In the Dark: Harm Management for People Already Sourcing Hormones

A note before this one — from me

This is a piece I thought hard about before writing, and not everyone I trust agreed I should. I wrote it anyway, for one reason: the system is, right now, failing some people completely — and when it does, they don’t stop; they’re left to do the most dangerous version of this entirely alone. That doesn’t make self-sourcing safe, and this piece won’t pretend it does. But refusing to say anything useful to someone already doing it isn’t caution. It’s an indictment of the situation we’ve been put in, dressed up as principle.

I started this site to answer one question: can I make this a little less lonely, a little safer, a little more understandable for the next person? This guide is the “a little safer” part, and nothing more. It is not here to talk you into anything, and it is not written to lecture you.

So if this is where you are right now: it was written for you. Read it, be careful, and please — do everything you can to get onto a medically supervised programme. Not because I’m required to say so, but because what I wish, every time I write one of these pages, is that nobody in the dark had to do it alone.

— Eden

The main guidance on this page is for adults who have already taken, or are currently taking, self-sourced hormones. It will not help anyone decide whether to start, and it does not tell anyone how to obtain, dose, apply or inject anything. Nothing on this page is intended to help anyone — least of all a minor — obtain or take hormones outside clinical care. A separate safety notice below is for a parent after a child’s exposure; the adult monitoring information must never be used for a child.

Why this page exists

Self-sourced hormones are, for many, a symptom; the deeper problem is a care system that, for too many people, has become unreachable. This piece does not recommend self-medication, and it will not tell anyone how to obtain, verify, dose, apply or inject a drug bought outside a clinic. It begins from a more uncomfortable premise: that people are already doing it, that stop-only advice — telling people to stop with no assessment, follow-up or reachable alternative — does not reliably end self-medication and can leave those who continue with no clinical support at all, and that for someone who cannot or will not stop right now, one question is more urgent than the rest: whether they do it connected to a clinician, or entirely alone. Harm reduction is not encouragement. Needle-and-syringe programmes did not cause injecting; they reduce its preventable harms and keep people connected to services long enough to reach the door marked help. Refusing to apply that logic here — out of squeamishness, or fear of how it photographs — is a decision to let the worst outcomes happen quietly.

The fight to make the door exist — to get every person, adult or child, the care and supervision they are owed — is real, and it is ours. It is not this page. This page is smaller and colder. It is for the people who are already doing it, tonight, and it is about one thing many urgently need and often lack: someone who has weighed the risk with them, and is close by when it turns.

The people this is for

Somewhere tonight, someone is measuring out a dose by the light of a phone. Maybe an adult who was on a waiting list long enough to stop believing in it. Maybe someone whose doctor said no, or said nothing, or said not while I’m your doctor. Maybe someone who cannot afford the fare to the one city where the private clinic is. Maybe a parent — steady, terrified, out of their depth — whose child was under care until a letter arrived and the care simply stopped. For most, this was not what they wanted. It was the end of a road: the care they could reach proved absent, unaffordable, or impossible to use in time. They are not doing this because they are careless.

Before anything else — for any parent whose child or teenager has already been given a self-sourced hormone: this page is not for that, and none of the adult monitoring below is a substitute for proper paediatric care. Do not assume a dose is either harmless or an emergency — the urgency depends on the substance, the amount, the route, the timing and the child’s symptoms. Contact a paediatric or adolescent gender or endocrine service, or a poisons or medical advice line, promptly, and tell them exactly what was given. They can advise on how urgent it is, and whether the child also needs paediatric, endocrine or specialist gender follow-up. Use emergency services if the child is seriously unwell — struggling to breathe, collapsing, having a seizure, or otherwise acutely ill. Which advice line to call is region-specific, and some are for clinicians only, so the local numbers are in the service box further down.

This is the part clinical warnings usually leave unwritten. It is easier — safer, cleaner, better for the reputation — to say do not do this, see a doctor, and let the sentence end there, knowing the door it points to is locked. I am not going to do that. Not because self-sourcing is safe; it is not, and this piece will be blunt about how it is not. But because pretending people will stop, when they will not, is not caution. It is abandonment with good posture.

What this is, and what it refuses to be

Let me be exact about the stance, because the whole piece rests on it and because the ambiguity is precisely where the danger lives.

I do not recommend self-sourcing hormones. Not for an adult, and emphatically not for a child. Gender-affirming therapy is medicine, and medicine is safest supervised — assessed, monitored, adjusted by someone trained to see a thing going wrong before the patient can feel it. Everything that follows starts from that position and never leaves it.

But a second fact sits beside the first, and holding both at once is the entire discipline of harm reduction: people are already doing this, and being told simply to stop does not reliably change that. This is not a fringe position. The World Professional Association for Transgender Health’s Standards of Care Version 8 state that non-prescribed hormone use should be approached through a harm-reduction lens, with the aim of connecting people to providers who can offer safe and monitored hormone therapy. The people this piece is written for arrive by different routes — cut off, refused, priced out, unable to wait, unwilling to disclose — and the safety in it applies whatever brought them here. The real question is therefore not whether people self-medicate — they do — but whether they stay connected to care, with a better chance of risks and complications being recognised, or harm themselves quietly and out of reach, in a way no one finds out about until it is too late to matter.

So this is a harm-management document, and its subject is monitoring, not method. It will tell you what to watch, which warning signs mean get help now, and how to stay tethered to a clinician even while you are sourcing outside one. It will not — and I want this in plain sight, not buried under a disclaimer — tell you where to buy anything, how to obtain it, how to check a product, what dose to take, or how to apply or inject it. Those are the lines between reducing harm and writing a manual, and I will not cross them. The dosing evidence lives in the clinical reviews — feminising hormone therapy and hormone monitoring and target ranges — behind a clinician’s judgement, which is exactly where it belongs. And there is a hard limit the other way, too: monitoring cannot confirm that a self-sourced product is genuine, that it contains the labelled strength, that it is sterile where sterility is required, or that it is free of contaminants — and it cannot prevent every serious complication. It lowers some risk. It removes none.

And there is one boundary firmer than all the rest. This document is written for adults making decisions about their own bodies. Where a child is involved, the logic does not soften — it hardens. An adult self-monitoring is exercising autonomy imperfectly; a child medicated at home is a patient with no one reading the numbers, and for a growing body those numbers move in ways a home cannot see and a kitchen cannot measure — the bloods, the pressure, the bone, the dose that has to change as the child does. What a parent cannot replace is qualified, accountable paediatric assessment and supervision, even where parts of legitimate care happen remotely or at home. That is the honest thing this page can say about it: not a protocol, but a direction — back to a service that can supervise. The fight to build that route, for every person of any age, is real, and it is ours. It is not this page.

Why the door is locked

Information alone does not explain why people self-source; barriers to supervised care are often an important part of it. But it would be dishonest to pretend words on a page never influence what people do — which is exactly why this page carries no procurement, no doses, and no application or injection technique. What it can do without inducing anything is describe the doors that were meant to open and did not, because a piece that treats self-sourcing as a personal failing rather than a systemic output has misunderstood the thing it is describing. To explain how people arrive here is not to recommend the journey. It is to refuse the lie that they chose it freely.

Start with the wait. In much of the world the route to supervised care runs through a list, and in many places the list has stopped being a queue and become a wall — measured not in weeks but in years, long enough that a referral can outlive the resolve of the person who made it, long enough that wait and refuse become, in practice, the same word. A system can deny care without ever saying no. It only has to be slow enough.

Then the cut-off, which is cruel in a particular way because it follows a kindness. Care that was flowing, and then stopped: a clinic closed, a contract ended, a policy changed, a threshold of age or eligibility crossed, a prescriber who retired or withdrew. People who were stable, monitored, doing everything that was asked of them, wake up unsupplied — and a person who has had care taken away does not stop needing it. They go looking. This is the door slammed on someone already through it.

Then the refusal at the front line — the clinician who will not prescribe, will not monitor, sometimes will not discuss it at all; the not while I’m your doctor that leaves a person medically homeless in the middle of their own care. Add cost, which turns the private alternative into a door with a price on it that not everyone can pay. Add distance — the single clinic a day’s travel and a fare away, so that geography quietly decides who gets seen. Each of these alone is survivable. Stacked, they are a machine, and the machine has an output, and the output is a person at a kitchen table at midnight.

And underneath all of it, the slow lesson the system teaches: that it is an obstacle to be got around rather than a place of help — that honesty will be used against you, that asking will be met with delay, that the safest posture toward the people meant to care for you is caution. Once someone has learned that lesson, the forum feels safer than the phone. That is not their error. It is the system’s, printed onto them.

None of this is universal, and it would be its own kind of dishonesty to pretend it were. The failures are real, and for some people they are total — but not everyone is failed, and for many, supervised care is still within reach. If it is within reach for you, reach for it. This piece describes what to do when the door is locked; it is not an argument for walking away from one that is open.

None of this makes self-sourcing safe, and none of it is a case for doing it — the sections that follow are blunt about the risks and they stand. It is a case for something else: for aiming the judgement at the arrangement rather than the person, and for seeing that the surest way to empty the dark is not a better set of instructions. It is a door that opens.

What the dark hides

Start by naming the danger accurately, because it is two dangers, not one. The hormone carries real risk — always, in any hands; supervision manages that risk, it does not abolish it. What self-sourcing often adds on top is further risk again: the product may be uncertain, the assessment incomplete, the follow-up fragmented or absent. These are established medicines, used across many parts of medicine, and taken under supervision — with the right checks beforehand and monitoring after — they are safe enough for routine long-term use; but their risks depend on the formulation, the route, the exposure, other medicines and the person. What may go missing outside a regulated pathway is the structured assessment, the known supply, and the follow-up that could catch some risks or help when things go wrong. And some of the most serious complications develop with no clear early warning — which is why a clinician cannot promise to catch them, and why knowing the warning signs yourself matters as much as any blood test.

Take one that matters for many people on oestrogen: a clot, in a deep vein or the lungs. Some oestrogen regimens raise the tendency to form clots — and how much depends heavily on the compound, the route, the exposure and a person’s other risk factors. You may carry meaningful risk without feeling anything at all, and a clot may be the first sign that the risk was there. No routine hormone-monitoring blood test predicts one coming; tests such as D-dimer are used by clinicians when a clot is already suspected, not as screening. What can lower the odds is an assessment beforehand that actually changes something — the route, a modifiable risk factor, the choice of drug. The history here is instructive, and it cuts against the panic as much as toward it. Some of the highest clot rates in older transfeminine cohorts were tied to regimens containing ethinylestradiol — an oestrogen no longer recommended for gender-affirming care because of its poor thrombotic profile — though those old cohorts also differed in age, smoking and how cases were counted, so the whole of the excess cannot be laid at one door. Route matters, too. Transdermal 17β-estradiol — patches or gels — avoids the first pass through the liver and appears to carry a lower clot risk than the same hormone taken as an oral tablet: a lower risk, not a zero one, and the evidence in trans people is thinner than the menopausal data it leans on. Injectable estradiol also avoids that first pass, but it produces substantial peak-and-trough swings rather than steady levels, and can produce supraphysiological estradiol concentrations depending on the preparation, the amount and the dosing interval. The evidence comparing its clot risk directly with oral or transdermal estradiol is very limited — so do not assume it matches the low-risk transdermal picture, and do not assume it is worse either; the honest answer is that it is not established, which is one more reason product, timing and clinical interpretation matter. In a clinic these are conversations you would have. In the dark they are decisions made blind.

Testosterone hides its danger differently. Over months it can raise the concentration of red cells in the blood — measured as haemoglobin and haematocrit. A marked rise needs prompt clinical review; it may add to the risk of clots, stroke or cardiovascular events, but how large that risk is — and how much the raised red-cell count itself is to blame rather than the things that travel with it — is not well established in people on gender-affirming testosterone. A raised count can also have other causes — smoking, sleep apnoea, lung disease — which is another reason not to read it alone. You will not feel it happening, and it is one of the things a blood test genuinely can catch before it shows as anything you would notice, which is why clinicians watch those numbers. Blood pressure can change on hormone therapy, too — the direction and size vary between people and regimens — which is reason enough to keep an eye on it. Serious liver injury, by contrast, is not a typical effect of standard 17β-estradiol regimens or of standard injectable or transdermal testosterone. Where it does happen, the cause is usually something specific — older oral androgens, or particular anti-androgens. Cyproterone acetate can cause liver injury, including rare severe cases; the risk looks greater with higher exposure, but injury has been reported at lower doses too, so there is no dose you can treat as safe. Flutamide has a well-established risk of severe liver injury and is generally avoided where safer options exist. Their use may call for baseline and follow-up liver testing under product-specific or local guidance, though testing cannot head off every severe reaction — and a self-sourced product cannot be assumed to be any of the safer preparations. None of these decides itself. Every one is a reason to get assessed — and depending on your history, your symptoms and what the tests show, a clinician might tell you to carry on, change something, pause, or stop. That is a judgement, and it is not one this page, or a home test, can make for you.

And then there is the product itself. A drug that comes through a regulated clinical and pharmacy pathway is a comparatively known quantity — a stated strength, quality controls, a pharmacist in the chain, traceability if something goes wrong — though no supply chain is entirely error-free. A drug that comes another way might be exactly what the label says — or under-strength, over-strength, or not only what it claims — and you have no reliable way to tell by looking; that uncertainty of identity and strength applies to tablets, gels and patches as much as to anything in a vial. For an injectable there is a further layer: whether it is sterile, and free of particulate or microbial contamination. A blood test helps, but only so far: it may show that your exposure is unexpectedly high or low, but reading it depends on the compound, the route, the timing and the clinical picture, and it cannot establish that a product was pure, sterile, or exactly the strength it claimed. That gap is not something you can close at the kitchen table. It is one more reason not to be there alone.

There is a subtler danger, and it is the one people miss. A single measurement rarely tells the story; risk usually lives in the trend — the count that is climbing, the pressure creeping up, the level drifting from where it began. Supervised care watches the line, not the dot. Self-sourcing, more often than not, has neither a baseline to start from nor a record to measure against. The dark does not only hide the danger. It erases the memory you would need to recognise it.

So when this piece says get assessed and monitored, it is not moralising. It is naming what turns some of this from invisible back into visible — and admitting what stays invisible even then. The hormone was always a risk, in any hands. What the dark takes away is the person who would have weighed that risk before it started, and stood close when it turned.

What monitoring can catch — and what it can’t

Monitoring is not bureaucracy, but it is not a crystal ball either, and the difference is the whole of this section. What it can do is identify some potentially important abnormalities — a red-cell count climbing, a blood pressure creeping up, an electrolyte pushed out of range by another drug — before they become more severe or cause obvious symptoms. What it cannot do is see a clot coming, or a stroke, or a reaction to a contaminated vial; those will not show on a routine panel. Some complications first appear as symptoms rather than on a test, and knowing the common warning signs can help you get care sooner — but symptoms can be absent, subtle, or noticed by someone else before you notice them yourself, so this is not a substitute for assessment either. Monitoring lowers the odds against you. It does not clear them, and it cannot make an unregulated product safe.

What to watch, and how often

Endocrine Society guidance for prescribed therapy recommends clinical review roughly every three months through the first year, with the actual blood tests chosen to fit the hormone, the accompanying medicines and the person — hormone levels, a red-cell count on testosterone, potassium where spironolactone is involved (hormone monitoring and target ranges). These are examples, not a complete test list, and repeating them without the right timing or clinical context can mislead as easily as inform. Once things are steady, review is usually less frequent. Two honest caveats travel with all of it: the schedule was written for known, prescribed regimens, so it is an imperfect fit for an unknown product; and it is a reference point for a conversation, not a guarantee — a clinician may want earlier, later or different testing depending on how uncertain the product is, your symptoms, other medicines and previous results.

What the blood is looking for depends on direction. On testosterone, haemoglobin and haematocrit — the red-cell measures — matter especially, because testosterone can raise them; a clinician also weighs testosterone exposure, blood pressure, symptoms and your other cardiovascular risks. On oestrogen, it is the hormone level plus whatever else your medicines and history call for, chosen by a clinician rather than off a menu — for instance, spironolactone usually means checking potassium and kidney function, because it can raise potassium, more so where kidney function is reduced, or where you also take common anti-inflammatory painkillers (NSAIDs, such as ibuprofen), certain blood-pressure medicines (ACE inhibitors or ARBs), potassium supplements or salt substitutes, any of which can push it higher still. That is one of the clearest reasons to tell a clinician or pharmacist everything you take, prescription or not — and to tell whoever reads your results about any recent change in product, route, source or timing, because it can shift both the risk and how a result should be read. The numbers themselves, and what to do about them, are a clinician’s to read.

Blood pressure is the one you can most easily keep an eye on yourself. A validated upper-arm monitor can help where a clinician or pharmacist recommends home measurement and can guide you on cuff size, technique and what the readings mean; treat readings that sit repeatedly above the level they give you as a reason to be seen, not to change anything alone.

Why a clinician, and not just a blood test

A blood-pressure reading you can take at home; blood tests can sometimes be arranged outside a clinic — but the correct test, the sampling, the timing and the interpretation still need appropriate clinical and accredited laboratory oversight, and they are not the whole of what supervision is. Some things change your risk and never show up on a home panel: whether you smoke, because smoking adds cardiovascular risk and can change which oestrogen route is sensible and how the whole balance is weighed; a personal or family history of clots; migraine, especially with aura, which a clinician weighs against the formulation, route and your other vascular risks rather than importing the contraceptive-pill rules wholesale; heart or kidney disease; other medicines that interact with what you are taking. And one that is missed more than any other: testosterone is not contraception. Pregnancy can still happen on it, and testosterone can harm a developing pregnancy, so anyone who could become pregnant needs proper contraceptive and pregnancy advice. Other conditions, medicines and reproductive considerations can matter too — which is exactly why a generic page cannot stand in for a baseline assessment. None of this is a checklist to clear yourself against — reading it, and weighing the parts against each other, is a clinician’s work, and the fact that it exists at all is one of the strongest reasons this piece keeps turning you back toward one. Nor is any of it the whole field: people also use cyproterone, GnRH analogues, progestogens, 5-alpha-reductase inhibitors and combinations, each with its own risks and its own monitoring. What is here is an illustration, not a complete plan.

Avoiding infection and injury

This page does not teach you how to apply or inject anything — application and injection differ between products, and the instructions belong with the specific product and with a person who can show you. Where the product and its manufacturer’s information can be reliably identified, follow the product-specific precautions and check them with a pharmacist — but treat neither packaging nor a leaflet as proof that a self-sourced product is genuine, because counterfeits copy both. What this page will do is name the hazards, because those are general.

Some hormone gels and other leave-on skin products can transfer to another person through contact with the treated skin — a documented harm, and the reason such products carry secondary-exposure warnings; the precautions protect children and anyone else who might touch the site. Patches are different: a patch can detach, and a used patch can still hold active drug, so both need careful disposal. Keep gels, patches and used patches away from children.

Injecting carries the harms that hygiene exists to prevent: reusing or sharing needles or syringes, using damaged or contaminated equipment, or injecting through broken or infected skin can all cause serious injury and infection. A nurse, pharmacist, sexual-health service or needle-and-syringe programme may be able to help — ask whether the service can provide sterile equipment, sharps disposal and appropriately trained injection-safety support; availability and scope vary, and many programmes are framed around drug use rather than hormones. That contact is worth having for its own sake, because it is a person, and a door. Do not treat written or recorded instructions as a substitute for product- and route-appropriate teaching from an appropriately trained healthcare or harm-reduction worker — and remember that even good teaching cannot make an unregulated product itself safe. Afterwards, watch the site: if it turns red, hot and swollen, get it seen the same day — and if that comes with fever, shivering or feeling very unwell, treat it as an emergency.

When it is an emergency

Some things do not wait for a blood test or a same-day appointment. If any of these happen, get emergency help now, whatever it costs in fear or awkwardness:

• New unexplained swelling, pain or tenderness in one leg — often in the calf or thigh, with or without warmth, redness, or an obvious difference between the legs — a possible deep-vein clot.
• Sudden breathlessness, chest pain that worsens when you breathe in, or coughing up blood — a possible clot in the lung.
• Sudden weakness or numbness in the face, arm or leg — especially on one side; sudden trouble speaking or understanding; sudden vision change; a sudden severe loss of balance or coordination; or a sudden, exceptionally severe headache — possible signs of a stroke or other brain emergency.
• New severe or lasting chest pressure, tightness or pain — especially with breathlessness, sweating, nausea or faintness, or discomfort spreading to an arm, the jaw, the back or the upper abdomen — a possible heart attack; it is not always dramatic or classic, particularly in women, older people and people with diabetes.
• Swelling of the face, tongue or throat, difficulty breathing, a spreading rash with feeling very unwell, or collapse — a possible severe reaction.
• Rapidly spreading redness or swelling, severe pain, pus or red streaking, fever or shivering, confusion, faintness, fast breathing, or simply feeling severely unwell — a possible serious infection, which does not always come with a high fever.

You do not need to be certain, and you do not need to diagnose yourself. Emergency staff would far rather see you and send you home than not see you at all. And never let the fear of being judged for self-sourcing cost you an hour — no clot cares how you came by the hormone, and no emergency clinician worth the name makes your care conditional on approving of it. Tell them what you are taking — plainly, without shame — because it is an important piece of information that can change how they assess and treat you. If you can do it safely, bring the product, or a clear photograph of its packaging and any batch or ingredient details; it may help them, even though a label cannot prove the product is genuine.

Between the routine and the emergency sit things that need acting on but are easy to misjudge. Some need urgent, usually same-day assessment: new or worsening breathlessness, especially if it has crept into things that never used to wind you; new yellowing of the eyes or skin, particularly with dark urine; a new, persistent or worsening headache without weakness, confusion, seizures, major vision changes or the other emergency signs above; a blood pressure that stays high. Do not let these sit for days. And treat as an emergency, straight away: sudden or severe breathlessness, or breathlessness with chest pain, fainting or coughing blood; a sudden severe headache, or a headache with new weakness, numbness, confusion, seizure, speech trouble, major vision change or loss of coordination; or a markedly high blood-pressure reading together with chest pain, confusion or neurological symptoms.

Which is the whole point this piece has been travelling toward. One of the most important things a person already self-sourcing can do is not do it entirely alone. That does not always mean the door you were turned away from; it can mean a telehealth service with a licensed clinician that includes blood monitoring, a bridging prescriber, a sexual-health clinic — any clinician willing to run the tests and read them, even one who will not prescribe. A clinician can sometimes assess your risks, arrange the right tests and treat complications even if they never prescribed or supplied the hormone; that is not the same as anyone confirming the product is safe, because routine clinical monitoring cannot confirm that the product itself is genuine, of the labelled strength, sterile or uncontaminated. But being assessed by someone who never handed you the drug puts a trained eye on the risk, catches the things that can be caught, and gives you someone to reach when the things that can’t announce themselves.

And give any clinician who treats you an accurate account of what you take, when it is relevant — emergency staff, prescribers, and above all a surgeon and anaesthetist. Raise your hormone use with a surgical and anaesthetic team well ahead of any planned operation, and do not suddenly stop or change it on your own: they have to weigh your individual clotting risk against the risk of abrupt withdrawal, and that is their call to make with you, not one to make alone. It feels like the most dangerous thing to disclose and it is very often the safest: a clinician who knows what is in your body can weigh what you cannot; one kept in the dark is guessing when the guessing matters most. Disclosure is not surrender. It is a hand on the wall in a dark room.

Sourcing outside a clinic is a risk taken. Doing it unassessed, unmonitored, unwilling to be seen — that is the risk that compounds every other. Do not add it to them. Stay visible.

In closing

Everything in this piece is a workaround, and workarounds should make you angry — not at the people improvising in the dark, but at the arrangement that put them there. There is a safest possible harm-reduction intervention for self-sourced hormones, and it is not a monitoring schedule, or a warning sign, or a hand on the wall. It is a clinic that answers the phone. It is a waiting list measured in weeks. It is a clinician who treats a trans patient as a patient. Every word above exists because those things, for too many people, are not there — and no amount of careful self-monitoring is a substitute for being cared for. This is triage on a battlefield that should not exist.

So hold two things at once, because both are true. Sourcing your own hormones is not safe, and this piece has not for a single sentence pretended otherwise. And the people doing it are not reckless or foolish — many are people who wanted an ordinary thing, care, and could not reach it, and chose to be seen and assessed rather than face it entirely alone. Being assessed and monitored is not safety: it cannot make the hormone safe, or the product genuine, or catch every complication. It is only better than the dark — and that is the single thing this page will call better, not the sourcing, not the going-without, only the refusal to face it unseen. They should never have been left to make the choice at all.

If you are one of them: get assessed, learn the signs, put a clinician in the loop, and refuse — however thoroughly the system has trained you to expect nothing — to disappear. And if you work in the system that made this necessary: this is what it looks like from where your patients are standing. The dark is full of people doing your job for you — alone, afraid, and far better than they should have to. Turn on a light.

Declarations

Competing interests. The author is a trans woman who supports access to safe, supervised gender-affirming care. This interest is not concealed — it is the reason the piece exists, and it is stated plainly in the author’s note at the top. The guidance was written and reviewed to a harm-reduction standard, reaching for what is safest for the reader rather than what is most persuasive for a cause; where the honest answer is that something is dangerous or uncertain, the piece says so.

Preparation and authorship. This piece was prepared with substantial assistance from a large language model (Claude, Anthropic), used for drafting, source retrieval, arithmetic and citation verification, structural editing, and adversarial review across several rounds. Every clinical claim, every citation, and the boundaries of the piece were reviewed and verified against primary sources by the author, who is responsible for all of them.

Sources

The graded, fully-cited evidence lives in the companion Evidence Reviews; the anchors below support the specific claims made here.

• Harm-reduction framing. Coleman E, et al. Standards of Care for the Health of Transgender and Gender Diverse People, Version 8. Int J Transgend Health. 2022;23(Suppl 1):S1–S259 — Introduction, “Health care services”, p. S8: non-prescribed hormone use should be approached with a harm-reduction lens, connecting people to providers who can prescribe safe, monitored therapy.
• Monitoring cadence and routes. Hembree WC, et al. Endocrine treatment of gender-dysphoric/gender-incongruent persons: an Endocrine Society Clinical Practice Guideline. J Clin Endocrinol Metab. 2017;102(11):3869–3903 — review roughly every three months in the first year, then less often; oral/transdermal/injectable 17β-estradiol; ethinylestradiol not recommended. ARUP Consult, Endocrine testing in transgender adults — haemoglobin/haematocrit before and through the first year on testosterone.
• Clot risk, route and formulation. Canonico M, et al. (ESTHER). Hormone therapy and venous thromboembolism among postmenopausal women. Circulation. 2007;115(7):840–845 — oral vs transdermal. Totaro M, et al. Risk of VTE in transgender people undergoing feminising therapy: a prevalence meta-analysis. Front Endocrinol. 2021;12:741866. Gender-affirming hormone therapy in the transgender patient: influence on thrombotic risk. Hematology Am Soc Hematol Educ Program. 2024 — ethinylestradiol with cyproterone acetate ~45-fold VTE; transdermal among the lowest-risk routes.
• Injectable estradiol. The use of injectable estradiol in transgender and gender diverse adults: a scoping review of dose and serum estradiol levels. Endocr Pract. 2024 — guideline-range doses can produce supraphysiological levels; direct comparative clot data versus oral/transdermal very limited.
• Anti-androgen hepatotoxicity. LiverTox: Cyproterone (NCBI Bookshelf NBK548024) — cyproterone and flutamide carry greater hepatotoxic potential than bicalutamide/nilutamide; severe cases, including at lower cyproterone doses.
• Spironolactone + hyperkalaemia. MHRA/GOV.UK Drug Safety Update — spironolactone with ACE inhibitors/ARBs: risk of potentially fatal hyperkalaemia; NSAIDs, potassium supplements and salt substitutes add to the risk.
• Topical-hormone transfer. FDA boxed warning (2009, strengthened 2022) on topical testosterone; product labelling — secondary exposure and virilization in children from skin contact; precautions to prevent transfer.