In this study, four (8.5%) of the children with negative initial skeletal surveys had forensically important findings on the follow-up films and 1 patient had a fracture requiring clinical management. In all of these cases, the information gained by the follow-up skeletal survey strengthened the diagnosis of non-accidental trauma which supported the subsequent safety plans for these children. Our findings and other research supports and affirms the AAP recommendations that all children < 24 months of age receive follow-up imaging when abuse is strongly suspected [2
]. In addition, the results of this study suggest that in cases when abuse is strongly suspected, follow-up imaging should be pursued regardless of the results of the initial evaluation.
The ALARA principle states that every reasonable effort should be made to maintain exposures to ionizing radiation as far below the dose limits as practical. This is an important consideration for follow-up imaging. In this study, 75% of the patients who had findings on follow-up skeletal survey had rib fractures. The frequent identification of occult rib fractures on follow-up imaging has been noted in other studies. Zimmerman et al noted that 66% of fractures detected on follow-up skeletal surveys were rib fractures [6
]. Kleinman et al noted that 94% of additional injuries noted on follow-up skeletal survey were rib fractures or classic metaphyseal lesions [4
]. Similarly, Harlan and colleagues found that 49.7% of fractures found on follow-up skeletal survey were rib fractures [7
]. The data from our study and previous investigations demonstrate that chest radiographs have a high yield for detecting occult injuries on follow-up skeletal surveys. The aforementioned study by Harlan determined that a limited follow-up skeletal survey could be performed without missing clinically significant new fractures [7
]. When limiting radiation is a priority, a chest radiograph in lieu of a full skeletal survey may be useful. Future studies are needed to further explore the optimal number of images in follow-up skeletal surveys.
A larger sample size may have produced a narrower confidence interval and increased the reliability of our results. Future studies using multiple centers are considered to address this further.
The literature states that a repeat skeletal survey should be completed approximately 2 weeks after the initial skeletal survey [3
], but variations in the timing of follow-up are to be expected. The follow-up skeletal survey was completed between 9-13 days in 14 (30%) of our patients and after 40 days in 4 (8.5%) of our patients. Two of these 18 children had positive findings on the follow-up evaluation. It is possible that some fractures may not have been detected with early follow-up or healed and subsequently missed with delayed follow-up. Therefore our results may underestimate the frequency of findings on follow-up skeletal surveys in this population.
The skeletal surveys were read by the attending pediatric radiology staff at our institution and therefore slight variability between radiologist interpretations may exist. While utilizing a single radiologist for all skeletal surveys would limit this potential variability, our methods reflect clinical practice and allow the results to be generalized to a larger population.