Purpose or Case Report: To evaluate the sociodemographic and exam-related characteristics associated with child life specialist intervention (CLSI) success or failure in acquiring non-sedated pediatric MRI
Methods & Materials: In this retrospective case-control study, the imaging database of a large academic pediatric center was systematically assessed to identify all pediatric patients (0-18 years) who underwent a clinically indicated, non-sedated MRI with CLSI from 8/2020-8/2021. Failed CSLI cases, considered if the resultant images were non-diagnostic due to degrading motion artifacts or if the exam was aborted, were matched in a 1:4 proportion to eligible controls (successful intervention). Collected data included patient sociodemographic (sex, race, language) and clinical characteristics (previous history of failed CLSI, recent MRI requiring anesthesia/sedation, special needs/neurodevelopmental disorder [SN/ND], etc.), as well as exam-related factors (indication, type of exam, scan duration, contrast requirement, etc.). Qualitative variables were reported as frequencies and percentages, and quantitative as median and interquartile ranges (IQR). Multivariate regression was modeled to study the association between a failed intervention and candidate variables. A Classification and Regression Tree (CART) was fitted to understand and graphically depict the relationship between variables and a failed CLSI. ROC curves were obtained to find the best combination of outcome predictors (using boolean operators). Analyses were performed in SPSS v26 (Armonk, NY)
Results: 270 patients (54 cases; 216 controls) were included (median age =10 [IQR 8-12]; 57.4% male). On multivariate analysis, older patients (OR: 0.826; CI, 0.732-0.933), studies performed for an oncologic indication (OR: 0.137, CI, 0.036-0.515), and those done in 3 T scanners (OR: 0.840, CI: 0.580-0.952), had lower odds of CLSI failure. On the other hand, patients with SN/ND (OR:3.89, CI: 1.74-8.68) had significantly higher odds of intervention failure. The CART showed that the subset of patients <12 years and with SN/ND had the highest probability of exam failure. This Boolean combination of predictors (age<12 years AND SN/ND) showed an AUC of 0.776, with a specificity of 76% and a sensitivity of 64%
Conclusions: Several patient- and study-related variables may predict CLSI failure. Our results suggest that patients <12 years and those with SN/ND have higher odds of exam failure. This may serve in the development of future scales for ICLS failure prediction