The research, led by Yan-Yan Zhu of Wenzhou Central Hospital, Affiliated to Wenzhou Medical University, together with colleagues from Taizhou Hospital of Zhejiang Province and Wenzhou Central Hospital, was published in BMC Nursing. The team set out to untangle the relationships among three constructs: patient safety culture, the organised set of values and practices that governs how hospital staff think about error; occupational exposure behaviours, the habits that determine whether nurses avoid needlestick injuries, splashes, and contact with hazardous fluids; and patient safety behaviours, the concrete actions nurses take to keep surgical patients from harm.

The study was conducted between March and June 2025 in the operating theatres of two tertiary general hospitals, facilities that sit at the top of the Chinese hospital hierarchy and handle the most complex surgical caseloads. Of 185 nurses invited to participate, 173 completed the questionnaire, a response rate of 93.5 percent that reflects the high level of engagement such surveys often achieve when delivered through institutional channels. After prespecified exclusions designed to remove incomplete or inconsistent responses, 130 questionnaires remained for analysis, corresponding to a data-retention rate of 75.1 percent. That attrition between collection and analysis is a routine but important detail in survey science, because the final analytical sample, not the initial pool, determines the statistical power and representativeness of the results.

Measurement was anchored by the Hospital Survey on Patient Safety Culture Version 2.0, a widely used instrument developed under the auspices of the Agency for Healthcare Research and Quality in the United States. HSOPSC 2.0 decomposes safety culture into composites such as hospital management support, supervisor expectations, handoffs and information exchange, and the perceived frequency of event reporting, allowing researchers to compute an overall positive response rate that summarises how favourably staff view the safety environment. For the two behavioural domains, the authors used self-developed scales assessing occupational exposure behaviours and patient safety behaviours. These instruments underwent preliminary psychometric evaluation within the sample itself, a step intended to confirm that the scales produced internally consistent and interpretable scores, although the researchers are careful to note that this evaluation was preliminary rather than a full validation program.

The headline numbers paint a moderately encouraging picture. The overall positive response rate for patient safety culture stood at 72.8 percent, meaning roughly three-quarters of survey items elicited responses favourable to safety. Mean scores were 39.8 plus or minus 5.6 for occupational exposure behaviours and 41.2 plus or minus 4.9 for patient safety behaviours, values that place the typical nurse above the midpoint of the scales but leave considerable room for improvement. In perioperative settings, where a single lapse in sharps handling or surgical site verification can cascade into serious harm, that residual gap is precisely what safety scientists worry about.

The correlational analysis revealed a consistent pattern of positive associations. Patient safety culture correlated with occupational exposure behaviours at r = 0.32 and with patient safety behaviours at r = 0.41, both statistically significant at p < 0.001. Occupational exposure behaviours, in turn, correlated with patient safety behaviours at r = 0.36, also highly significant. In practical terms, nurses working in units with stronger safety cultures reported both more careful protective practices against occupational hazards and more diligent safety behaviours directed at patients. The moderate size of these correlations, all in the range conventionally described as small to moderate, indicates real but partial overlap between organisational climate and individual conduct; culture matters, but it does not mechanically determine behaviour.

To move beyond simple bivariate associations, the researchers built a multivariable linear regression model predicting patient safety behaviours while adjusting for study centre and a set of prespecified demographic and work-related covariates. In this fully adjusted joint model, patient safety culture retained a significant association with patient safety behaviours, with a standardised coefficient of 0.29 and a 95 percent confidence interval of 0.15 to 0.43, p < 0.001. Occupational exposure behaviours also remained independently associated, with a coefficient of 0.24 and a 95 percent confidence interval of 0.08 to 0.40, p = 0.002. The persistence of both effects when entered simultaneously suggests that the two domains carry partly distinct information: a nurse’s protective habits against occupational exposures predict patient-directed safety behaviours even after accounting for the prevailing safety culture, implying that personal safety practice and organisational climate each contribute something the other does not capture.

Perhaps the most conceptually interesting result came from the exploratory indirect-association analysis. Using statistical decomposition with bias-corrected and accelerated bootstrap confidence intervals, the team estimated an indirect association between patient safety culture and patient safety behaviours operating through occupational exposure behaviours, quantified at B = 0.74 with a 95 percent confidence interval of 0.21 to 1.60. In plain language, part of the reason nurses in strong safety cultures behave more safely toward patients appears to run through their safer handling of occupational hazards. The authors frame this explicitly as exploratory: the decomposition identifies a potential pathway but does not establish intervention effectiveness, and the confidence interval, while excluding zero, is wide enough to counsel caution about the precise magnitude of the mediated component.

Several limitations deserve emphasis, and the study itself acknowledges their weight. The cross-sectional design means all variables were measured at a single point in time, so the direction of influence cannot be established; it is equally plausible that nurses who habitually practise safely come to perceive their workplace culture more favourably, or that unmeasured third factors, such as staffing levels, workload, or training quality, drive both culture scores and behaviours. The sample of 130 nurses drawn from two tertiary hospitals in one Chinese province limits generalisability to other regions, hospital tiers, and health systems. The two behavioural scales were self-developed and only preliminarily evaluated within the sample, so their properties beyond this dataset remain uncertain. Self-reported behaviour is also vulnerable to social desirability bias, particularly around safety compliance, where respondents may feel pressure to report textbook practice. The researchers compensated for collinearity concerns by checking variance inflation factors and by adjusting for centre effects, but no statistical adjustment can fully repair the inferential limits of a cross-sectional design.

Even with those caveats, the findings carry practical significance for hospital administrators and perioperative nursing leaders. If occupational exposure behaviours constitute a behavioural domain that bridges organisational safety culture and frontline patient safety practice, then safety programs that treat staff protection and patient protection as separate silos may be missing a synergy. Training modules on sharps safety, personal protective equipment use, and exposure reporting could serve double duty, reinforcing the same habits of attention, communication, and error prevention that underpin safe surgical care. Conversely, investments in culture-building, such as non-punitive event reporting and visible management commitment, may yield dividends not only in staff morale and injury reduction but in measurable patient safety behaviour. The study’s authors stop short of prescribing interventions, and rightly so, since their data show associations rather than causal effects. But they note that the results identify potential areas for further safety assessment, and that framing is a reasonable roadmap: hospitals that want to understand their surgical safety risk profile would do well to measure occupational exposure behaviours alongside conventional culture surveys, treating the two as complementary lenses on the same underlying system. As perioperative medicine grows more complex and surgical volumes rise worldwide, evidence that the guardrails protecting nurses and the guardrails protecting patients are statistically entangled offers a simple, actionable insight: in the operating theatre, a culture of safety is not divisible, and the behaviours it produces tend to travel together.

Subject of Research: Associations among patient safety culture, occupational exposure behaviours, and patient safety behaviours in operating theatre nurses

Article Title: Associations between patient safety culture, occupational exposure behaviours and patient safety behaviours among operating theatre nurses: a multicentre cross-sectional study

Article References: Zhu, Y.-Y., Ma, Y.-J., Liu, X.-M., Li, L.-Y., & Yu, S.-M. (2026). Associations between patient safety culture, occupational exposure behaviours and patient safety behaviours among operating theatre nurses: a multicentre cross-sectional study. BMC Nursing. https://doi.org/10.1186/s12912-026-05408-y

Image Credits: AI Generated

DOI: 10.1186/s12912-026-05408-y

Keywords: patient safety culture, operating theatre nurses, occupational exposure, perioperative nursing, cross-sectional study, HSOPSC 2.0, nurse safety behaviours, occupational health, hospital safety management, needlestick injury prevention, BMC Nursing, China

The research, led by Ping Gao and Xiaomei Chen of Chengdu Wenjiang District People’s Hospital along with colleagues at several Chinese institutions, addressed a well-known gap in clinical training. Nursing interns, who have just transitioned from classroom to bedside, are among the most vulnerable members of the hospital workforce. They perform injections, venipunctures, and sharps disposal under time pressure and supervision that is often inconsistent. Traditional safety training, typically delivered as lectures or static pamphlets, tends to fade quickly once the demands of real clinical work take over. Although animation-based education has proven effective in other areas of clinical teaching, its potential for preventing needlestick injuries had been largely unexplored until now.

The intervention at the heart of the study was what the researchers call caricature storyline-based audiovisual education. Rather than presenting isolated facts about sharps safety, the educational materials wove preventive principles into a narrative told through exaggerated, cartoon-like animated images, which the authors note were generated with the assistance of the AI tool Doubao. The storyline format is more than an aesthetic choice. Educational psychology suggests that narrative structure helps learners encode and retrieve procedural information, because characters, conflicts, and resolutions give abstract rules a concrete emotional and situational anchor. For a novice nurse, watching a caricatured character fumble a recapping maneuver and face the consequences may be more memorable than memorizing a checklist.

To test whether the approach actually worked, the team used a sequential mixed-methods design, combining rigorous quantitative measurement with in-depth qualitative exploration. They recruited 260 nursing interns enrolled in junior college or undergraduate nursing programs, all of whom had completed one to two months of clinical practice before entering the study. This timing was deliberate: the participants had enough ward experience to understand the risks firsthand, but were still early enough in their training for education to shape lasting habits. Every participant received the caricature-based audiovisual education, and the researchers measured outcomes at three points: before the intervention, immediately afterward, and again three months later.

The quantitative results were striking. Using validated scales, the team assessed four dimensions: knowledge about needlestick injuries, attitudes toward prevention, actual safety behaviors, and preventive motivation. Repeated-measures analysis of variance revealed significant main effects of time across all four outcomes, with p values below 0.001 and partial eta-squared values ranging from 0.968 to 0.992, effect sizes so large they are rarely seen in behavioral research. Corrected post-hoc comparisons confirmed that every score rose significantly from baseline to both the post-test and the three-month follow-up. In practical terms, the interns did not merely perform better on a quiz immediately after watching the videos; they internalized the material in a way that persisted.

The follow-up data revealed an especially interesting pattern. Knowledge and attitude scores plateaued between the post-test and the three-month mark, remaining statistically stable with p values of 0.473 and 0.541 respectively. Behavior and preventive motivation, however, continued to climb significantly during the same period, both with p values below 0.001. This suggests that the intervention did more than transmit information. It appears to have triggered a self-reinforcing cycle in which motivated interns kept practicing safer behaviors, and those behaviors in turn deepened their commitment to prevention. The most concrete outcome measure was perhaps the simplest: only six needlestick injuries were reported across the entire group during the follow-up period.

Numbers alone cannot explain why the intervention worked, which is where the qualitative arm of the study proved valuable. A subsample of 49 participants took part in ethnographic observations and semi-structured interviews, allowing the researchers to see how the education translated, or failed to translate, into daily ward practice. Thematic analysis of this material yielded four themes and six subthemes, capturing how the interns perceived the intervention and how it reshaped their thinking about sharps safety. The researchers report that most nursing interns rated the intervention positively, describing it as engaging and relevant to their real clinical situations, and suggesting that the format could be extended to other healthcare disciplines beyond nursing.

The study’s methodological choices strengthen its credibility. By requiring participants to have completed at least a month of clinical practice, the researchers ensured that baseline knowledge and attitudes reflected genuine ward experience rather than classroom theory. The three-time-point measurement design allowed them to distinguish immediate learning effects from durable behavioral change, a distinction that many training evaluations miss. The combination of repeated-measures ANOVA with corrected post-hoc testing guards against the inflated error rates that can plague before-and-after studies, and the very large effect sizes, while partly a reflection of a single-group design without a control arm, indicate that the intervention moved every measured dial.

That single-group design is also the study’s main limitation, and the authors and readers alike should interpret the findings with appropriate caution. Without a randomized control group receiving standard lecture-based training, it is difficult to isolate the specific contribution of the caricature storyline format from the general benefit of receiving any structured, engaging education on needlestick prevention. Self-reported behavior scales can also diverge from observed practice, although the ethnographic observations in the qualitative arm partially offset this concern. Still, the consistency of improvement across four distinct outcome dimensions, and the continued rise in behavior and motivation over three months, make a compelling case that the intervention produced real change rather than a fleeting testing effect.

The implications reach well beyond one hospital in Chengdu. Needlestick injuries impose enormous costs on health systems, from post-exposure prophylaxis and testing to the psychological toll of waiting weeks for serology results. If a low-cost, scalable package of animated videos can meaningfully reduce these injuries among the newest and most vulnerable members of the clinical workforce, hospital educators have a practical tool they can deploy widely. The authors suggest the model may support the standardization of clinical practice and help reduce the occurrence of needlestick injuries more broadly. As hospitals worldwide grapple with how to instill safety culture in a constantly rotating stream of trainees, this study offers a reminder that sometimes the most effective teacher is a well-told story, even one drawn in exaggerated cartoon lines.

Subject of Research: Caricature storyline-based audiovisual education for preventing needlestick injuries among nursing interns

Article Title: Effects of caricature storyline audiovisual education on knowledge, attitudes, behavior, and preventive motivation regarding needlestick injuries among nursing interns: a mixed methods study

Article References: Gao, P., Cheng, L., Wang, Z., Zeng, X., Xu, X., Huang, J., Li, L., Li, W., Tang, M., Wang, Q., & Chen, X. (2026). Effects of caricature storyline audiovisual education on knowledge, attitudes, behavior, and preventive motivation regarding needlestick injuries among nursing interns: a mixed methods study. BMC Nursing. https://doi.org/10.1186/s12912-026-05373-6

Image Credits: AI Generated

DOI: 10.1186/s12912-026-05373-6

Keywords: needlestick injuries, nursing interns, audiovisual education, caricature animation, nursing education, occupational safety, mixed methods study, preventive motivation, clinical training, BMC Nursing, healthcare workers, sharps safety

Protection motivation theory, first articulated by Ronald Rogers in the 1970s and refined in the decades since, proposes that protective behavior is the product of two parallel appraisal processes. The first is threat appraisal, in which an individual judges how susceptible they are to a hazard, how severe its consequences would be, and how much fear the hazard arouses. The second is coping appraisal, in which the individual evaluates whether a recommended response would actually work (response efficacy), what that response would cost them in time, money, or discomfort (response costs), and whether they are personally capable of carrying it out (self-efficacy). When the perceived threat outweighs the perceived barriers, protective motivation rises; when barriers loom larger, motivation collapses. In the context of a needlestick injury, this translates into a stark set of questions: How likely is the source patient to carry hepatitis B, hepatitis C, or HIV? How severe would infection be? Would screening and follow-up actually reduce harm, and can the nurse realistically navigate the hospital’s reporting bureaucracy to complete it?

The scale’s construction followed a rigorous multi-stage pipeline. The initial item pool was generated through a systematic literature review and semi-structured interviews with clinical nurses, ensuring that the questionnaire reflected the lived experience of frontline staff rather than armchair theorizing. The pool was then refined through two rounds of Delphi expert consultation, a structured method in which a panel of subject-matter experts independently rates and comments on draft items across successive rounds until consensus emerges. This approach is particularly valued in instrument development because it filters out ambiguous, redundant, or theoretically misaligned items before expensive data collection begins. The result was a candidate questionnaire ready for empirical testing in the field.

For the psychometric evaluation, Long recruited 431 clinical nurses using purposive sampling, a technique that deliberately selects participants with direct relevance to the research question—in this case, nurses with clinical exposure to needlestick injury risk. The full sample was used for classical item analysis, examining metrics such as item discrimination and floor and ceiling effects. The sample was then randomly split into two halves: subsample A, comprising 215 nurses, for exploratory factor analysis, and subsample B, comprising 216 nurses, for confirmatory factor analysis. This split-sample strategy is considered best practice in scale validation because it prevents the same data from being used both to discover a factor structure and to confirm it, which would inflate apparent fit. Separately, 77 nurses completed the scale a second time after two weeks, allowing the researchers to compute test–retest reliability. Content validity was assessed using content validity indices, in which expert raters judge the relevance of each item to the underlying construct.

The exploratory factor analysis on subsample A identified six common factors corresponding exactly to the theory-driven dimensions: perceived susceptibility, perceived severity, fear arousal, response efficacy, response costs, and self-efficacy. Together these factors explained 75.176 percent of the total variance, an unusually high figure that suggests the six dimensions capture the bulk of the meaningful variation in nurses’ protection motivation. Confirmatory factor analysis on subsample B produced mixed but broadly supportive fit indices: the comparative fit index and incremental fit index both reached 0.903, hovering at the conventional threshold of 0.90 for acceptable fit, while the root mean square residual of 0.038 and standardized root mean square residual of 0.063 fell comfortably within accepted limits. The root mean square error of approximation, however, registered 0.105, above the commonly cited 0.08 threshold, indicating that while the six-factor model is a reasonable approximation of the data, some local misfit remains—likely reflecting overlapping variance between conceptually adjacent dimensions.

Reliability figures were strong across the board. Cronbach’s alpha for the overall 24-item scale reached 0.955, indicating exceptional internal consistency, and composite reliability for the individual dimensions ranged from 0.851 to 0.959, all well above the 0.70 benchmark. The average variance extracted, a measure of how much of a construct’s variance is explained by its own indicators relative to measurement error, ranged from 0.590 to 0.855, exceeding the conventional 0.50 threshold for all six dimensions. Split-half reliability, computed using both the Spearman–Brown coefficient (0.771) and Guttman’s coefficient (0.736), was moderate to good, and the two-week test–retest coefficient of 0.936 demonstrated that the scale yields highly stable scores over time—an essential property for an instrument intended to track motivation before and after interventions. Content validity was equally impressive, with item-level content validity indices ranging from 0.93 to 1.00 and a scale-level index of 0.97 based on the averaging method.

Discriminant validity, the degree to which the six dimensions are empirically distinguishable from one another, was tested using the Fornell–Larcker criterion, which requires that the square root of each construct’s average variance extracted exceed the correlations between that construct and all others. The criterion was satisfied for most construct pairs, with one notable exception: response efficacy and response costs sat at the boundary of the criterion, meaning the data could not cleanly separate the perceived benefits of screening from the perceived burdens of undergoing it. This finding is theoretically interesting rather than merely statistical. It suggests that for many nurses, the calculus of whether screening “works” and whether screening is “worth it” are psychologically intertwined—perhaps because the efficacy of screening is judged largely in terms of whether its inconvenience and expense justify the protection it affords. The author explicitly recommends that the stability of the factor structure and the distinction between these two constructs be re-examined in independent samples.

The practical implications of the NMSP-NSI extend well beyond the psychometrics laboratory. Needlestick injuries expose nurses to bloodborne pathogens including hepatitis B virus, hepatitis C virus, and HIV, and post-exposure protocols—immediate washing, reporting, source-patient testing, and serial screening of the injured worker—are only effective if nurses actually complete them. Underreporting and incomplete follow-up are chronic problems in hospitals worldwide, driven by factors the new scale is designed to capture: underestimation of susceptibility, minimization of severity, fear of stigma or job repercussions, uncertainty about whether screening helps, and perceived logistical burdens. By pinpointing which of the six motivational dimensions is weakest in a given nurse or department, hospital administrators and occupational health teams can tailor education and process redesign precisely where they are needed, rather than delivering generic safety lectures that may miss the actual psychological bottleneck. A nurse who already understands the severity of HIV but doubts her self-efficacy in navigating the reporting system needs a very different intervention from one who simply does not perceive the injury as a threat.

The study was conducted entirely online, with participants receiving key information about the purpose, content, and confidentiality of the research before indicating consent—an approach approved by the Research Ethics Committee of Wuming Hospital of Guangxi Medical University under decision number WM-2025(11) and conducted in accordance with the Declaration of Helsinki. The work was funded by the 2023 Guangxi University Young and Middle-aged Teachers’ Scientific Research Basic Ability Enhancement Project (no. 2023KY0077) and the 2025 Self-funded Scientific Research Project of the Health Commission of the Guangxi Zhuang Autonomous Region (no. Z-A20250654). The author acknowledges the contributions of the participating nurses, the Delphi expert panel, and staff at the eight hospitals that assisted with data collection.

As an open-access publication released under a Creative Commons Attribution-NonCommercial-NoDerivatives license, the article makes its instruments and findings freely available to occupational health researchers globally, and the author notes that the version shared is a citable, DOI-bearing early-release that will be superseded by the final Version of Record. Whether the NMSP-NSI will prove equally valid in healthcare systems outside China, among other at-risk professions such as physicians, laboratory technicians, and sanitation workers, and whether motivational scores actually predict real-world screening behavior over time, are questions for future research. But the study marks a meaningful step toward treating post-exposure screening not merely as a protocol to be enforced, but as a psychological decision to be understood—and, with the right tools, nudged in the direction of safety.