When the Answer Isn't Obvious: How We Approach Persistent Symptoms

William Love, MD, FMCP-M

Aug 19

6 min read

Updated: Oct 1

Taking symptoms seriously without manufacturing explanations

William Love, MD, FMCP-M | Board-Certified Internal Medicine | Love Functional Medicine

A patient has been evaluated. The evaluation was appropriate. Serious disease has been reasonably excluded. The results were largely normal. And the symptoms have not gone away. What happens next divides along a predictable line, and both of the usual responses are failures of reasoning.

There is a particular clinical situation that medicine can struggle to handle well. It is not rare.

One response treats the normal results as the end of the inquiry, which leaves the patient with an unresolved experience and no framework for it. The other treats the persistence of symptoms as proof that something is hiding, and begins searching for it — often with tests chosen for availability rather than for what they could establish.

Both fail, and they fail in opposite directions. What follows is how this practice tries to occupy the space between them.

Listening is not agreeing

The practice's stated commitment is that careful attention comes first. That is a claim about method, not about warmth.

Attention is diagnostic. The sequence in which symptoms appeared, what reliably provokes them, what the patient has already noticed about their own pattern — these are data, and they are frequently the most informative data available. These patterns can be difficult to capture when the clinical question is narrow or time is limited. A longer synthesis visit creates room to reconstruct them deliberately.

But attending seriously to what a patient describes is not the same as accepting every explanation that has been proposed for it. Those are different acts, and conflating them is how patients end up with confident answers that later turn out to be wrong. The commitment is to take the account seriously, not to ratify it.

Taking symptoms seriously is not accepting every proposed explanation

Patients arriving with persistent symptoms have usually encountered explanations along the way — from clinicians, from reading, from other patients, from people selling something. Some of those explanations are sound. Some are plausible but unestablished. Some are not supported at all.

The obligation is to evaluate each on its merits, which means being willing to say that a proposed explanation does not fit — and to say it without implying the symptoms are imagined. Those two things are frequently heard as the same message. They are not, and separating them clearly is part of the work.

Normal results are not nothing

A normal result is often treated as an absence of information. It is not. It is evidence, and it does real work.

It lowers the probability of the conditions the test measures well. It constrains what remains plausible. It can establish a baseline that makes a future change interpretable. What it does not do is measure everything, and a normal result for a test that does not assess the relevant process tells you very little about that process.

So the right question is never whether the tests were normal. It is what those particular tests measured, how well they measure it, and what that leaves open.

Unexplained is not hidden, and possible is not probable

This is where reasoning most often goes wrong, and it goes wrong sympathetically.

If symptoms persist without explanation, it feels natural to conclude that an explanation exists and has been missed. Sometimes that is true. Earlier evaluation may not have reached the right question, and finding it is precisely what a synthesis visit is for.

But the absence of an explanation is not evidence for any particular explanation. It does not preferentially support one candidate over others. A condition being possible — consistent with the symptoms, not excluded by testing — is a much weaker claim than a condition being probable, and the distance between those two words is where a great deal of harm is done. A possible diagnosis can become a working assumption, then a label, then a treatment plan, without ever having crossed the threshold that would justify any of it.

Holding possible and probable apart is unglamorous and it is most of the discipline.

More testing is not more information

More testing does not necessarily produce more useful information. As testing expands without a specific clinical question, the probability of incidental and false-positive findings rises, while each abnormal result creates another decision that has to be interpreted in context.

Most diagnostic testing carries some possibility of false-positive, incidental, or clinically insignificant findings. Order enough tests in a person without a specific pretest question and abnormal results will appear by arithmetic alone. Each one then demands explanation, often generating further testing, sometimes generating a diagnosis that would not have been made had the test not been ordered.

This is why tests are chosen for what a result could change — the assessment, the management, the monitoring, the prognosis, or the next diagnostic step — rather than for coverage. A broad panel ordered without a question does not produce a broad answer.

Uncertainty is not dismissal

Patients have often heard uncertainty delivered as a verdict: we do not know, so there is nothing to pursue. Understandably, uncertainty then starts to sound like disbelief.

Stated properly, it is the opposite. Useful uncertainty is specific. It names what is known, what is suspected and on what grounds, what remains genuinely open, and what evidence would change the assessment. That last element is what separates honest uncertainty from a shrug — it converts an open question into something with a next step and a way of being resolved.

A physician who can say "I do not know, and here is what would tell us" is being more useful than one who supplies a confident answer the evidence does not support.

A diagnosis is a model, not an identity

A diagnostic label is a compression. It groups findings well enough to guide decisions, and its value lies entirely in how well it does that.

Trouble begins when a label stops functioning as a working model and starts functioning as a description of the person. Findings that fit get emphasized. Findings that do not get set aside. The label stops being revisable at exactly the moment it most needs revision.

The commitment here is that the model serves the patient. When new information does not fit, the model changes. The willingness to abandon a diagnosis is part of what makes it worth having proposed one.

What this looks like in practice

These are not abstractions. Each one corresponds to something concrete.

It looks like reading the normal results as carefully as the abnormal ones. It looks like a written assessment that separates what is known from what is suspected. It looks like declining a test and explaining why. It looks like telling a patient that a diagnosis they were offered elsewhere does not fit their findings — and telling them their symptoms are nonetheless real and worth pursuing. It looks like saying that a question belongs to their cardiologist rather than to this practice.

And sometimes it looks like concluding that the most useful thing is not more medicine. That conclusion is a legitimate result of careful evaluation, not a failure to find one.

Questions readers commonly ask

Does this mean most persistent symptoms have no explanation?

No. Many do, and some are identified by asking a question earlier evaluation did not reach. The position is narrower: the fact that symptoms persist is not by itself evidence for any specific explanation, and the explanation has to be established rather than assumed.

If my tests were normal, is there any point in another evaluation?

Sometimes. It depends on what the previous tests measured, what question they were answering, and whether the pattern has ever been assessed as a whole rather than in parts.

Will you tell me you do not know?

When that is the accurate answer, yes — along with what is suspected, what would change the assessment, and what is worth doing next.

About Love Functional Medicine

Love Functional Medicine is a physician-led practice in Mechanicsburg, Pennsylvania, focused on complex symptom patterns that persist despite normal or incomplete evaluation. Adults physically located in Pennsylvania may request a consultation to learn whether this approach fits the question that remains. This is a consultative practice; primary care and specialty relationships remain essential and are not replaced.

This practice does not provide urgent or emergency care. For urgent concerns, contact your primary care clinician; in an emergency, call 911 or go to the nearest emergency department.

Consultations are scheduled through Healthie, the practice's secure booking system.

William Love, MD, FMCP-M — Love Functional Medicine, Mechanicsburg, Pennsylvania

When the answer isn't obvious, the next step may be the 90-minute whole-case consultation.