A lab result is more than a number. The evidence behind it matters.

Every day, clinicians rely on laboratory results to understand their patients’ health, evaluate treatment response, and decide what to do next.

In most cases, these results provide the information clinicians need. Laboratories follow established processes, and physicians reasonably trust the numbers they receive.

But there is an important distinction between receiving a laboratory result and knowing whether that result provides sufficient evidence for a particular clinical decision.

A laboratory value can be correctly reported and still leave important questions about the conclusion drawn from it.

Our research has focused on understanding this gap and determining how clinical evidence can be made more reliable, transparent, and useful.

A decades-old method meets modern treatment decisions

Consider LDL cholesterol (LDL-C), commonly known as bad cholesterol.

LDL-C plays a central role in cardiovascular prevention. Physicians use it to assess risk, monitor cholesterol-lowering therapy, and determine whether patients have reached their treatment goals.

Yet LDL-C is often not directly measured. Instead, it is calculated using other components of a routine blood test.

One widely used calculation method, the Friedewald equation, was introduced in 1972. It has served clinical medicine for more than five decades.

However, extensive research has identified limitations, particularly when triglycerides are elevated or LDL-C concentrations are low.

Modern equations have improved estimation, and the 2026 ACC/AHA dyslipidemia guideline recommends newer methods over Friedewald in relevant clinical settings.¹

This progress matters because modern cardiovascular care increasingly relies on specific treatment thresholds.

For example, a lipid panel with total cholesterol 150, HDL-C 45, and triglycerides 180 mg/dL gives an LDL-C estimate of 69 mg/dL by Friedewald and about 75 mg/dL by Sampson/NIH. Against a documented goal below 70 mg/dL, the same sample falls on opposite sides of the threshold. This illustrates a classification difference; it does not establish which estimate is more accurate.

Neither the laboratory nor the clinician has necessarily made a mistake.

The issue is that the interpretation of a reported value can depend on how that value was obtained.

And that difference may not be obvious when a clinician reviews the result.

The evidence gap clinicians rarely see

Clinical workflows are built around laboratory results being available when physicians need them.

A clinician sees the reported value, considers the patient’s condition, and makes a decision.

Our customer-discovery interviews highlighted an important reality: clinicians generally trust reported laboratory values and do not routinely investigate how every result was generated.

That is understandable. Physicians must interpret large volumes of information, and routinely reconstructing the measurement history behind every laboratory result would be impractical.

But sometimes the context behind a result matters.

Was the value measured or calculated? Were there limitations associated with the method used? Does the available information consistently support the conclusion? Can the result be meaningfully compared with previous measurements?

These questions become particularly important when treatment decisions depend on specific thresholds or changes over time.

The challenge is not that clinicians should distrust laboratory results. It is that important limitations should be visible when they matter, without creating additional work.

What our research revealed

To better understand this problem, our research team investigated how established LDL-C estimation methods behave when used to classify patients against clinical thresholds.

In a peer-reviewed study published in the Journal of Clinical Lipidology, we examined 11,182 selected lipid panels from the NIH All of Us Research Program.²

We found that established equations agreed on which side of the 70 mg/dL threshold a result fell in 85.7% of cases.

When the equations agreed, their classification matched the available routine direct-designated LDL-C comparator in 94.6% of cases.

When the equations disagreed, majority-vote concordance with that comparator fell to 60.6%.

These findings revealed an important pattern: most results produced consistent classifications, while a smaller group showed substantially greater disagreement.

Nevertheless, it identified a meaningful research opportunity: can information already present in routine laboratory records help identify results that deserve closer examination?

That question became central to GOVERN’s development.

A second problem: The number may be preserved, but its origin may not be

Another part of our research examined something more fundamental.

When a laboratory result appears in an electronic health record, how reliably can we establish how it was produced?

In a separate peer-reviewed study published in The Journal of Applied Laboratory Medicine, we investigated records designated as direct LDL-C measurements.³

In the national cohort we studied, we found that at least 7.6% of these records were numerically compatible with Friedewald-calculated values under a stringent comparison criterion.³

This raised concerns about whether some results described as direct measurements might actually represent calculated values.

Importantly, matching a calculation does not prove that an individual result was substituted. Establishing its origin requires supporting laboratory records.

The broader lesson extends beyond cholesterol.

As clinical information moves between laboratories, electronic health records, and other systems, a numerical value may remain intact while important information about its origin or measurement method becomes difficult to establish.

Preserving a number is not always the same as preserving the evidence behind it.

These findings reinforce the importance of measurement provenance and traceable clinical information, particularly as healthcare systems increasingly rely on automation and artificial intelligence.

Why GOVERN exists

GOVERN addresses the gap between laboratory reporting and clinical decision-making.

Its purpose is not to replace laboratories, question every reported value, or make treatment decisions for clinicians.

Instead, GOVERN works quietly within clinical workflows.

It gathers and connects relevant clinical evidence, preserves links to supporting records, and evaluates whether the available information supports the specific question being considered.

When important information is missing or conflicting, GOVERN presents a concise, source-linked explanation for clinician review.

For routine cases, GOVERN stays out of the clinician’s way.

The original laboratory result remains unchanged, and the clinician retains responsibility for care.

The objective is simple: make important evidence limitations easier to recognize without requiring clinicians to investigate every result themselves.

See how GOVERN reviews the evidence.

Evidence that remains useful beyond the laboratory result

The need for reliable clinical evidence does not end when a physician reviews a laboratory report.

Consider a patient receiving cholesterol-lowering treatment.

Their care team may need to compare baseline and follow-up measurements to determine how the patient responded.

Later, the patient may require a medication that needs prior authorization.

Now the team must assemble laboratory history, previous treatment documentation, clinical justification, and other information required by the insurer.

These records may exist across different clinical systems, and their meaning can depend on how the underlying measurements were obtained.

GOVERN keeps that evidence connected. Its medication-access evidence packet brings laboratory results and treatment records together, links each item to its source, and organizes the available evidence against the plan’s stated requirements.

The packet shows what is documented and what is missing, so the care team can review the supporting information before submitting an authorization request. GOVERN does not make the clinical or coverage decision, and the packet does not guarantee approval.

Research first. Evidence in everything we build.

GOVERN is built on peer-reviewed research into the problems it addresses. Its in-chart evidence note and medication-access packet are built capabilities. The studies discussed here establish the problems they address; they do not demonstrate the software’s clinical performance or impact on patients and workflow. Those claims require separate evaluation.

We apply the same standard to each capability: define its intended use, evaluate its performance, and state its limitations alongside the results.

Our commitment is to let evidence guide what we build and claim.

Making evidence count

The future of clinical care will increasingly depend on reliable information.

More sophisticated treatments, more precise clinical thresholds, and growing documentation requirements make it important to understand not just what a laboratory result reports, but what its supporting evidence can establish.

For patients, that could mean fewer important questions being overlooked.

For clinicians, it could mean clearer evidence without additional interruptions.

For healthcare systems and medication-access teams, it could mean less effort reconstructing information that already exists.

GOVERN is built around that opportunity.

The laboratory reports the result. The clinician makes the decision. GOVERN helps make the evidence in between visible, traceable, and useful.