The current system by which we typically diagnose developmental and behavioral disorders in children is dysfunctional, as specialists are vastly outnumbered by at risk-children in need of appointments. For delays like autism, the earlier the diagnosis, the earlier the intervention can start and the more effective it becomes. We are in the early stages of using technology to advance a variety of areas in healthcare, such as using AI to look at pictures of your skin and diagnose skin cancer and identify disease-causing gene mutations. There is a natural place for technology in healthcare, and child development is an area in which it can make a big difference.
There are several problems present in the area of developmental diagnoses, most of which stem from an overall lack of developmental specialists. There are only about 13 pediatric specialists for every 100,000 patients, with this number only projected to worsen. This shortage is in part due to the economic disadvantages for doctors in the field – the training time generally leads to an extra two to three years of schooling, and the average Medicaid reimbursement is about 30 percent less than Medicare. A 2011 Pediatrics study even noted that going by solely financial reasons, the only pediatric subspecialties worth pursuing are critical care, neonatology and cardiology, which doesn’t leave much room for children at high risk of a developmental disorder.
With many physicians understandably choosing not to foot the financial burden of this specific field, we have a pretty major crisis on our hands when it comes to children not receiving therapeutic interventions within the most effective timeframe. Due to the way physicians are trained, many primary care physicians, including pediatricians, are often uncomfortable with delivering developmental diagnoses. The time between a parent initially expressing concern over their child’s behavior and being able to finally meet with a specialist can reach a year or more. All this time spent without a proper diagnosis often leads to missed opportunity for critical interventions. Access to these specialists can be difficult as well, with the average family needing to drive 44 miles just to visit one. For low-income families or those in rural areas, these roadblocks are life-changing.
We can use the aforementioned technology to begin to patch up a system that’s been broken for a while. This particular sector of technology has an incredibly applicable place in behavioral health. We are essentially using the human brain to teach a machine to learn how the human brain works, so its use in neurology is especially pertinent. Utilizing it to deliver diagnoses for disorders like autism showcases a true symbiotic relationship between doctors and data science.
Machine learning technology can essentially operate as a triage service for developmental disorders. For instance, with Cognoa, parents complete a questionnaire exploring their children’s behavioral tendencies. Using an algorithm, expert insights and data from thousands of previous cases, these results are then compiled into a report which identifies the child’s risk level for a developmental delay. If a child is identified as high-risk, this can help push the family higher up in the queue to see a specialist, as referenced in the corresponding study. This is an example of a method bringing out the best of both human and machine intelligence and allowing them to work together.
Again, this is not a self-operating machine nor is the algorithm replacing medical professionals, by any means. But it can serve as the middleman between patient and doctor when resources are tight. It gives both patients and their families a chance to follow their intuition and push for medical intervention where necessary.
As for doctors, we need to take a more active role in the pursuit of AI and machine learning advancements. It has a strong current and potential use in drug discovery, overall research, diagnosis and treatment – the entire healthcare system as a whole. Once we start keeping up with trends and welcoming this technology into our practices, we can start recognizing its true value. We can start empowering patients without sacrificing physician expertise, and we can start using technological advancements to make both ourselves and the medical system stronger altogether.
Sharief Taraman, MD, is the assistant division chief of pediatric neurology and medical informaticist at Children’s Hospital of Orange County. He also serves as the Vice President of Medical at Cognoa, a digital health company revolutionizing how children with developmental delays are diagnosed and supported.
Source: BioScience by Sharief Taraman, MD, Children’s Hospital of Orange County