Release the tourniquet as the first tube fills, not after the last

Design

Single-centre, parallel-group randomised trial

Population

790 adult emergency patients (ESI 3-5) with direct 21-gauge venepuncture

Primary outcome

Proportion of gel tubes with haemolysis index 1+ or more

Effect

17.2% vs 26.1%; absolute reduction 8.9 points (95% CI 3.1-14.5)

Haemolysis is the commonest pre-analytical error in emergency department samples. This single-centre randomised trial enrolled 792 adults triaged as non-critical (ESI 3-5) and drew blood by direct 21-gauge venepuncture with a cuff inflated to 60 mm Hg.

Cuff release was either started as blood entered the first tube and completed as it filled, or left until the last tube was filled. Haemolysis (index 1+ or more on the gel tube, read blind) occurred in 17.2% with early release and 26.1% with late release: an absolute reduction of 8.9 percentage points (95% CI 3.1-14.5), about one haemolysed sample prevented for every 11 patients. No early-release patient had more than 60 seconds of stasis, against 29% with late release. Complication estimates were imprecise.

The finding is limited to direct venepuncture with a standard cuff pressure. It does not apply to cannula draws, winged sets or elastic tourniquets, or to the sickest patients.

This is a cheap, teachable change for phlebotomy protocols that the laboratory can audit through its haemolysis index.

• Start releasing the tourniquet as blood enters the first tube.
• Keep total stasis under about 60 seconds.
• Early release cut haemolysis from 26.1% to 17.2% in the trial.
• The evidence covers direct 21-gauge venepuncture, not cannula or winged-set draws.
• Audit haemolysis index rates before and after a protocol change.

Why it matters

A free change in technique removes a large share of repeat samples and spurious results the laboratory otherwise has to flag.

Don't overread it

The trial excluded cannula draws and critically ill patients, where much emergency haemolysis arises.

The statistics, in plain English

An absolute reduction of 8.9 points means about 9 fewer haemolysed samples per 100 patients; the interval (3.1-14.5) excludes no effect. A number needed to treat of 11 means changing technique for 11 patients prevents one haemolysed sample.

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.