This multi-institutional cohort study spanning nine years (2005–2014) involved considerable variation in clinical protocols for both REPs and apexification procedures. As acknowledged by the authors, the lack of consistent irrigation protocols, intracanal medicament selection, and treatment timing introduces confounding variables that were not fully controlled in the statistical analysis. While the study employed multivariate modeling to assess predictors of radiographic root area (RRA) change, the influence of these treatment heterogeneities on both RRA and resorption healing outcomes remains insufficiently addressed.
Furthermore, the retrospective nature of part of the dataset and the non-randomized assignment of treatment types (REPs vs APEX) based on clinical judgment, particularly with respect to trauma severity (e.g., avulsion vs non-avulsion), introduces potential selection bias. For example, MTA apexification was more commonly used in cases with severe trauma, such as avulsion, which naturally correlates with worse prognosis. As such, the more favorable radiographic healing outcomes seen with REPs may be confounded by their preferential use in less severe cases, rather than the inherent superiority of the technique.