ASA 1 and ASA 2 patients are generally suitable for ED-led procedural sedation. ASA 3 or above warrants anaesthetic / ITU input. Patient selection is where most exam stems start, and getting it wrong from the outset means everything else you do is downstream of an unsafe decision. Remember: in the exam, you are the consultant — you're not asking a senior to make this call for you. You're deciding who else to involve.

Three things go into the decision:

• Patient factors (ASA grade). ASA 1 (healthy) and ASA 2 (mild systemic disease) patients can be sedated by an appropriately trained ED clinician. ASA 3 (severe systemic disease) and above warrant anaesthetic / ITU input — not because ED can't do it, but because the risk profile changes and the rescue plan has to change with it.
• Airway assessment. A focused airway assessment is part of every pre-sedation checklist. Use Mallampati alongside the LEMON criteria (Look externally, Evaluate 3-3-2, Mallampati, Obstruction, Neck mobility). Anticipated difficult airway is a reason to escalate, not just to be cautious.
• Environment and department capacity. The ED is a VUCA environment — volatile, uncertain, complex and ambiguous. The exam may describe a scenario where the resus room is full, the only available nurse is taking a critically unwell patient, or a major incident has just been declared. Don't compromise the safety of the whole department for one procedure. Recognising when to defer or escalate is a testable judgement.

Practically, sedation should happen in a designated area — resus or a dedicated procedure room — with full monitoring, suction and resuscitation equipment immediately available, and staff trained in airway rescue present. None of these are optional.

The standard rule is 2 hours for clear fluids and 6 hours for solids, but fasting is NOT mandatory for emergency procedures if the benefits of proceeding outweigh the risks. This is the exception SBA stems lean on — the safe-looking answer (delay) is wrong, and the right call is to proceed with documented risk-benefit.

Scenarios where waiting six hours is the wrong clinical call:

• Ongoing limb ischaemia from a dislocated joint or compromised perfusion
• An unstable arrhythmia for cardioversion
• An open fracture, a peri-mortem situation, or an active airway problem

In these stems, fasting status doesn't override the procedure.

If you proceed unfasted, you must document the risk-benefit decision — the indication, the risks explained to the patient or family, and your clinical reasoning. In the exam, YOU are the decision-maker. No one is going to make this call for you. “Patient not fasted; procedure time-critical (limb-threatening dislocation); aspiration risk explained and accepted; benefits judged to outweigh risks; decision documented” is not bureaucratic — it's the line that protects the patient and you when the case is reviewed.

The four agents you need to know are propofol, ketamine, midazolam (usually with fentanyl), and nitrous oxide (Entonox). The exam doesn't ask you to recite mg/kg figures — it asks you which drug is appropriate for which scenario, what the major complications are, and which agents have a reversal option.

Adult sedation doses

From the RCEM 2022 Best Practice appendix and the RCEMLearning adult procedural sedation reference.

Emergency drugs — reversal and airway rescue

Sources: RCEM Best Practice Procedural Sedation Aug 2022 (appendix); RCEMLearning adult procedural sedation reference; BNF; UK ALS standard doses; LITFL laryngospasm.

A few exam-relevant principles that follow from the drug profile rather than the drug name:

• Dosing is weight-based and titrated to effect per local protocol. Start lower and slower in older patients — they desaturate faster, recover more slowly, and have a higher complication rate.
• Reversal agents are not a free pass. Flumazenil has a shorter half-life than midazolam, so re-sedation is possible after reversal. Naloxone has a shorter duration than most opioids — repeat doses or an infusion may be needed.
• Propofol and ketamine have no reversal agent. If you give them, you have to be prepared to support the patient through the depth you've achieved.

The mandatory monitoring set is continuous pulse oximetry, continuous capnography, continuous ECG (3-lead, for ongoing heart rate and rhythm), blood pressure every 5 minutes, and ongoing clinical observation of respiratory rate and depth. You always want the heart rate — and ECG leads are the most reliable way to track it. Capnography is the headline for the exam — the single most reliably tested monitoring point on the topic.

Why capnography is the key

SpO₂ tells you the patient is currently oxygenated. It does not tell you that they're ventilating. A sedated patient on supplemental oxygen can be apnoeic for some time before their saturations drop — by which point you've lost a chunk of warning. Continuous capnography (end-tidal CO₂) detects hypoventilation BEFORE desaturation.

What to look for on the trace:

• Loss of waveform — apnoea or disconnection.
• Rising baseline ETCO₂ — hypoventilation, CO₂ retention.
• Obstructive pattern (sloped, shark-fin morphology) — partial airway obstruction.
• Falling ETCO₂ with maintained respiratory rate — drop in cardiac output or ventilation/perfusion mismatch.

The exam point is blunt: if capnography is not in the list of options for “what monitoring is required”, that option is wrong. Treat it like a forced-choice marker.

Staffing — the rule that catches people out

Sedation requires a dedicated sedation practitioner who is separate from the proceduralist. Their job is to deliver the drug, monitor the patient, recognise deterioration, and lead any rescue. They must be trained in airway management and resuscitation. They must not also be the one reducing the shoulder, draining the abscess, or cardioverting the patient.

The minimum staffing is therefore two clinicians — one proceduralist and one dedicated to sedation/monitoring. In practice you usually want a third (a runner / scribe / family liaison), but the testable threshold is two.

Recovery monitoring — after the procedure

Monitoring doesn't stop when the procedure ends. The RCEM guideline is explicit: vital signs are monitored until the patient is awake, alert, and showing no evidence of respiratory depression — using the same parameter set (HR, NIBP, RR, SpO₂, continuous capnography). The guideline doesn't prescribe a fixed recording interval for the recovery period — frequency follows clinical state and local protocol.

• Designated recovery area with monitoring and resus equipment immediately available.
• A clinician capable of managing complications stays immediately available until discharge criteria are met — not just “in the department somewhere”.
• Sedation depth documented through to full alertness on a recognised scale (e.g. ASA conscious depression grades).
• If a reversal agent was given, monitor for up to 2 hours after — flumazenil's half-life is shorter than most benzodiazepines, so re-sedation is real. Same logic for naloxone vs longer-acting opioids.
• Midazolam recovery is longer than propofol/ketamine because of its half-life. Don't anchor your recovery time to the agent you'd like to use.

The complications you must recognise on sight and have a managed sequence for are respiratory depression / apnoea, laryngospasm, aspiration, paradoxical agitation, and cardiovascular instability. The first step in any of them is the same.

STOP the procedure and stop the sedation. This is the line the exam tests as the “most appropriate immediate action” in deterioration stems. It alerts the team, removes the ongoing insult, and creates the space to manage the airway. Everything else flows from this.

Respiratory depression and apnoea (the most common serious complication)

Sequence:

• Stop the procedure and stop further sedation.
• Airway manoeuvres — jaw thrust, head tilt-chin lift, oropharyngeal airway if tolerated.
• Bag-valve-mask ventilation with high-flow oxygen.
• Call for specialist help (anaesthetics) early — don't wait until you've tried everything yourself.
• Reversal agents if appropriate — flumazenil for benzodiazepines, naloxone for opioids. Nothing for propofol or ketamine.
• Escalate to intubation if basic measures fail.

Laryngospasm

Most often associated with ketamine. The algorithm escalates — conservative measures first, drugs last:

1. Stop the procedure and call for help.
2. 100% O₂ via tight-fitting mask + jaw thrust + applied PEEP. This alone often breaks the spasm.
3. Larson's point stimulation — firm bilateral pressure behind the lobule of the pinna (bounded by mandible, mastoid process and skull base), combined with jaw thrust.
4. Deepen sedation with propofol (low-dose, 0.25–0.8 mg/kg).
5. Suxamethonium if still refractory — 0.1–0.5 mg/kg IV low-dose, or 1–2 mg/kg full IV dose; 3–4 mg/kg IM if no IV access. At this point you are committed to a definitive airway and the situation has escalated to full RSI.

Common SBA trap: candidates jump straight to suxamethonium because it's the named drug. The algorithm doesn't — it works conservative measures first. Recognising where the stem is in the sequence is what's tested.

Aspiration

Risk is increased in unfasted patients and obtunded patients. Suction must be immediately available — it's part of the pre-sedation checklist for a reason. Position, suction, oxygen, and escalate; if there's clinical evidence of aspiration, plan for admission and observation.

Paradoxical reactions

Midazolam can cause agitation rather than sedation, particularly at extremes of age and in patients with neurological disease. The clue in the stem is a patient becoming more disinhibited or distressed after the dose, not less. Recognise it, stop dosing midazolam, and reach for a different agent or call for help — don't keep titrating up.

Cardiovascular instability

Propofol classically causes hypotension, particularly in volume-depleted or elderly patients. Titrate slowly, have IV fluids running, and have a vasopressor plan in your head before you start. Ketamine, by contrast, is generally cardiovascularly stable — which is part of why it's used in unwell patients.

Before discharge the patient must have GCS 15, stable observations for at least 30 minutes after the last dose of sedation (longer for midazolam), be walking unaided or back to baseline mobility, be tolerating oral fluids, and have adequate pain control. The exam tests these as a checklist — miss one and the answer changes.

Clinical criteria — all must be met

• GCS 15 and back to baseline mental status. Not “responsive” or “rousable” — back to baseline.
• Stable observations for a minimum of 30 minutes after the last dose, not 30 minutes after the procedure ended. Midazolam in particular needs longer because of its longer half-life and risk of re-sedation.
• Walking unaided, or back to baseline mobility. A patient who normally uses a frame should be back on the frame, not having to be helped down the corridor.
• Tolerating oral fluids.
• Pain adequately controlled with a clear analgesia plan to take home.

Discharge advice — must be given AND documented

• Responsible adult to accompany the patient home and stay with them for 24 hours.
• No driving for 24 hours.
• No alcohol for 24 hours.
• No important decisions or signing legal documents for 24 hours — sedation impairs judgement well beyond the obvious recovery period.
• Written advice provided and documented in the notes.

Sedation is one of the topics where candidates score poorly in FRCEM Final. SLO 6 (where sedation sits) carries about 13 marks across the exam, and sedation is one topic among several within it. The patterns the exam rewards in sedation are predictable and narrow — nail them and you bank marks others lose.

The high-yield exam points, distilled:

• Capnography is mandatory. If it's not in the answer options for monitoring, that option is wrong. Treat as a hard rule.
• Capnography trace recognition. You're expected to identify loss of waveform (apnoea/disconnection), rising baseline (hypoventilation), shark-fin morphology (partial obstruction), and falling ETCO₂ (drop in cardiac output or V/Q mismatch).
• Laryngospasm algorithm. Conservative measures first — 100% O₂ + jaw thrust + PEEP → Larson's point stimulation → deepen sedation → suxamethonium last. Don't jump to sux.
• Dedicated sedation practitioner cannot be the proceduralist. A single clinician doing both is wrong by definition. Minimum two clinicians.
• Fasting has an exception for emergencies when benefits outweigh risks. The exam wants you to spot the time-critical scenario and proceed with documented risk-benefit consent.
• Department-level scenarios are testable. If the stem describes the resus room full and one nurse, recognising that sedation is unsafe right now is the answer. Don't compromise whole-department safety for one procedure.
• Older patients need lower doses, slower titration, and longer recovery. Higher complication risk. The right answer almost never says “use the standard dose”.
• Sedation depth ladder — minimal, moderate, deep, general anaesthesia. ED procedural sedation typically aims for moderate to deep. You must be able to rescue a patient one level deeper than the depth you intended — a patient given moderate sedation can drift into deep sedation, and the team has to be able to manage that.
• Reversal agents at a glance — flumazenil reverses benzodiazepines, naloxone reverses opioids, propofol and ketamine have no reversal.
• First step in deterioration is always STOP the procedure. Not flumazenil. Not intubation. Stop, then escalate.

If you read the seven sections above and can answer each with the headline rule from memory, you've covered the testable surface area of the topic.