In opioid titration, sedation is not the endpoint

Why it matters

The PACU is where opioids are given fastest and where the person giving them is most tempted to stop when the patient’s eyes close. The sources for this day watched that moment. In Paqueron 2002, titration was stopped at sleep onset, and 13 of the 52 patients who slept (25%) still had a visual analogue pain score above 50 mm at 30 min paqueron-2002. In Frasca 2007, patients who left the PACU sedated but still in pain reported severe pain at 24 h in 50% of cases against 6% of matched controls frasca-2007.

The consequence of getting this wrong is a patient who is asleep, in pain, and about to wake up on a ward where the next dose is an hour away.

Mechanism

Morphine acts on the reticular activating system and on the nociceptive pathways at different rates and different doses. Sedation is a central effect that can appear at plasma concentrations that have not yet produced analgesia, particularly in a patient who is exhausted, has had a long anesthetic, or is receiving boluses faster than the drug equilibrates with the brain. Analgesia follows, or does not, on its own curve. The two are therefore separable at the bedside, and a titration rule that uses the first as a proxy for the second will stop early in a predictable fraction of patients. Function, a patient who reports relief and can breathe deeply, cough and move, is the endpoint that both effects have to pass through.

Evidence

Paqueron 2002 titrated 73 PACU patients with intravenous morphine and stopped at sleep onset; 52 slept and 21 did not paqueron-2002. At sleep onset the Ramsay score rose from 1.7 (0.4) to 2.4 (0.6) and the bispectral index fell from 95 (5.0) to 89.8 (10.2); pain scores fell comparably in both groups, from 78 (17) to 39 (21) in sleepers and from 64 (16) to 30.4 (11) in the awake group, but 13 of 52 sleepers (25%) still had a VAS above 50 mm at 30 min paqueron-2002. The authors’ conclusion, verbatim in sense: morphine-induced sedation should not be considered an indicator of an appropriate level of analgesia during intravenous morphine titration paqueron-2002.

Frasca 2007 compared 26 patients with a Ramsay score above 3 and a verbal rating score of 3 or more at PACU discharge with 52 matched controls frasca-2007. The sedated group reported severe, moderate and no PACU pain in 58%, 16% and 26% against 18%, 25% and 57%; a bad first night’s sleep in 54% against 10% (P=0.001); severe pain at 24 h in 50% against 6% (P less than 0.0001); and dissatisfaction in 20% against 2% frasca-2007.

Aubrun 2003 describes the titration itself in 3045 patients: a bolus of 2 mg (60 kg or less) or 3 mg (more than 60 kg) every 5 min, with a threshold VAS of 30 aubrun-2003. Mean initial VAS was 73 (19); mean dose to relief 0.17 (0.10) mg/kg, median 4 boluses (range 1 to 20); and a VAS of 70 or more predicted a requirement above 0.15 mg/kg with sensitivity 0.77 and specificity 0.54, along a sigmoid rather than linear relationship aubrun-2003. The earlier quality program that compared four regimens found that unlimited 5-min boluses with early subcutaneous morphine produced the lowest end-PACU pain score, 26 (17) mm, with 73% relieved, at the cost of sedation in 61% against 27% with the most restrictive regimen aubrun-2001.

The ASA practice guideline on acute pain management in the perioperative setting exists as a society source for this day; its identifiers were read but its text was not, so nothing from it is quoted asa-acute-pain-2012.

What this does not show

Every source is observational. Paqueron 2002 is a cohort, Frasca 2007 is a case-control study, and the Aubrun series are quality programs and cohorts paqueron-2002 frasca-2007 aubrun-2003 aubrun-2001. They show that sedation and analgesia dissociate and that leaving the PACU sedated and in pain predicts a bad day. They do not test a titration rule against another, and no trial on this page shows that changing the endpoint changes the outcome.

The Aubrun regimens are one institution’s protocol. The bolus sizes, the interval and the threshold are what that group used, not a standard aubrun-2003. Local practice differs. the opioid, bolus size, interval, pain-score threshold and sedation limit used for PACU titration differ between institutions; follow local protocol and know its endpoint.

The guideline that would carry a general recommendation has not been read: That guideline’s own statements on opioid titration, and whether it is still current, have not been read. The 2012 review by the Aubrun group is a narrative review and stays a finding aid rather than support.

The multimodal recommendation this page lost now has a source, though a qualified one. A 2024 multidisciplinary consensus statement asks that multimodal analgesia be encouraged throughout the perioperative pathway where not contraindicated, naming paracetamol, non-steroidal anti-inflammatory drugs, alpha-2 adrenergic agonists, N-methyl-D-aspartate antagonists, corticosteroids and regional analgesia, and that intraoperative analgesia be multimodal and extend into the postoperative period elboghdadly-2024. Its own account of its method matters as much as its content: it did not include a full systematic review with quantitative or qualitative synthesis, nor grading of recommendations, so its recommendations are unnumbered and ungraded and it says several are supported by modest evidence elboghdadly-2024. Take it as a statement of what a multidisciplinary group agreed, arrived at by a four-round Delphi with a 75% consensus threshold, rather than as a graded recommendation elboghdadly-2024.

The United States guideline usually credited with this material is in the ledger and its recommendation text has still never been read, so nothing is quoted from it here chou-2016. A newer document was checked as a possible replacement for it and does not replace it: different jurisdiction, different societies, explicitly ungraded, and it cites the older guideline as its own reference rather than superseding it elboghdadly-2024.

One currency question on this page is open and the reader should know it. A 2026 American Society of Anesthesiologists practice guideline exists on perioperative pain management using local and regional analgesia for cardiothoracic, mastectomy and abdominal surgery in adults and children asa-2026-regional-analgesia. Its recommendations have not been read, so nothing here is quoted from it and no source on this page has been marked superseded by it. Whether it changes any of what is above is an open comparison, not a settled one.

At the bedside

Titrate to a report of relief and to function, and write down what the endpoint was. A patient who says the pain is tolerable, breathes deeply and can cough has reached it; a patient who has stopped answering has not, and a quarter of those patients are still in pain paqueron-2002.

Treat sedation during titration as a limit on the rate, not as the goal. If the patient is drowsy and still reports pain, the choice is between slowing the boluses and adding a non-opioid, not between more morphine and stopping aubrun-2001.

Do not discharge a patient who is sedated and still in pain as if the sedation had solved the pain; the case-control data say that patient has a bad night and is still in severe pain the next day frasca-2007.

Check yourself

Three items. Every option carries an explanation. Progress is not saved.