Jonathan D. Klein, William L. Cull, Karen G. O'Connor, Sandra G. Hassink. American Academy of Pediatrics, Elk Grove Village, IL; A.I. DuPont Hospital for Children, Wilmington, DE.
Presented at the April 2011 Pediatric Academic Societies Annual Meeting.
Background: Measurement of BMI %ile in children is recommended by the AAP and by CDC for identification of overweight and obese children. In recent years, AAP and others have emphasized strategies to address the obesity epidemic, including clinical screening using BMI %ile.
Objective: We evaluated pediatricians' self-reported use of BMI %ile screening in 2006 and again in 2010 to assess whether pediatricians were adopting BMI %ile screening recommendations.
Design/Methods: National, random sample, mailed AAP Periodic Surveys (PS) #65, 2006 (N=1,622; response=62%) and #76, 2010 (N=1,628; response=55%) were compared. Surveys asked identical questions concerning the computation or plotting of BMI. Analysis was limited to post-GME pediatricians spending ≥
50% time in general pediatrics (n=656 and n=552, respectively). Multivariate logistic regression was used to compare BMI screening rates across survey years while controlling for other pediatrician characteristics (age, gender, practice type, practice location, and percentage of patients on public insurance). BMI screening was
defined as pediatricians reporting that they computed BMI or plotted BMI at most or all well child visits for children over age 2 years.
Results: Self-reported use of BMI %ile screening increased from 59% in 2006 to 88% in 2010. This increase (aOR = 6.1, 95% CI = 4.2-8.7) was significant within the multivariate model. Several other factors were also significant predictors of BMI screening: female pediatricians were more likely than male pediatricians (aOR = 1.6, 95% CI = 1.1-2.2), and pediatricians working in teaching hospitals or medical schools were more likely than solo or 2-physician practitioners (aOR = 2.6, 95% CI = 1.4-4.7) to screen. Pediatricians with a higher percentage of public insurance patients were less likely than other pediatricians to screen (aOR = .88, 95% CI = .46-.97). Among the pediatricians who provided screening in 2010, 51% used an electronic medical record (EMR) to compute BMI.
Conclusions: Nearly, 9 of every 10 general pediatricians now report screening for obesity using BMI %ile most of the time or always. This large increase from 2006 indicates that awareness of childhood obesity and of the benefits of BMI screening has spread to many primary care pediatricians. Increased use of EMRs and
additional outreach, especially among pediatricians serving publicly insured populations, will be needed to achieve universal BMI %ile screening.