With Tamara S. Hannon, MD, MS, and Elena Christofides, MD
Given the rising prevalence of diabetes in the pediatric population, the value of a remote screening method to foster an earlier, more consistent diagnosis, was pursued with the aim of improving clinical outcomes.
Tamara S. Hannon, MD, MS, associate professor of pediatrics at Indiana University School of Medicine in Indianapolis, Indiana, and her colleagues, sought to employ the Child Health Improvement Through Computer Automation (CHICA), a computerized clinical decision support system, to decrease screening barriers and improve rates of diagnosis and follow-up for prediabetes and type 2 diabetes (T2D) in children;1 the study was published in JAMA Pediatrics.
The findings indicated that the proportion of youths meeting the outcomes criteria for T2D—body mass index (BMI) and at least 2 other risk factors— was an astonishing 41.3%.1
Based on the authors’ literature review, they had expected greater than 20% of their patients 10 years or older would have a BMI at or above the 85th percentile and at least 2 risk factors for T2D.2
While the CHICA analysis did not increase the proportion of youths identified with diabetes risk factors, it more than quadrupled the rate of screenings and demonstrated greater compliance in follow-up visits.
“Computer decision support is going to be utilized for screening in the future,” Dr. Hannon, the study’s primary author, told EndocrineWeb when discussing the clinical implications of this study. “Healthcare providers can be pointed quickly to conditions that individuals may be at risk for, helping clinicians to be more effective in implementing a care plan."
Similar research results3 suggested that web-based and computer decision support systems may be the evolution in medical diagnosis and treatment, as the data can be accessed readily by clinicians from remote locations, and reduces potential errors in patient analysis.
However, despite the ease of ordering HbA1c and fasting blood glucose, the authors found a low rate of actual overall screening, even for children meeting BMI criteria and having at least 2 other risk factors for diabetes. The researchers noted 3 common barriers to screening:
“CHICA is perfect for a problem that has evidence-based guidelines that can be partially automated to make the work-flow smoother," said Dr. Hannon. The goal is to facilitate the process so physicians will follow the screening guidelines more closely.
The results were so positive, that the researchers are implementing the use of the CHICA-T2D module in all of the control clinics used in their study.
Looking at the Study Process
This study was performed in 4 primary care practices in the Eskenazi Health Center Primary Care system from January 1, 2013 to December 1, 2016. The patients, 10 – 16 years old, had a median age of 13. The study design was a cluster-randomized clinical trial in which 283 patients and an equal number were followed as controls.
The American Diabetes Association guidelines for T2D in the pediatric setting were applied to the CHICA-T2D and paired with an electronic medical record (EMR) to produce a prescreening form containing 20 questions directed at parents. CHICA-related questions covered family history, race or ethnicity, and maternal gestational diabetes while the EMR module inserted diagnostic information such as codes, laboratory data, prescriptions, and physician orders.
The intervention groups used the CHICA system that incorporated the T2D module while the control groups used a generic CHICA model without the inclusion of diabetes guidelines or EMR data.
From a private practice perspective, Elena Christofides, MD, chief executive officer at Endocrinology Associates, Inc., in Columbus, Ohio, felt that the study questions were not as useful as it appeared, initially.
While acknowledging that CHICA-T2D improved screening, Dr. Christofides felt that spending more money on screening methods would not necessarily help physicians diagnose patients more effectively.
“The control group, which was basically physician judgment, had a better return on investment of time and energy spent on screening these patients because they had a higher diagnosis rate,” said Dr. Christofides. She indicated that she always is looking for a return on investment and cost-effective implementation. From that standpoint, her assessment was that the computer decision support systems present the illusion of being cost-effective but were actually money that she couldn't justify spending for this technology.
Offering a counter perspective, Dr. Hannon’s hope lies in integrating the CHICA module into other health care systems programs to make it more widely available, which ultimately would put more pediatric endocrinologists on a level playing field to help screen and diagnosis many more children with prediabetes or T2D. And, in this way, foster preventive strategies and improved outcomes.
1. Lee JM, Eason A, Nelson C, Kazzi NG, Cowan AE, Tarini BA. Screening practices for identifying type 2 diabetes in adolescents. J Adolesc Health. 2014;54(2):139-143.
2. Drobac S, Brickman W, Smith T, Binns HJ. Evaluation of a type 2 diabetes screening protocol in an urban pediatric clinic. Pediatrics. 2004;114(1):141-148.
3. Novo J, Hermida A, Ortega M, Barreira N, Penedo MG, Lopez JE, Calvo C. Hydra: a web-based system for cardiovascular analysis, diagnosis and treatment. Computer Methods Programs Biomed. 2017 Feb;139:61-81.