Respiratory Depression in the PACU and a PCA Programming Discrepancy

Every element points to opioid effect - small pupils, slow respiratory rate, preserved response to pain, and a well-controlled surgical site - but the source is layered:

• Systemic opioid: 1 mg IV hydromorphone intraoperatively plus a demand dose. If the pump actually delivered 2 mg, that single demand is ten times the ordered dose and by itself explains profound hypoventilation in an opioid-naive 95 kg man.
• Neuraxial opioid: 300 mcg of intrathecal morphine is above the usual 100–200 mcg range. It has slow onset and a long duration, with the classic window for delayed respiratory depression at 6–12 hours, but early depression within the first hour also occurs, especially with concurrent systemic opioid.
• OSA and obesity: reduced functional residual capacity, blunted arousal, and additive sedation.

The pump discrepancy is not a footnote: it changes the working diagnosis from expected side effect to probable medication error, changes the expected magnitude and duration of the opioid load, and requires that the pump be stopped and the programming verified before anything else is given. It also means you must look for a second patient or a second pump from the same batch, and it triggers reporting obligations that a simple opioid side effect would not.

Immediate actions, in order:

• Stop the PCA pump and disconnect it; verify the programming against the order and keep the pump as evidence.
• Stimulate and open the airway with jaw thrust; apply bag-mask ventilation if the respiratory rate and tidal volumes are inadequate. Apply capnography if available.
• Titrate oxygen to a saturation of 92–94% while ventilating; oxygen alone masks hypoventilation and is not treatment.

Naloxone strategy:

• Start at 40–80 mcg IV increments, repeated every 2–3 minutes until respiratory rate and minute ventilation recover - not a full 0.4 mg bolus. Full reversal in a patient with a fresh knee arthroplasty and intrathecal morphine produces abrupt pain, hypertension, tachycardia, agitation, and occasionally pulmonary edema.
• The goal is ventilation, not full wakefulness.
• Because naloxone lasts 30–60 minutes and intrathecal morphine lasts many hours, start a continuous infusion at roughly two-thirds of the total arousal dose per hour, and plan for 24 hours of monitoring with continuous pulse oximetry and capnography.
• If CPAP was used at home, restart it once the airway is stable.
• Notify the surgeon, the acute pain service, and the pharmacist; re-examine the patient frequently because renarcotization is expected.

This is a medication error with patient harm, and disclosure is an ethical and professional obligation, not a discretionary courtesy.

• To the patient and family: explain in plain language what happened - the pump was programmed at ten times the ordered dose, he received more opioid than intended, he required naloxone, and he is expected to recover. Apologize. Do this as soon as he is coherent enough to participate, and again with family present.
• Internally: notify the attending surgeon, the acute pain service, pharmacy, nursing leadership, and risk management, and file a safety report the same day. Preserve the pump and the programming history.
• Expected clinical follow-up: extended monitoring for renarcotization, reassessment of the intrathecal morphine dose given his OSA, and a documented plan for CPAP and overnight oximetry.

System changes worth demanding:

• Independent double-check of PCA programming by two clinicians with a second witness signature, plus a smart-pump library that flags doses outside usual limits for hydromorphone.
• Standardized order sets that separate intrathecal from systemic opioid limits.
• Capnography in the PACU and on the ward for OSA patients receiving neuraxial opioids.
• A blameless review that asks why 2 mg was programmable, not only who typed it.