In Reply.—We appreciate Dr Pavan's letter in reference to our recent report in Pediatrics and congratulate her practice on the efforts to implement standardized developmental screening for their young patients. We agree that practice-based improvement methods, such as those sponsored by the National Initiative for Children's Healthcare Quality, hold tremendous promise in this regard. However, in the case of office-based developmental screening, such efforts (although necessary) may be insufficient.As the success of the National Initiative for Children's Healthcare Quality has taught us, physician behavior change is a systems-level issue. The paradigm that physicians simply have to be educated regarding the importance of a particular problem (eg, developmental screening) is dated, and such efforts thankfully are giving way to more substantive improvement methods such as the learning collaborative. These efforts focus on changing office systems—fundamental issues such as maintaining efficient patient flow, keeping patient registries, and reassessing who performs what role during a patient visit. They emphasize measurement and rapid response and define success as sustained incremental change over time.Although we know that standardized developmental screening tools are relatively sensitive and specific for identifying children with developmental problems, our study found that 71% of pediatricians use clinical assessment alone to assess children's developmental status.1 Although a knee-jerk response may be that physician knowledge must improve or physician behavior must change, we argue that physician behavior should more appropriately be viewed as appropriate to the context in which it occurs. In other words, office systems, third-party reimbursement schemes, and the interface between medical providers and broader community-based developmental services condition providers to practice a certain style of medicine, and in the case of developmental screening, this style is not supported by the preponderance of evidence.Although different developmental screening tools take varying amounts of time to complete and interpret, they all do take time. If pediatricians spend an average of 18 minutes with each patient,2 even a 2-minute screening tool adds a significant amount of time. Advocates of developmental screening, therefore, are in direct competition with advocates of screening for maternal depression, domestic violence, family substance use, and any other important conditions for which standardized, systematic screening approaches have been shown to be superior to clinical impression.Therefore, we see coordinated quality-improvement efforts as one of many pieces necessary for the translation of evidence-guided care for children with developmental problems into practice. If clinic systems are going to be changed to accommodate developmental screening, they should also be changed in such a way that makes acting on a positive screen possible. Because many developmental services, such as early intervention programs, are outside of the traditional medical system, this means an efficient interface and bidirectional communication with community-based resources is necessary. Last, if any office-based change is to be sustained, third-party reimbursement policies should reflect the importance of these changes.We therefore praise Dr Pavan for her participation in an important learning collaborative. With equally laudable efforts in lobbying for fair reimbursement strategies and pursuing community-specific efforts around the integration of developmental services, we hope that someday children with developmental problems will be better served.