ABSTRACT
BACKGROUND: Most studies on the incidence rate (IR) and post-exposure reporting rate (RR) of needle-stick injuries (NSIs) were performed using retrospective surveillance, which is vulnerable to recall bias. This study aimed to identify the agreement between IRs and RRs obtained from prospective and retrospective surveillance. METHODS: The prospective surveillance was performed with 716 nurses working at 3 hospitals from August to September in 2012. They prospectively reported when they experienced the NSIs, and the investigator retrospectively calculated the RR from records in the infection control unit or health care unit during the same periods when they reported the number of NSIs. The retrospective surveillance was carried out with 312 nurses who participated in the prospective surveillance. They retrospectively answered the question on the number of NSIs and post-exposure reporting after recalling the experienced NSI from August to September in 2012. RESULTS: The IR of NSIs was 9.8 per 100 nurses by the prospective surveillance and 36.4 per 100 nurses by the retrospective surveillance, which was statistically significantly different (P<0.001). The RR of NSIs was 14.3% by the prospective surveillance and 8.5% by the retrospective surveillance, which was not statistically significantly different. CONCLUSION: We recommend using a prospective approach for calculating the IR of NSIs to reduce the risk of recall bias. However, the RR of NSIs can be calculated using both prospective and retrospective approaches.