McKinsey pro-bono effort helps shed light on “hidden” cause of child blindness

Many people assume that blindness is an issue with the eye—clinicians and special educators included.

But many children with blindness have perfectly healthy eyes. Instead, their blindness is due to cortical/cerebral visual impairment (CVI), a brain-based visual impairment caused by damage to the visual processing areas of the brain.

Because the visual impairment comes from the brain, it most often presents in children who have another neurological diagnosis. CVI can present in many ways, such as difficulty recognizing materials, processing motion, or safely navigating through space. These factors can make CVI difficult to diagnose, leaving kids without the specialized care they need.

“This has really been a hidden epidemic,” says Ilse Willems, senior director of the CVI Center at Perkins School for the Blind, an international NGO in Boston for children with CVI that also advocates for CVI awareness, resources, and support. “We’ve never had the data to show just how prevalent CVI really is.”

Ilse Willems and Dave Power of Perkins School for the Blind
Ilse Willems and Dave Power of Perkins School for the Blind
Ilse Willems and Dave Power of Perkins School for the Blind

That’s starting to change. Perkins, undertook a sweeping research study to understand the extent to which CVI is underdiagnosed in the US, and McKinsey joined to provide critical data analyses. The study identified about 180,000 cases of CVI in the US, even though only 24,000 were officially diagnosed—and this is still likely an undercount.

The work confirmed what those at Perkins had long suspected.

“We received authoritative verification of what is, to some, a mystery affliction,” says Dave Power, president and CEO of Perkins. “The study told us there's more incidence than we might've imagined.”

A methodical approach

Collecting data to analyze the prevalence of CVI would not be easy. Sarah Pickard, a doctor who has experience caring for CVI patients and is now an expert in the firm’s Life Science practice, had a clear grasp of the challenge ahead—for instance, there is not a single unique code to identify CVI in the data.

“We’re dealing with the lack of a unique medical insurance diagnosis code, a lack of CVI training for practitioners, and a lack of sufficient resources to address the amount of need,” says Sarah.

McKinsey's Sarah Pickard
McKinsey's Sarah Pickard
McKinsey's Sarah Pickard

Sarah and the McKinsey team started with donated data from Komodo Health, a healthcare patient data company, and then interviewed clinical experts to understand how to identify CVI patients within this real-world, anonymized data. McKinsey data scientists then used this information to build an algorithm, giving the team a machine-learning engine to find children with CVI in the patient data.

The algorithm uses several characteristics from the data of patients with a diagnosis of CVI to assess against patients without a CVI diagnosis. Using these factors, it’s able to determine the likelihood of CVI being present in patients without a diagnosis.

“What we found is that for every one patient that was diagnosed there were about four more that weren't diagnosed,” says Sarah. “And we believe this may be a conservative estimate.”

Putting the data to work

The McKinsey team mapped out the data into a visualization that they then overlayed with county-level data of a variety of social risk factors, including access to physicians, percent uninsured, poverty rate, and access to transportation.

We really generated what I would say is new information for the field that likely would not have been able to be generated without these resources.

Sarah Pickard, McKinsey expert

This patient data combined with population health information is helping guide Perkins’ next steps in advocating for CVI patients. Now, Perkins can target its work for the best impact, and back up its initiatives with the McKinsey data.

Their work includes creating training materials for medical providers, teachers of students with visual impairments and others who work with them, such as physical therapists and speech therapists; conducting training on diagnosing CVI, including ways to test in a clinical setting, especially for nonverbal patients; and advocating for the creation of a medical code for CVI. These tools will be especially useful in areas where data shows CVI prevalence is high, but resources are scant.

A girl flies a rainbow kite; A toddler looks up smiling
A girl flies a rainbow kite; A toddler looks up smiling.
Photos courtesy of Perkins School for the Blind.
A girl flies a rainbow kite; A toddler looks up smiling

Providers for and parents of children with CVI from around the world look to Perkins for support and information and attend their annual conference in Boston. Because CVI is a neurological disorder, with the right therapies, children have the potential to really improve functional vision.

“When a child makes progress, it’s like a miracle, but you can't pull off those miracles without a playbook, and that’s what we’re providing,” says Dave. “And now we have the data we need to do it.”

Never miss a story

Stay updated about McKinsey news as it happens