Do protective devices prevent needlestick injuries among health care workers?

Objectives: To determine the effectiveness and direct cost of two protective devices - a shielded 3 ml safely syringe (Safety-Lok; Becton Dickinson and Co., Becton Dickinson Division, Franklin Lakes, N.J.) and the components of a needleless IV system (Interlink; Baxter Healthcare Corp., Deerfield, ill.)- in preventing needlestick injuries to health care workers. Design: Twelve-month prospective, controlled, before-and-after trial with a standardized questionnaire to monitor needlestick injury rates. Setting: Six hospital inpatient units, consisting of three medical units, two surgical units (all of which were similar in patient census, acuity, and frequency of needlesticks), and a surgical-trauma intensive care unit, at a 900-bed urban university medical center. Participants: All nursing personnel, including registered nurses, licensed practical nurses, nursing aides, and students, as well as medical teams consisting of an attending physician, resident physician, interns, and medical students on the study units. Intervention: After a 6-month prospective surveillance period, the protective devices were randomly introduced to four of the chosen study units and to the surgical-trauma intensive care unit. Results: Forty-seven needlesticks were reported throughout the entire study period, 33 in the 6 months before and 14 in the 6 months after the introduction of the protective devices. Nursing staff members who were using hollow-bore needles and manipulating intravenous lines accounted for the greatest number of needlestick injuries in the pre-intervention period. The overall rate of needlestick injury was reduced by 61%, from 0.785 to 0.303 needlestick injuries per 1000 health care worker-days after the introduction of the protective devices (relative risk = 1.958; 95% confidence interval, 1.012 to 3.790; p = 0.046). Needlestick injury rates associated with intravenous line manipulation, procedures with 3 mi syringes, and sharps disposal were reduced by 50%; however, reductions in these subcategories were not statistically significant. No seroconversions to HIV-1 or hepatitis B virus seropositivity occurred among those with needlestick injuries. The direct cost for each needlestick prevented was $789. Conclusions: Despite an overall reduction in needlestick injury rates, no statistically significant reductions could be directly attributed to the protective devices. These devices are associated with a significant increase in cost compared with conventional devices. Further studies must be concurrently controlled to establish the effectiveness of these devices.