Bolus-only hydromorphone PCA was not shown to match bolus plus infusion

Design

Multicentre, randomised, non-inferiority trial

Population

226 patients with refractory cancer pain on IV hydromorphone PCA, China

Primary outcome

≥50% reduction in resting pain at 72 hours

Effect

89.4% vs 94.2%; 90% CI -11.2 to 1.1 (non-inferiority not shown)

This Chinese multicentre randomised non-inferiority trial enrolled 226 patients with refractory cancer pain starting intravenous hydromorphone patient-controlled analgesia (PCA). They received either bolus doses only or boluses plus a continuous background infusion. Published in Supportive Care in Cancer on 22 October 2026.

At 72 hours, at least a 50% reduction in resting pain was reached by 89.4% with bolus-only and 94.2% with bolus plus infusion. The 90% confidence interval for the difference (-11.2 to 1.1 points) crossed the -10% margin, so bolus-only dosing was not shown to be non-inferior. Both groups improved substantially, and PCA-related adverse events were similar.

In postoperative pain, background infusions are often avoided because they add sedation without improving relief. Refractory cancer pain is different: patients are often opioid-tolerant and have continuous pain, and this trial found no safety penalty for the infusion.

This is one trial from one country with a modest difference, but it supports keeping a background dose when PCA is used for refractory cancer pain.

• Consider a background infusion when starting IV PCA for refractory cancer pain
• Base the background rate on the patient's previous 24-hour opioid requirement
• Monitor sedation and respiratory rate, especially in the first 24 hours
• Review bolus use daily and adjust the background rate accordingly
• Do not carry over the postoperative habit of bolus-only PCA to cancer pain without thought

Why it matters

It challenges carrying postoperative PCA habits over to patients with continuous cancer pain.

Don't overread it

A failed non-inferiority test does not prove bolus-only dosing is worse; the difference was about 5 points.

The statistics, in plain English

Non-inferiority needed the lower end of the confidence interval to stay above -10 points. It reached -11.2, just past the margin, so the trial could not rule out that bolus-only dosing was meaningfully less effective.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.