Early in my career as a neonatologist, group B streptococcus (GBS) was one of the infections we feared most. Babies who appeared healthy could become critically ill within hours or days, leaving physicians to tell parents that their newborn was suddenly fighting a life-threatening infection.
Eventually, evidence gave us a way to prevent many of those infections. Researchers found that screening pregnant women for GBS, and treating those who tested positive with antibiotics during labor could dramatically reduce transmission to newborns. Professional societies issued recommendations. But practice did not change overnight.
It took years of clearer guidelines, physician education, and standardized workflows before universal GBS screening became embedded into routine prenatal care. I spent much of my career developing and implementing evidence-based guidelines across neonatal and population health programs. That experience taught me something healthcare needs to pay more attention to today: A guideline can tell clinicians what good care looks like, but on its own it cannot make that care happen.
That matters as medicine advances faster than our healthcare system’s ability to incorporate those advances into everyday practice.
THE GAP BETWEEN EVIDENCE AND ACTION
Clinical research asks whether a diagnostic or treatment works. Implementation science asks what happens next: If we know something works, what will it take to make sure patients receive it?
GBS screening offers useful lessons. Early recommendations gave clinicians multiple approaches. Over time, guidance became more decisive and universal screening became the standard. But guidelines were only part of what changed.
Healthcare built a reliable system around them. The recommendation became a workflow. There was clear timing set for screening, a standard test, a straightforward result, and an established intervention. Responsibility for each step needed to be clear until the process became routine.
That is the lesson we should carry forward as we consider some of the most promising advances in medicine today.
GENOMIC MEDICINE IS AN IMPLEMENTATION CHALLENGE
While innovation is here, there are implementation challenges. Genomic sequencing is a good example. Evidence supporting exome and genome sequencing has expanded significantly, particularly for children with suspected genetic conditions. Clinical guidelines increasingly recognize its role in helping patients reach a diagnosis earlier.
Guidelines, though, are a leading indicator of change, not the endpoint.
Putting genomic testing into practice is more complex. For a clinician, the diagnostic journey starts with connecting subtle clinical clues to a potential hereditary etiology. What follows is a demanding progression of matching symptoms to the right test, guiding the family through pre-test choices, and deciphering complex findings into a tailored care plan.
Once a diagnosis is confirmed, it reshapes the trajectory on various levels by modifying treatments. It establishes surveillance protocols and potentially identifies at-risk relatives. Every transition, though, creates another opportunity for adoption to break down.
Meanwhile, operational friction looms over the entire path. Insurance roadblocks and limited access to specialized genetics colleagues, along with cumbersome documentation routines often overshadow the clinical merits. Even the strongest guidelines will fail to shift behavior when the everyday reality makes acting on them nearly impossible.
CLOSE THE GAP BETWEEN GUIDELINES AND PRACTICE
If we want medical innovation to reach patients, we need to design for implementation from the beginning. Can the right patient be identified easily? Does the clinician know when testing is appropriate? Is ordering simple? Is coverage predictable? Can the result be understood and acted upon? Is it clear who owns the next step?
And critically, can the process work not only at leading academic medical centers with deep specialty resources, but in the places where most patients actually receive care?
Today, we have tools that can accelerate this transition. Clinical decision support can help identify patients who may benefit from testing. Digital ordering can reduce administrative burden. AI can help clinicians navigate complex information. Telehealth can extend access to genetics expertise.
But the goal should be to remove the friction between what medicine knows and what clinicians can realistically do.
WE BUILD THE STANDARD OF CARE
Healthcare often talks about achieving standard of care as though it happens when enough evidence accumulates or a professional society updates its recommendations. My experience has taught me to think about it differently.
A true standard of care exists when the right care happens reliably across physicians, health systems, geographies, and patient populations without requiring extraordinary effort from the clinician or patient. Guidelines are essential, but they have to be translated into coverage and workflows. They need to be included in education and clinician confidence, as well as provide access.
The history of GBS screening shows us that evidence can transform medicine, but only when we build the systems that allow clinicians to act on it. Genomic medicine is a more complex implementation challenge, but the lesson is the same.
We have spent decades getting better at discovering what works in medicine. Now we need to become just as good at making it work in the real world.
Linda Genen, MD, MPH, is chief medical officer at GeneDx.
Â
