of essential items relevant to validation studies with updating only

Characteristic	Number (%) N = 23
Rationale provided for updating the model
Yes	23 (100.0)
Authors showed that less extensive update methods were inadequate prior to considering more extensive revisions^a
Yes	2 (8.7)
If the model is derived or updated by methods (c) to (f), was any method of shrinkage of predictor weights or regression coefficients applied^b
Uniform shrinkage	0 (0.0)
Penalized estimation	0 (0.0)
Other (method by van Houwelingen)	1 (3.7)
Unclear or not reported	26 (96.3)
Results from model updating (i.e. model specification, model performance, recalibration) presented (i.e. updated intercept, regression coefficients, discrimination with CI or SE, calibration)
Yes	13 (56.5)

^aLess extensive updates: (a) Recalibrating the intercept only, (b) Recalibrating the intercept and adjust the other regression coefficients by a common factor, (c) Category b plus extra adjustment of a subset of the existing coefficients to a different strength, and (d) Category c plus adding new predictors. Extensive revisions: (e) Re-estimating all of the original regression coefficients and (f) category e plus adding new additional predictors
^bThe denominator is 27 (N = 4 derivations and N = 23 external validation with updating)

ISSN: 2397-7523