To compare the percutaneous injury rate associated with a standard versus a safety resheathable winged steel (butterfly) needle.

Before-after trial of winged steel needle injuries during a 33-month period (19-month baseline, 3-month training, and 11-month study intervention), followed by a 31-month poststudy period.

A 1,190-bed acute care referral hospital with inpatient and outpatient services in New York City.

All healthcare workers performing intravascular-access procedures with winged steel needles.

Safety resheathable winged steel needle.

The injury rate associated with winged steel needles declined from 13.41 to 6.41 per 100,000 (relative risk [RR], 0.48; 95% confidence interval [CI95], 0.31 to 0.73) following implementation of the safety device. Injuries occurring during or after disposal were reduced most substantially (RR 0.15; CI95, 0.06 to 0.43 Safety winged steel needle injuries occurred most often before activation of the safety mechanism was appropriate (39%); 32% were due to the user choosing not to activate the device, 21% occurred during activation, and 4% were due to improper activation. Preference for the safety winged steel needle over the standard device was 63%. The safety feature was activated in 83% of the samples examined during audits of disposal containers. Following completion of the study, the safety winged steel needle injury rate (7.29 per 100,000) did not differ significantly from the winged steel needle injury rate during the study period.

Implementation of a safety resheathable winged steel needle substantially reduced injuries among healthcare workers performing vascular-access procedures. The residual risk of injury associated with this device can be reduced further with increased compliance with proper activation procedures.