Underreporting of blood and body fluid exposures among nurses: prevalence, reporting cascade, and modifiable...

Occupational exposure to blood and body fluids is common among nurses, and incomplete reporting limits occupational surveillance and access to post-exposure assessment. This study estimated underreporting at the nurse level and the proportion of aggregate exposure counts that were unreported, and examined associated individual and workplace characteristics.

A census-based analytical cross-sectional survey was conducted between November 2025 and March 2026 among registered nurses at a tertiary infectious disease referral hospital in China, using a structured questionnaire with a 12-month recall period. Underreporting was defined among exposed nurses as failure to formally report at least one exposure. Modified Poisson regression with robust variance was used to estimate adjusted prevalence ratios (aPRs). Negative-binomial and hurdle models with an exposure offset served as sensitivity analyses for exposure opportunity.

Of 1,110 eligible nurses, 872 responded (78.6%); 392 (45.0%) reported at least one exposure, contributing 747 aggregate exposure counts. Underreporting affected 232 exposed nurses (59.2, 95% CI 54.3-63.9), and 373 of 747 exposure counts (49.9%) were unreported. Among under-reporters, 48.3% reported receiving none of the five assessed services after any exposure during the recall period, compared with 0.6% of complete reporters; clinical indications and exposure-specific service receipt were unavailable. Underreporting was associated with individual perceptions of safety climate (aPR per SD 0.77, 95% CI 0.70-0.84), vague or absent procedural knowledge (aPR 1.56, 95% CI 1.31-1.86), non-permanent employment (aPR 1.30, 95% CI 1.07-1.57), and night-shift frequency (aPR per level 1.14, 95% CI 1.04-1.25).

Underreporting was common at both the nurse level and across aggregate exposure counts. Perceived safety climate, procedural knowledge, employment security, and night-shift work were associated with underreporting and are potential targets for intervention research; the cross-sectional design does not establish causation or event-specific clinical consequences.