The Invisible Crisis: Why Long COVID May Be More Prevalent Than We Think
For years, the medical community has struggled to pin down the true scope of long COVID. While official registries rely on specific diagnostic codes, a groundbreaking study from Mass General Brigham (MGB) suggests that we may be missing a massive portion of the population suffering in silence.
By leveraging artificial intelligence to analyze electronic medical records across 58 hospitals, researchers found that 1 in 6 COVID-19 patients may be dealing with chronic, post-acute conditions that go undiagnosed as “long COVID.”
Beyond the Diagnostic Code
The core issue lies in how doctors document patient encounters. According to Hossein Estiri, director of the Clinical Augmented Intelligence Group at MGH, the federally approved long COVID code is utilized in fewer than 7 percent of potential cases. When a patient presents with persistent fatigue, chronic pain, or sudden-onset heart issues, physicians often treat the symptom rather than labeling the root cause.
The AI algorithm used in the study identified a pattern of “hidden” conditions—including neurological disorders and prediabetes—that lacked any other clinical explanation. This suggests that the burden of long COVID is not just a legacy of the early pandemic, but a rising trend that continued to climb through mid-2025.
The Debate Over Prevalence
Not everyone agrees on the scale of the issue. Experts like Dr. Eric Topol of Scripps Research have urged caution, noting that some AI-identified cases may involve mild symptoms that resolve relatively quickly. The debate centers on how we define “recovery” versus “chronic illness.”
However, proponents of the AI-driven approach argue that our current surveillance systems are far too conservative. Dr. Shawn Murphy, a specialist in biomedical informatics, suggests that recognizing the true breadth of these cases is the only way to compel insurance providers to expand coverage for necessary treatments.
What Which means for Future Healthcare
As we look toward the future of public health, the integration of AI in clinical diagnostics will likely become the standard for identifying complex, multi-system disorders. If the federal government acknowledges these broader prevalence numbers, we could see significant shifts in:
- Insurance Coverage: Expanded reimbursement for long-term management of post-viral syndromes.
- Clinical Guidelines: A move toward symptom-based treatment protocols that acknowledge post-acute COVID triggers.
- Public Awareness: A reduction in the “medical gaslighting” often reported by patients whose symptoms don’t fit traditional diagnostic boxes.
Did You Know?
The CDC defines long COVID as a chronic condition present for at least three months following an infection. Because the symptoms vary so widely—from brain fog to cardiovascular issues—it remains one of the most complex diagnostic challenges in modern medicine.
Frequently Asked Questions (FAQ)
Q: Is long COVID still increasing?
A: Research indicates that cases continued to rise through mid-2025 across multiple U.S. Regions, suggesting that the condition remains a persistent public health challenge.
Q: Why don’t doctors use the official long COVID diagnosis code more often?
A: Because the disorder has no single “cure” and symptoms are incredibly diverse, many clinicians focus on treating specific symptoms (like pain or fatigue) rather than applying a blanket diagnostic label.
Q: Can AI really detect a medical condition better than a doctor?
A: AI doesn’t replace the doctor; it acts as a screening tool that can analyze millions of data points to find correlations—such as sudden-onset prediabetes following an infection—that a human provider might not link to a past virus during a routine visit.
Have you or a loved one navigated the complexities of post-viral recovery? Share your experiences in the comments below, or subscribe to our health newsletter to stay updated on the latest breakthroughs in chronic condition research.
Related reading
