When Medical Information Leads Us in the Wrong Direction
Why a convincing scientific article doesn't always mean you need another blood test.
Hampton & South Medical Centre | Patient information.
We live in an age when almost anyone can access sophisticated medical information. Patients can search scientific journals, read specialist guidelines and ask artificial intelligence tools to interpret their symptoms and laboratory results.
This is generally a good thing. Patients who understand their health are better equipped to participate in decisions about their care.
But there is a potential problem.
A patient discovers an unusual blood test result, searches for an explanation and finds a scientific article describing a disease associated with that abnormality. Further searching uncovers additional articles supporting the possibility.
Before long, a convincing explanation has emerged, along with a growing list of investigations that appear necessary.
The problem is that the information may be entirely accurate, but its relevance to the individual patient may have been misunderstood.
Finding a possible explanation is not the same as establishing a likely diagnosis.
A simple example: a mildly abnormal liver test
Imagine a healthy adult who undergoes a blood test and discovers that one liver enzyme, alanine aminotransferase (ALT), is slightly above the laboratory reference range.
The patient feels well but searches the internet and discovers that elevated ALT can be associated with autoimmune hepatitis, viral hepatitis and other potentially serious liver diseases.
A scientific article explains that autoimmune hepatitis can sometimes be discovered in people who have no symptoms.
Naturally concerned, the patient requests an extensive battery of autoimmune tests, viral investigations and liver imaging.
The information they have found may be entirely correct.
But does it apply to them?
A mildly elevated ALT is a finding, not a diagnosis. Its significance depends on the degree of elevation, previous results, other liver tests, medications, alcohol consumption, metabolic risk factors and the patient's medical history.
A GP may recommend further investigations immediately, arrange repeat testing or refer the patient to a specialist, depending on the circumstances. Persistent abnormalities require appropriate assessment and should not simply be dismissed.
The patient has correctly discovered that autoimmune hepatitis is one possible explanation. However, they have not established that it is the most relevant explanation in their particular circumstances.
The challenge is determining which investigations are appropriate, rather than simply ordering every test mentioned in an article.
The missing ingredient: clinical judgement
Medical training involves considerably more than memorising diseases and their associated symptoms.
Over years of clinical practice, doctors develop what is sometimes called tacit knowledge: knowledge gained through experience that can be difficult to communicate through a textbook, checklist or website.
This includes recognising which minor laboratory abnormalities are commonly encountered, which patterns warrant concern, and which investigations are likely to influence treatment.
There is a fundamental difference between knowing that a disease can produce a particular result and knowing how likely that disease is to explain the result in an individual patient.
Consider the laboratory reference range.
For many tests, reference intervals are established to encompass approximately 95% of an apparently healthy reference population. This means some healthy individuals will have results outside those intervals.
A result falling just outside the quoted range does not automatically indicate disease.
Blood test results can also vary because of normal biological variation, time of day, recent illness, medication and laboratory measurement.
An experienced clinician considers the magnitude of an abnormality, its pattern over time, associated findings and the patient's overall circumstances.
A seemingly minor abnormality can occasionally be important. Conversely, a result flagged in red on a laboratory report may have little clinical significance.
The distinction requires judgement rather than simply identifying whether a number is above or below a reference limit.
How we can be led in the wrong direction
The process often begins with a reasonable question.
A patient notices an unexpected result and searches for possible causes. A search engine produces a list of conditions, including rare but serious diseases.
The patient then searches for more information about one of those diseases and finds a scientific article describing the association.
Perhaps an AI tool is asked to investigate further. It produces an impressive explanation, complete with references and a list of suggested tests.
Each successive search makes the original hypothesis appear increasingly plausible.
Yet the research may never address the most important question:
How likely is this condition in someone with my particular symptoms, history and pattern of test results?
A scientific article may accurately describe findings in patients who already have a particular disease. It does not necessarily tell us how frequently that disease occurs among otherwise healthy people with one borderline laboratory result.
This is a common source of misunderstanding.
The information is not necessarily wrong. It is being applied in the wrong context.
Searching repeatedly for evidence supporting one explanation can also lead us to overlook other possibilities, including the possibility that the original finding is not clinically significant.
Doctors are not immune to this problem either. Clinical training and experience help reduce the risk, but good practice also requires doctors to reconsider their initial impressions when new evidence emerges.
More tests do not necessarily mean better healthcare
This brings us to another common problem: the increasing popularity of large batteries of blood tests.
Some complementary healthcare providers, including naturopaths, recommend extensive panels covering vitamins, minerals, hormones and other biochemical markers, sometimes without a clear clinical indication.
Patients may also request these investigations themselves after encountering advertisements for comprehensive health checks or personalised wellness testing.
The underlying assumption is understandable: surely testing more things must provide a more complete picture of our health?
Unfortunately, it is not that simple.
Imagine ordering 20 independent tests, each with a reference interval encompassing 95% of healthy people. Even if the person is healthy, there is a substantial chance that at least one result will fall outside its reference range.
In practice, laboratory results are not always independent, and reference intervals differ, but the principle remains: the more tests ordered, the more incidental abnormalities we are likely to encounter.
An unexpected result can lead to further blood tests, imaging and specialist consultations. These may uncover additional incidental findings, each requiring another decision about whether investigation is necessary.
This phenomenon is sometimes called a diagnostic cascade.
The Royal Australian College of General Practitioners has described how excessive testing can generate incidental findings, further investigations and potential patient harm.
Reference: RACGP, We live in testing times – teaching rational test ordering in general practice.
https://www.racgp.org.au/afp/2014/may/we-live-in-testing-times
The potential consequences include unnecessary expense, anxiety, false alarms and occasionally treatment for conditions that would never have caused harm.
None of this means that doctors should avoid investigating genuine abnormalities. It means that investigations need a clear purpose.
Before ordering a test, the important questions are: What condition are we considering? How likely is it? What will we do if the result is positive, negative or borderline?
For further discussion, see our related article on battery testing.
Testosterone: a common example of testing without a clear indication
One increasingly familiar request in general practice is for testosterone testing.
Men may have read that testosterone levels decline with age or encountered advertisements linking low testosterone to fatigue, reduced energy, weight gain and other common experiences.
They may request a testosterone level as part of a routine health check, even when there are no clinical features suggesting androgen deficiency.
However, declining testosterone levels with age do not automatically indicate a pathological hormone deficiency requiring treatment.
A man who has developed normal male secondary sexual characteristics, such as facial hair, and has normal sexual function may have no obvious reason to suspect androgen deficiency. These findings alone do not exclude it, but they contribute to the overall clinical assessment.
Symptoms such as reduced libido, erectile dysfunction, infertility, loss of body hair or unexplained osteoporosis may warrant further assessment, depending on the circumstances.
The Endocrine Society of Australia advises that testosterone testing should follow clinical suspicion of pathological androgen deficiency rather than being used for general population screening.
Reference: Endocrine Society of Australia, Position statement on male hypogonadism: assessment and indications for testosterone therapy.
When testing is appropriate, testosterone levels need to be measured under suitable conditions, generally in the morning, and an unexpectedly low result requires confirmation and interpretation.
A single low result does not necessarily establish a diagnosis.
Likewise, testosterone replacement is appropriate for established pathological hypogonadism, regardless of age, but is not automatically indicated for age-related decline alone.
The aim is to diagnose and treat genuine disease, not to treat every numerical change associated with getting older.
Medicare does not fund every requested test
There is another important consideration: Medicare has specific requirements governing when certain pathology tests qualify for a rebate.
Patients may assume that a blood test is automatically covered because their GP has requested it.
That is not necessarily the case.
A doctor must consider both the clinical appropriateness of an investigation and the relevant Medicare Benefits Schedule (MBS) requirements.
Two examples illustrate the distinction.
Vitamin D
Vitamin D testing is not recommended as a routine screening test for everyone.
Medicare restricts rebates for vitamin D testing to specified clinical circumstances, including suspected osteoporosis or osteomalacia, malabsorption, chronic and severe lack of sun exposure, particular medication use and certain other recognised risk factors.
A patient does not necessarily qualify for a Medicare-funded vitamin D test simply because they would like to know their level.
The complete eligibility requirements are available under Medicare Benefits Schedule item 66833.
https://www9.health.gov.au/mbs/fullDisplay.cfm?q=66833&type=item
Vitamin B12
Medicare requirements for vitamin B12 testing were revised on 1 July 2025.
The standard vitamin B12 test, item 66838, is generally limited to once every 11 months.
However, an additional item, 66842, allows more frequent testing when specified clinical circumstances are present.
These include patients requiring monitoring of B12 treatment, those with diets low in vitamin B12, certain gastrointestinal conditions, relevant medication use and persistent symptoms meeting the item requirements.
These exceptions matter. The rules do not mean that patients with a genuine medical need must simply go without testing.
References:
Medicare Benefits Schedule item 66838:
https://www9.health.gov.au/mbs/fullDisplay.cfm?q=66838&type=item
Medicare Benefits Schedule item 66842:
https://www9.health.gov.au/mbs/fullDisplay.cfm?criteria=&q=66842&qt=item&type=item
The examples illustrate why a GP may decline a request or explain that a particular test is not eligible for a Medicare rebate.
A clinically appropriate investigation can sometimes fall outside Medicare's funding requirements. Conversely, a patient's willingness to pay privately does not automatically make a test medically worthwhile.
Doctors have obligations too
Doctors cannot simply order whatever investigations a patient requests without considering their clinical relevance.
Medicare benefits are intended for clinically relevant services, and medical practitioners are expected to understand and comply with the applicable requirements.
Inappropriate practice can result in professional review and consequences such as repayment of Medicare benefits, counselling, reprimands or restrictions on Medicare billing.
This does not mean a GP should refuse every test that falls outside routine practice.
There will be circumstances where an unusual investigation is justified, particularly when symptoms, examination findings or previous results warrant further assessment.
However, the decision should arise from a reasoned clinical assessment rather than simply from a patient's request or a recommendation generated by an internet search.
It is also worth remembering that different doctors may make different decisions in situations where the evidence is uncertain.
The fact that another practitioner is willing to order a test does not, by itself, establish that the investigation is necessary, just as declining a test does not automatically mean that a patient's concerns have been dismissed.
Where does artificial intelligence fit in?
AI has made medical information more accessible than ever before.
It can explain complicated terminology, summarise scientific articles and help patients prepare questions for their doctor.
However, AI systems can also produce highly persuasive explanations that give insufficient attention to the probability of a diagnosis or the limitations of the available information.
An AI tool asked, "Could this abnormal result be caused by a particular disease?" may correctly answer yes and produce an extensive explanation supporting that possibility.
That is not the same question as, "Given everything we know about this patient, what is the most appropriate next step?"
The latter requires a broader assessment of the clinical picture, including information that may not have been provided to the AI system.
Even a response containing legitimate scientific references can be misleading if it focuses on a possible diagnosis without adequately considering how likely that diagnosis is.
AI can assist clinical reasoning, but it should not be mistaken for a complete clinical assessment.
Should patients stop researching their health?
Absolutely not.
Patients should be encouraged to understand their medical conditions, ask questions and participate in decisions about their treatment.
Sometimes patients identify important information that deserves further investigation. Doctors also need to remain open to new evidence and reconsider their assessments when circumstances change.
However, it is worth approaching medical information with a degree of caution.
Rather than asking only whether a particular disease could explain a symptom or abnormal result, it is often more useful to ask how likely the explanation is, what other possibilities exist, and whether the proposed investigation will meaningfully influence management.
A GP who recommends against an investigation may be recognising that the proposed test is unlikely to provide useful information or could generate more uncertainty than it resolves.
Sometimes the right decision is to investigate immediately. Sometimes it is to monitor the situation, repeat a test or seek specialist advice.
And sometimes the right decision is not to order another test at all.
The take-home message
Good healthcare is not about ordering the greatest number of tests. It is about ordering the right tests, for the right reasons, at the right time.
Medical information helps us understand what is possible.
Clinical judgement helps us determine what is relevant.
The two work best together.
This article provides general health information and is not a substitute for an individual medical assessment. Decisions about investigations should take account of a patient's symptoms, medical history, examination findings and clinical circumstances.
Many Australians assume they won’t qualify for government concessions once they retire — particularly if they own assets or have super.
New medications known as GLP-1 receptor agonists are increasingly being used to assist with weight loss. These include: