Margaret, a 52-year-old teacher in Manchester, first noticed the fatigue two years ago. She would fall asleep by 7 PM, gained roughly 10 kilograms without changing her diet, and felt as though a fog had settled over her thinking. Her GP ran a thyroid panel. The results: TSH 3.9 mIU/L, within the laboratory reference range of 0.5–4.5. “Normal,” the letter said. Margaret was told to monitor symptoms and return in six months. She is one of an estimated 5% of the UK population with hypothyroidism, but only about 30% of those on levothyroxine are adequately treated, according to NHANES data. The gap between symptoms and lab flags is not a mystery of medicine—it is a failure of system inertia.

The 6-Week Wait That Should Be 2 Days

When a patient presents with classic hypothyroid symptoms—fatigue, cold intolerance, weight gain, constipation, dry skin, and cognitive slowing—the standard response is to check thyroid-stimulating hormone (TSH). If TSH exceeds the lab’s upper limit, usually 4.5 mIU/L, the protocol is clear: start or adjust levothyroxine. But what if TSH is 3.9 and the patient feels terrible? The GP cannot act on symptoms alone. Most clinical guidelines require a confirmatory test after 6–12 weeks, and only then, if TSH has risen above the threshold, can a dose change be initiated.

That means a patient who walks in today with a TSH of 3.9 will be told to wait two to three months for a recheck. During that time, symptoms persist. Work productivity suffers. Sleep quality deteriorates. And for women of childbearing age, the stakes are higher: untreated hypothyroidism during pregnancy increases risks of miscarriage and preterm birth. A 2019 audit by the NHS Wessex Clinical Commissioning Group, published in the British Journal of General Practice, found that the average time from initial presentation of hypothyroid symptoms to a dose adjustment was 10.2 weeks across 12 GP surgeries. In private practice, where patients can pay for a same-day appointment and a private lab, the turnaround can be 48 hours.

The disparity is not driven by clinical complexity. Adjusting levothyroxine is straightforward: increase by 25 mcg, recheck TSH in 6–8 weeks, repeat. The bottleneck is the reference range. Until the lab flags a value as abnormal, the GP has no justification to act within the system. Some GPs will use clinical judgment and adjust anyway, but they risk being flagged for “non-guideline” prescribing, which can trigger audits or complaints. Most choose to wait.

The irony is that the 0.5–4.5 mIU/L range was set decades ago, based on small studies of healthy volunteers that excluded people with positive thyroid antibodies. Modern population data show that the optimal TSH for most adults is below 2.5. The National Academy of Clinical Biochemistry and the American Thyroid Association have both recommended lowering the upper limit to 2.5. But labs rarely update their panels. Recalibrating a reference range requires new reagents, new software validation, and a willingness to increase the number of flagged results—which some labs fear will overwhelm clinicians.

Why Lab Reference Ranges Are Stuck in the 1970s

The TSH reference range of 0.5–4.5 mIU/L was established in the 1970s using first-generation radioimmunoassays. Those assays had higher coefficients of variation and included a broader, healthier population. Today’s third-generation immunometric assays are far more precise, with coefficients of variation below 5%. Yet most labs continue to report the same old cutoffs. The Endocrine Society has published guidelines since 2012 recommending a narrower range—typically 0.5–2.5—but adoption has been slow. A 2022 survey by the Association for Clinical Biochemistry and Laboratory Medicine (now part of the Royal College of Pathologists) found that fewer than 20% of UK clinical biochemistry labs had updated their TSH reference ranges in the previous five years.

Why the resistance? Changing a reference range is not trivial. It requires the lab to run a new reference interval study using local population data, which costs roughly £50,000–100,000 per lab. Multiply that by the roughly 200 NHS labs in England, and the total is in the tens of millions. Labs argue that the money is better spent on new equipment or staffing. But the cost of not changing is also high: a 2019 study in the Journal of Clinical Endocrinology & Metabolism estimated that lowering the upper TSH limit to 2.5 would reclassify roughly 15% of the adult population as having elevated TSH, potentially doubling the number of people treated for hypothyroidism. That would increase prescribing costs—but also reduce the downstream costs of untreated disease.

There is also a philosophical debate. Some endocrinologists argue that treating everyone with TSH above 2.5 would medicalize a large number of asymptomatic people. The evidence that levothyroxine improves quality of life in patients with TSH between 2.5 and 4.5 is mixed. A 2020 Cochrane review found that levothyroxine improved symptoms in patients with overt hypothyroidism (TSH >10) but not consistently in those with subclinical disease. Critics say that lowering the threshold would lead to overdiagnosis and unnecessary medication. Supporters counter that the trials excluded patients with positive antibodies—the very group most likely to benefit. In clinical practice, the decision should be individualized, but the current range prevents individualization by defaulting to “normal.”

Meanwhile, labs in Australia and parts of Europe have quietly adopted narrower ranges. The Royal College of Pathologists of Australasia now recommends an upper limit of 4.0–4.5 depending on the assay, but some Australian labs use 3.5. In Germany, the upper limit is commonly 4.0. In the UK, NICE published a draft update in 2022 that considered lowering the threshold for treatment in symptomatic patients with TSH between 4.0 and 10, but the final guidance was delayed amid disagreements. As of mid-2026, the draft remains in limbo.

The Mechanism: How a Static Range Delays Treatment

Imagine a 35-year-old woman with a family history of Hashimoto’s thyroiditis. She presents with fatigue, hair thinning, and a feeling of being “slowed down.” Her TSH is 3.9, her free T4 is 12 pmol/L (within range), and her thyroid peroxidase antibodies are elevated at 200 IU/mL. According to current guidelines, she has subclinical hypothyroidism: a normal free T4 with a mildly elevated TSH. The recommended approach is watchful waiting, with repeat testing in 6–12 months. But the patient’s symptoms are real, and her antibody status predicts a high risk of progression to overt hypothyroidism—about 5% per year.

The GP is caught. The lab report says “normal.” The patient is suffering. The GP can either document the situation and wait, or prescribe off-label. Most choose to wait. Over the next year, the patient’s TSH creeps up to 4.8, then 5.6. Only at that point does the lab flag it, and the GP can finally start levothyroxine. By then, the patient has lost a year of function. That year could have been prevented by a simple dose of 25 mcg of levothyroxine daily, at a cost of roughly £20 per year. The delay is not due to ignorance—it is due to a number printed on a lab report.

The same mechanism plays out in reverse for patients already on levothyroxine. A patient whose TSH is 0.01 (suppressed) may feel anxious, tachycardic, and lose bone density. But if the lab range is 0.5–4.5, the low TSH is flagged, and the dose is reduced. That works. But a patient on 100 mcg whose TSH is 0.6 is considered “in range” even if they have symptoms of over-replacement. The GP may not reduce the dose until TSH drops below 0.5. The range, again, creates a false sense of adequacy.

The problem is compounded by the fact that TSH is a lagging indicator. After a dose change, it takes 6–8 weeks for TSH to reach a new steady state. So the watchful waiting period becomes a double delay: first to get the lab to flag the value, then to wait for the TSH to respond to treatment. The total time from symptom onset to therapeutic dose can easily exceed six months. For a condition that affects metabolism, cognition, and energy, that is an eternity.

Who Pays the Price of Waiting

The human cost of delayed levothyroxine adjustment is hard to quantify but impossible to ignore. Fatigue alone reduces work productivity. A 2021 study from the University of Birmingham estimated that untreated hypothyroidism costs the UK economy roughly £1.2 billion annually in lost productivity. For the individual, the cost is measured in missed promotions, strained relationships, and diminished quality of life. Brain fog—a term patients use to describe difficulty concentrating, memory lapses, and slow thinking—is one of the most distressing symptoms. It is also one of the first to improve with adequate levothyroxine, often within two weeks of dose adjustment.

Pregnant women face particular risk. TSH requirements increase during pregnancy, and inadequate treatment is linked to miscarriage, preterm birth, and lower IQ in offspring. Guidelines recommend maintaining TSH below 2.5 in the first trimester. Yet many women enter pregnancy with TSH between 2.5 and 4.5, deemed “normal” by their lab, and are not dose-adjusted until after a miscarriage or a high TSH is detected at booking. A 2019 study in the BMJ found that women with TSH between 4.0 and 10 had a 40% higher risk of pregnancy loss compared to those with TSH below 2.5. Early adjustment could prevent many of these losses.

Elderly patients also pay a price. Hypothyroidism slows reflexes, reduces muscle strength, and increases fall risk. A 2022 meta-analysis in the Journal of the American Geriatrics Society found that untreated subclinical hypothyroidism was associated with a 25% increase in fall-related fractures. Once a hip fracture occurs, the cascade of immobility, institutionalization, and mortality is steep. Adjusting levothyroxine earlier could prevent some of these fractures, but the static range delays intervention until TSH is frankly elevated.

Emergency departments bear the downstream burden. Patients with undiagnosed or undertreated hypothyroidism often present with non-specific symptoms: weakness, confusion, constipation, or falls. They undergo costly workups—CT scans, blood panels, ECGs—before someone checks thyroid function. A 2023 audit at Guy's and St Thomas' NHS Foundation Trust in London found that roughly 8% of emergency department visits for non-specific complaints could be attributed to thyroid dysfunction, most of which was known but undertreated. The cost of one emergency visit (£150–300) far exceeds the cost of a year of levothyroxine (£20) and a GP appointment (£40). The system is spending more money by waiting than it would by adjusting.

A Simple Fix That Nobody Wants to Fund

The solution is technically straightforward: lower the TSH reference range to 0.5–2.5 mIU/L, as recommended by the Endocrine Society. This would automatically flag patients with TSH between 2.5 and 4.5, allowing GPs to initiate shared decision-making about treatment. The assays already have the precision to support this narrower range. Mass spectrometer TSH assays, now common in reference labs, have coefficients of variation below 3% at the upper limit. The technology is not the barrier.

The barrier is cost. Updating reference ranges across all NHS labs would cost an estimated £15 million in one-time recalibration and validation. That sounds like a lot, but consider: the NHS spends roughly £350 million annually on unnecessary thyroid imaging—ultrasounds and nuclear medicine scans for nodules that are almost always benign. A 2024 study in the British Journal of General Practice found that over 80% of thyroid ultrasounds ordered for elevated TSH showed no clinically significant nodules. Redirecting a fraction of that imaging budget to lab recalibration would pay for the change many times over.

There is also the cost of increased prescribing. If the range is lowered, an additional 5–10% of the adult population might be offered levothyroxine. At £20 per patient per year, that is roughly £100–200 million annually. But much of that cost would be offset by reduced visits for fatigue, reduced emergency admissions, and improved productivity. A health economics analysis published in Thyroid in 2021 estimated that lowering the TSH threshold to 2.5 in symptomatic patients would be cost-effective, with an incremental cost-effectiveness ratio well below the NICE threshold of £20,000 per quality-adjusted life year.

Yet the political will is lacking. NICE’s 2022 draft update on thyroid disease assessment was expected to address the reference range issue, but it was shelved after pushback from the Royal College of Pathologists and some endocrinologists who argued that the evidence was insufficient. The draft remains unpublished. Meanwhile, the American Thyroid Association updated its guidelines in 2023 to recommend treatment for symptomatic patients with TSH between 2.5 and 4.5 and positive antibodies. But UK labs have not followed. The result is a patchwork: some GPs treat based on symptoms and ignore the lab range, while most adhere strictly to the numbers. The variability is a recipe for inequity.

A Counter-Argument: The Case for Caution

Not everyone agrees that lowering the TSH threshold is the right move. Some endocrinologists point to the risk of overtreatment. Levothyroxine, while generally safe, can cause iatrogenic thyrotoxicosis if dosed too high, leading to atrial fibrillation, osteoporosis, and anxiety. A 2021 study in the Lancet estimated that up to 20% of patients on levothyroxine have suppressed TSH (<0.1), indicating over-replacement. Lowering the threshold could increase that number if GPs are not meticulous about follow-up. The argument is that the current range, while imperfect, at least prevents a flood of new prescriptions that may not benefit everyone.

There is also the issue of lab capacity. If the range is lowered, the number of flagged TSH results could double overnight. Labs already struggle with turnaround times; a sudden surge in abnormal results might lead to more calls from GPs, more repeat tests, and longer waits for everyone. Some lab directors argue that the system would be overwhelmed, at least temporarily. A 2023 commentary in the Annals of Clinical Biochemistry warned that a blanket change without adequate resources could backfire, causing more confusion than clarity.

Finally, the evidence for treating subclinical hypothyroidism in asymptomatic patients is weak. The 2020 Cochrane review found no clear benefit of levothyroxine for quality of life, depression, or fatigue in patients with TSH between 4.0 and 10 who had no symptoms. Treating everyone with TSH above 2.5 would inevitably include many who would never progress to overt disease and who would gain no benefit from medication. The risk of medicalization is real: labelling someone with a disease they might never notice can cause anxiety, repeated testing, and unnecessary healthcare use.

These counter-arguments are not trivial. They explain why many labs and guidelines have been hesitant to change. But they also miss the point: the current system does not allow for nuance. The lab range is a blunt tool that labels a symptomatic patient as normal. The fix is not to mandate treatment for everyone above 2.5, but to flag that range so that a conversation can happen. The current range suppresses that conversation. A narrower range would allow GPs to use their clinical judgment, rather than being overruled by a decades-old number.

What GPs Can Do Right Now

While waiting for systemic change, GPs have tools at their disposal. The first is to treat symptoms, not just lab numbers. A patient with classic hypothyroid symptoms, a TSH of 3.9, and positive antibodies is a good candidate for a trial of levothyroxine. Several guidelines, including those from the American Thyroid Association, support a therapeutic trial in such cases. The key is to document the clinical decision-making: note the symptoms, the antibody status, and the shared decision with the patient. This protects against audit criticism.

Second, GPs can use clinical judgment to override the lab flag. The lab range is a statistical construct, not a diagnostic commandment. If a patient on levothyroxine has a TSH of 0.6 but reports palpitations and anxiety, reducing the dose is reasonable even if the TSH is “in range.” The same applies to the patient with TSH 3.9 and fatigue. The GP can start 25 mcg of levothyroxine, recheck TSH in 8 weeks, and adjust. The risk of overtreatment is low at that dose, and the potential benefit is high.

Third, GPs can advocate for local lab range review. Clinical commissioning groups and hospital trusts can request that their lab update reference ranges to align with current evidence. Some labs have already done so. For example, the lab at University College London Hospital uses an upper limit of 4.0 for TSH, and clinicians there report fewer delays. GPs can also refer borderline cases to endocrinology, where specialists are more comfortable treating subclinical disease. However, endocrinology waiting lists are long—often 6–12 months—so referral is not a quick fix.

Fourth, GPs can use point-of-care TSH testing, where available. Some GP practices now have in-house TSH analyzers that give results in 15 minutes. With point-of-care testing, the GP can discuss the result immediately with the patient and, if appropriate, start treatment that day. The cost of the test is roughly £10–15, comparable to a lab test, but it eliminates the 6-week wait for a recheck. A 2024 pilot in four GP surgeries in the West Midlands, reported by the NHS Innovation Accelerator, found that point-of-care TSH reduced the time to dose adjustment from 10 weeks to 2 weeks, with high patient satisfaction.

Finally, GPs can document shared decision-making. The patient should understand that the decision to treat a TSH of 3.9 is based on their symptoms and preferences, not on a strict threshold. Some patients will choose to wait; others will want to try medication. Both choices are valid. The GP’s role is to provide the evidence and let the patient decide. The current system denies that choice by labeling the TSH as normal, which discourages discussion.

The Cost of Inaction: A System That Prefers Waiting

The longer the reference range remains unchanged, the more patients will suffer avoidable symptoms. Hypothyroidism affects roughly 5% of the UK population, or about 3.3 million people. Of those on levothyroxine, NHANES data suggest only 30% have a TSH within the optimal range of 0.5–2.5. That means over 2 million people in the UK are likely undertreated or overtreated. The annual NHS spend on levothyroxine is around £50 million, a relatively small sum. But the cost of dose adjustment—a GP visit and a lab test—is roughly £40 per patient per year. If 2 million patients need adjustment, that is £80 million annually. However, the savings from reduced emergency visits, fewer fractures, and improved productivity far outweigh that cost. A 2022 analysis by the King’s Fund estimated that optimizing thyroid treatment could save the NHS £200 million annually.

Yet there is no financial incentive for the system to adjust faster. GPs are paid by the number of consultations, not by outcomes. Labs are reimbursed per test, not per accurate result. The pharmaceutical industry has little incentive to push for narrower ranges, since levothyroxine is off-patent and cheap. The only parties who lose are patients, who suffer in silence. The system prefers waiting because waiting is cheap in the short term. It is only expensive in the long term—but the long term is someone else’s problem.

The case of Margaret, the Manchester teacher, illustrates the point. After a year of fatigue and cognitive decline, her TSH finally rose to 5.2. She was started on 50 mcg of levothyroxine. Within six weeks, her energy improved, she lost 4 kilograms, and her thinking cleared. She told her GP she felt “like a different person.” The delay was not due to medical uncertainty—it was due to a lab range that had not been updated in 50 years. The fix is simple, cheap, and evidence-based. But the counter-arguments remind us that change must be careful: a wholesale shift without resources could cause harm. The middle path—flagging borderline results and encouraging shared decision-making—is both prudent and overdue.

Until that will materializes, GPs must work around the system, one patient at a time. And patients must advocate for themselves, asking their GP: “My TSH is 3.9, but I feel awful. Can we try a low dose of thyroid hormone?” The answer should not depend on a number printed on a lab report. It should depend on a conversation between a doctor and a patient.

This article is for informational purposes only and does not constitute personalised medical advice. If you have symptoms of thyroid disease, consult your GP.