We use this method to prevent deformities in children with bone fragility, but we cannot explain the risks of the growth perturbations in long term follow-up period.
The safety limit is around 20% of the transverse diameter of the growth plate. Nordentoft, EL: Experimental epiphyseal injuries. Acta Orthop Scand 40: 176-192, 1969
This is a difficult question. 20% seems to be a reasonable percentage. But here one can discuss on the possibility of an asymmetrical damage to the growth plate and the consequent frontal or sagittal deviations around the adjacent joints.
In any case it is better if the physis is not distrurbed by osteosynthesis material at all. In little children one can use a bundle of K-wires which are not so harmful for the physes.