Sedation & Anticoagulation Management
- Two separate jobs that both happen before you ever touch the needle: keep the patient comfortable and still (sedation), and keep them from either bleeding out or clotting off (anticoagulation).
- Most IR procedures run on moderate sedation — the patient drowses but still breathes on their own and answers you. The whole game is titrate slowly and watch the breathing.
- Bleeding risk is sorted by procedure: low-risk procedures barely care about your labs; high-risk procedures want a tidied-up coagulation profile first.
- You hold blood thinners before and restart them after on a drug-by-drug schedule — there is no single universal "stop 5 days before" rule, and guessing here genuinely hurts people.
- When in doubt, the question isn't "is this number perfect?" — it's "what's the bleeding risk of this procedure versus the clotting risk of stopping this drug?"
Every interventional procedure has two quiet little problems to solve before the fun part starts. First: the patient is awake, anxious, and about to have a needle threaded into them — so you need them calm and still. Second: you're about to make a hole in something, often a blood vessel, so you need their blood to behave. Not too clotty, not too runny. Goldilocks blood.
Neither of these is glamorous. Both of them are how you avoid a very bad afternoon.
Sedation: the dimmer switch, not the off switch
Think of consciousness as a dimmer switch rather than an on/off light. General anesthesia slams the switch to OFF — the patient is fully out, can't protect their airway, and needs an anesthesiologist running the machine. Most IR doesn't need that. Most IR runs on moderate sedation (the old name was "conscious sedation," which always sounded like a contradiction), where you dim the lights enough that the patient is sleepy and comfortable but still breathing on their own and able to respond when you talk to them.
The classic combo is a benzodiazepine for anxiety and amnesia plus a short-acting opioid for pain — two different jobs, two different drugs. The cardinal sin is pushing both fast, because their effects on breathing stack. So the rule is dull and lifesaving: give a little, wait, reassess, give a little more. Titrate to the patient in front of you, not to a number on a chart.
Sedation lives on a continuum, and patients slide down it without asking permission. You can aim for moderate sedation and accidentally land in deep sedation — where the airway is at risk. That's why someone whose only job is watching the patient (monitoring breathing, oxygen saturation, blood pressure) is non-negotiable, and why reversal agents are kept within arm's reach.
Two reversal agents are worth knowing cold, because they're the "undo" button: naloxone reverses opioids, and flumazenil reverses benzodiazepines. Useful to remember they wear off faster than the drugs they're reversing — so reversing someone isn't the end of the story, it's the start of a longer watch.
Who actually needs to be assessed harder
Before any sedation, you size up the patient's airway and overall risk — the same instinct anesthesiologists formalize. Someone with a difficult-looking airway, severe sleep apnea, or a teetering heart isn't a "give a little and watch" candidate; that's a "call anesthesia" candidate. Knowing where your comfort zone ends is the skill. This dovetails with the broader pre-procedure workup you do during informed consent in IR.
Anticoagulation: risk lives in the procedure, not the patient
Here's the mental model that makes the whole anticoagulation mess click: the bleeding risk is a property of the procedure, not the patient. A skin biopsy and a kidney biopsy carry wildly different stakes if the blood is thin, so the field sorts procedures into low- versus high-bleeding-risk buckets, and that bucket decides how fussy you get about labs and held medications.
| Bleeding-risk category | Examples | How fussy about coagulation |
|---|---|---|
| Low risk | Superficial drains, line removals, shallow aspirations | Minimal — relaxed lab thresholds, fewer drugs held |
| High risk | Solid-organ (liver/kidney) biopsy, deep abscess drainage, arterial access and large-bore procedures | Strict — correct coagulopathy, hold the relevant blood thinners |
The two labs people reach for are the INR/PT (tracks the warfarin pathway) and the platelet count (the actual bricklayers of a clot). High-risk procedures generally want the INR brought down toward normal-ish and the platelets above a working floor; low-risk procedures shrug at modest derangements. I'm deliberately not quoting hard cutoff numbers, because the published thresholds get revised and they differ by procedure — look up the current society guidance for the exact line rather than trusting a number you half-remember.
The single most expensive mistake here is treating all blood thinners as one thing. Warfarin, the direct oral anticoagulants (DOACs like apixaban or rivaroxaban), heparin drips, and antiplatelets like aspirin and clopidogrel each have different hold-before and restart-after timing. There is no universal "stop it 5 days prior." Each drug has its own schedule, and DOAC timing also flexes with kidney function.
The two-sided question you're always really asking
Holding a blood thinner isn't free. Stop someone's anticoagulation and you may have prevented a procedural bleed while quietly inviting the stroke or clot the drug was preventing in the first place. So every decision is a tug-of-war: the bleeding risk of doing the procedure on thin blood versus the clotting risk of stopping the drug. A patient with a recent stent or a mechanical heart valve sits very differently in that tug-of-war than someone on aspirin "just in case," and those high-stakes cases earn a real conversation with the prescribing team rather than a reflex hold.
Don't forget the back end. The procedure isn't over when the needle comes out — when does the blood thinner go back on is its own decision, balanced against the fresh puncture you just made. Restart too early and you bleed; too late and you clot. This is also where access-site management lives — see closure devices and access complications.
If you remember one thing: sedation and anticoagulation are both exercises in titration against risk. Sedation, you titrate the drug to the patient's breathing. Anticoagulation, you titrate the held medications to the procedure's appetite for bleeding. Neither rewards a fixed recipe — both reward thinking about the specific human on the table. (And once they're calm and their blood is behaving, you still get to worry about contrast and radiation in IR — but that's a problem for the next page.)