A nebulised tranexamic acid protocol cut returns to theatre for post-tonsillectomy bleeding by two-thirds

Design

Quality improvement project with two Plan-Do-Study-Act cycles, single institution, 2020-2025

Population

501 children under 18 presenting with post-tonsillectomy haemorrhage, median age 7.2 years

Primary outcome

Need for surgical control of bleeding

Effect

43.9 per cent pre-intervention, 29.8 per cent with unprotocolised tranexamic acid, 15.2 per cent after the standardised protocol (P < 0.0001)

Five hundred and one children under 18 presenting with post-tonsillectomy haemorrhage were followed across three periods at one institution: a pre-intervention baseline from January 2020 to June 2022 (171 patients), a first improvement cycle in which nebulised tranexamic acid was given without a protocol (205 patients), and a second in which a standardised protocol was implemented (125 patients). Median age was 7.2 years.

The proportion needing surgical control of bleeding fell at each step: 43.9 per cent, then 29.8 per cent, then 15.2 per cent (P < 0.0001) - a 65.4 per cent relative reduction. The balancing measures did not move: no change in length of stay, transfusion rates, readmissions or tranexamic acid-related complications.

This is a quality improvement project, not a randomised trial, and that limits it in one specific way: there is no concurrent control, so a secular change in practice over five and a half years could contribute. Against that, the stepwise fall aligned with each intervention, the effect is large, and the intervention is cheap, available and already familiar. For a unit taking children back to theatre at night for tonsillar bleeds, the case for writing a protocol - dose, nebuliser, who gives it, when to escalate - is strong, and the absence of any signal in the balancing measures is what makes it reasonable to act on before a trial exists.

• Write the protocol rather than relying on the drug being available; the unprotocolised cycle achieved only half the effect.
• Specify the dose, the delivery device, who administers it and the escalation threshold.
• Keep the theatre pathway ready - this reduced returns to theatre, it did not remove them.
• Audit transfusions, readmissions and length of stay alongside, as this project did.
• Remember the design: a stepwise quality improvement project with no concurrent control.

The statistics, in plain English

The stepwise fall across three sequential periods is more persuasive than a simple before-and-after, because the intermediate cycle - drug available but no protocol - landed between the two, which is what you would expect if the protocol is doing the work. But sequential periods cannot exclude a background trend, and the groups are not randomised or matched. The unchanged balancing measures are reassuring about safety at this sample size; 501 patients cannot exclude a rare harm.

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.