Accidental exposure to blood (AEB) poses a potential risk for bloodborne pathogen infections in healthcare workers (HCWs) and subsequently in patients during care. The most common pathogens transmitted through AEB include human immunodeficiency virus (HIV), hepatitis B and C, and certain bacteria. Reference Tarantola, Abiteboul and Rachline1 AEBs can occur in various ways, such as percutaneous needlestick injuries, cuts, scrapes, or splashes to the mucous membranes like the eyes, nose, or mouth. Needlestick injuries are the most common cause of AEB during care tasks. Reference Mengistu, Tolera and Demmu2 High-risk care tasks for AEBs typically involve handling sharp objects, administering medications, managing bodily fluids, and working with contaminated equipment. However, AEB can also occur during other nursing tasks and environmental activities, such as cleaning contaminated surfaces or disposing of hazardous waste. To mitigate AEBs, it is vital to adhere to standard precautions and to utilize personal protective equipment, including gloves, masks, and eye protection, during high-risk care. Proper disposal of needles and other sharp objects is also essential. Reference Hinkin, Gammon and Cutter3 AEB serves as a marker for risk, quality, and safety of care among HCWs because the accident rate might reflect the performance of healthcare facilities in terms of cost and economic impact. Reference Mannocci, De Carli and Di Bari4 AEB risk increases during nursing care or environmental activities, particularly for HCWs regularly exposed to invasive procedures with significant blood contact. Reference Adams, Stojkovic and Leveson5 Activities like wound care, equipment handling, and waste disposal can raise the likelihood of AEB, which contributes to stress and has physical and psychological consequences for HCWs. However, understanding the impact of various healthcare-related factors on AEB risk remains challenging. Although behavioral and management factors influence the risk of healthcare-associated infections in patients, Reference Zingg, Holmes and Dettenkofer6 research identifying these factors for HCWs has been limited. Studies highlight the importance of teamwork, leadership, and proper equipment, Reference Farokhzadian, Dehghan Nayeri and Borhani7 as well as the need for a strong safety culture among HCWs. Reference Gershon, Karkashian and Grosch8 Considering this evidence, a French cross-sectional study Reference Jones, Hocine and Salomon9 examined the impact of various factors on HCW stress and fatigue in intensive care units (ICUs). These researchers found that employment and organizational factors were significantly associated with stress and fatigue outcomes, even after controlling for demographic factors. Addressing factors at both individual and organizational levels is crucial for improving HCWs health. With this background in mind, we sought to identify both individual behavioral and organizational factors that could influence the occurrence of AEB in HCWs and to better elucidate the role of various potential risk factors.
Methods
Study design and participants
The STRIPPS study (no. NCT03532321) was a 1-year follow-up multicenter, prospective study investigating the individual and organizational factors that predict occupational exposure to blood among 730 HCWs in Paris university hospitals. 10 The study was carried out in 4 general-care hospitals between February 2018 and July 2019 and included nurses, nursing assistants, midwives, and physicians as participants. We included both permanent and fixed-term contract HCWs with work contracts lasting at least 1 year, matching the survey duration. We excluded those with contracts <1 year, as well as external personnel (eg, nursing float pool). The sample size was a convenience sample based on a previous study. Reference Jones, Hocine and Salomon9 Data were collected from these HCWs by randomly selecting 8 wards per participating hospital from those that employed at least 30 HCWs.
Data collection
Data were collected in all participating individuals, every 4 months for a total follow-up period of 1 year, by 2 different interviewers. The collection times were designated as t0, t1, t2, and t3, corresponding to the first collection during the inclusion visit and follow-up visits at 4, 8, and 12 months, respectively. For the first collection (t0), dates and times of visits were randomly assigned for each ward. For subsequent collections, individual appointments were made considering different work shifts (day and night) to ensure a comprehensive representation of HCW schedules.
Data were collected through questionnaire-based interviews at both the ward and individual levels. Potential participants were informed of the study through an information letter and gave verbal consent at the beginning of each interview. Participants were guaranteed confidentiality and anonymity of responses.
Ward-level variables
The hospital health executives (nurse managers) were interviewed at t0 to collect data for each of the 32 wards. The data collected pertained to the medical specialty of the ward, the number of beds per ward, the proportion of double rooms, the frequency of tasks performed outside the ward, the ratio of HCW to patients, and the use of external staff. Further inclusion and exclusion criteria were previously reported. Reference Daouda, Bun and Ait Bouziad11
Individual-level variables
Interviewers collected a range of information about the HCWs including demographics and details about their work organization. Validated questionnaires, such as the Effort-Reward Imbalance (ERI), Reference Siegrist, Starke and Chandola12 the Job Content Questionnaire (JCQ), Reference Karasek, Brisson and Kawakami13 the Perceived Stress Scale-10 (PSS-10), Reference Cohen, Kamarck and Mermelstein14,Reference Lesage, Berjot and Deschamps15 and the Pichot fatigue scale, Reference Pichot and Brun16 were used to standardize measures of overcommitment, social support, and stress and fatigue levels, respectively. This information was collected at 4 different times (ie, t0, t1, t2, and t3) to track changes over time.
Outcome
The outcome variable was the self-reported occupational exposure to blood among HCWs. An AEB is defined as any unintended contact with blood or blood-containing body fluids, which can occur through percutaneous injuries, cuts, scrapes, or splashes to the skin or mucous membranes. At each visit, HCWs were asked about the number of AEBs they had experienced within the previous 4 months. Only declared AEBs were considered. The accidents were further described in terms of their context, cause, and nature of injury. Notably, this information was obtained through self-reporting by the HCWs during the study visits.
Missing data
To handle missing data, we utilized multiple imputation on validated questionnaire items (JCQ, PSS-10, Pichot, and ERI questionnaires) with the R mice package. 17 The imputation was performed on both continuous and categorical variables in longitudinal data. Missing data for all questionnaire items in the imputation model were assumed to be missing at random.
Statistical analyses
First, we conducted a descriptive analysis to summarize the data collected at the individual and ward levels. We assessed changes in individual-level variables over time using 2-sided student tests for continuous variables and χ Reference Mengistu, Tolera and Demmu2 tests for categorical variables. Next, we identified factors associated with AEBs in participating HCWs. Bivariate analyses were conducted on all individual-level variables to determine which variables were relevant for inclusion in the multivariate analysis. Variables with P ≤ .20 were considered for inclusion in the model. Based on these results, we performed a multivariate analysis with longitudinal data to investigate the association between risk factors and the outcome variable. We used a linear mixed-effects model, including the hospital as a random effect to account for the unobserved heterogeneity across hospitals. We used the hospital variable as the random effect because it represents a higher-level grouping in the data hierarchy and demonstrated a significant effect in the bivariate analysis. To select the best model, we used the Akaike information criterion (AIC) and compared it with alternative models to ensure the inclusion of the most suitable random and fixed effects. All data analyses were conducted using R package lme4 software (R Foundation for Statistical Computing, Vienna, Austria). Reference Bates, Maechler and Bolker18
Ethical approval
The study protocol obtained both an agreement from the French Committee for the Protection of Persons (CPP) on November 14, 2017, and clearance from the French Data Protection Authority (CNIL) on December 14, 2017 (IDRCB no. 2017-A02939-44).
Results
Demographic and work characteristics of the study sample
In this study, a sample of 730 HCWs was analyzed. The sample comprised 384 nurses (52.6%), 300 nursing assistants (41.1%), 35 physicians (4.8%), and 11 midwives (1.5%). The female:male sex ratio was 5:1, with 610 female respondents (83.6%). The majority of HCWs were permanent staff (n = 644, 88.2%) compared to temporary staff (n = 66, 9.0%) and contractual staff (n = 19, 2.6%). The average number of years of experience was 10.5 (±9.7), and 380 respondents (52.1%) had supervising responsibilities. On average, HCWs worked 37.6 hours per week (±5.8 hours), and 614 (84.1%) had advance knowledge of their schedule, but 322 (44.1%) had never participated in creating it. Furthermore, 302 (41.4%) staff did not take their rest immediately after night shifts. In terms of transportation, 365 participants (50.0%) reported a daily car use versus 303 (41.5%) using public transportation and 62 (8.5%) using other options (ie, walking, biking, or motor biking). The daily commute duration was <1 hour for 306 participants (41.9%), between 1 and 2 hours for 321 participants (44.0%), and >2 hours for 103 participants (14.1%).
Characteristics of participating wards in the study
This study included 32 wards from various medical fields, including surgery and obstetrics (14 wards, 43.8%), medicine (11 wards, 34.4%), and ICUs (7 wards, 21.9%). The average number of beds per ward was 35.5 (±18.5), and ∼20% of ward rooms were double rooms. The patient-to-physician ratio and patient-to-paramedic ratio were 2.9 and 0.8, respectively. The scheduling of work varied across participating wards. Most HCWs (80%) organized work in three 8-hour shifts, whereas 16% of wards used two 12-hour shifts. More than 80% of wards required HCWs to leave the ward on occasion, and most wards utilized interim staff.
Details of accidental exposures to blood
In total, 108 instances of occupational blood exposure were reported among 71 HCWs. Table 1 provides details about AEBs, including the nature of the injury, the mechanism of occurrence, and the task being performed when the accident occurred, grouped by medical specialty and occupation. Of the 108 reported blood accidents, 52 occurred among 29 HCWs in the ICU, 40 among 37 HCWs in surgery and obstetrics, and the remaining 16 among 5 HCWs in other medical specialties. The incidents included 59 splashes, 44 needlestick injuries, and 5 cuts from sharp objects. The main reasons were handling a mounted needle (57%), followed by handling contaminated instruments (17%) and other mechanisms (13%). Most injuries occurred during tasks such as blood sampling (41%), infusion (12%), surgery (12%), nursing and hygiene (11%), and other care (9%). Surgeons and midwives, who carry out procedures involving skin punctures or cuts in surgery and obstetrics departments, had higher rates of occupational blood exposure, at 37.5% and 20%, respectively. ICU nurses had the second-highest rate of blood exposure, at 10.8%.
Incidence Rate of Accidental Exposure to Blood per 1,000 Person Years by Medical Specialty and by Occupation (n=108)
a Indicence per 1,000 person years; the number of accidents is specified in brackets, unless otherwise indicated.
Note. Other medical specialties: cardiology, geriatrics, gastroenterology, infectious diseases, internal medicine, nephrology, oncology, pulmonology, rheumatology, urology.
Bivariate and multivariate analyses on predictors of AEBs among HCWs
Table 2 shows the results of bivariate analyses conducted on variables associated with AEBs among HCWs. These analyses helped identify potential individual-level predictors that might be associated with the occurrence of AEBs in HCWs. Table 3 presents the multivariate model selected using Akaike information criterion (AIC), which shows significant association between the occurrence of AEBs and various individual-level predictors. These predictors included younger age (relative risk [RR], 4.25; 95% confidence interval [CI], 1.20–9.94; P = .026), occupation as nurses (RR, 2.43; 95% CI, 1.25–4.52; P = .009) or midwives (RR, 2.90; 95% CI, 1.32–4.45; P = .012), irregular work schedules (RR, 3.18; 95% CI, 1.83–5.11; P < .001), rotating shifts (RR, 3.11; 95% CI, 1.72–4.83; P = .001), and lack of support from supervisors (RR, 1.16; 95% CI, 1.06–1.28; P = .001). No significant variation over time was observed. Additionally, the use of external nurses was a significant ward-level predictor associated with the occurrence of AEBs (RR, 2.02; 95% CI, 1.19–3.35; P = .010).
General and Organizational Characteristics of HCWs on Occupational Exposure to Blood Using Bivariate Analysis (Logistic Regression)
Note. AEB, accidental exposure to blood; HCW, healthcare worker; ICU, intensive care unit; ND, No Data Available; SD, standard deviation.
a No. (%) unless otherwise indicated.
b Other medical specialties: cardiology, geriatrics, gastroenterology, infectious diseases, internal medicine, nephrology, oncology, pulmonology, rheumatology, urology.
* Indicates statistical significance.
Multivariate Analysis of Accidental Exposure to Blood Using a Stepwise Linear Mixed-Effects Model Selection Based on Akaike Information Criterion
Note. RR, relative risk; CI, confidence interval.
a Refers to the perceived support from supervisors as assessed by the Karazek questionnaire. This subscale measure the extent to which HCWs perceive their supervisors to be unsupportive or indifferent to their needs and concerns. A higher score indicates a greater perceived lack of support from supervisors, which has been identified as a potential risk factor for occupational blood exposure.
Discussion
The main findings of this longitudinal study highlight the importance of considering both individual and organizational factors when addressing AEBs among HCWs. We identified several significant factors, including occupation, age, work schedule consistency, rotating shifts, social support from supervisors, and the frequent use of external nurses. However, there was no evidence of relationship between stress and fatigue and the occurrence of AEBs.
The study revealed that physicians, nurses, and midwives, who have more frequent and direct contact with patients, are more likely to be exposed to AEBs. Reference Wicker, Jung and Allwinn19 The increased risk among these occupations could be due to the nature of their work, involving invasive procedures, handling of sharp instruments, and frequent patient interactions. Reference Adams, Stojkovic and Leveson5 Additionally, younger HCWs may be at higher risk due to their lack of experience and knowledge of infection control and safety procedures. Reference Motaarefi, Mahmoudi and Mohammadi20 Targeted training and education to especially those who are relatively inexperienced is paramount; education has been shown to be effective in reducing AEBs. Reference Van der Molen, Zwinderman and Sluiter21
Inconsistent work schedules and rotating shifts can increase the risk of AEBs. Our findings suggest that healthcare facilities should consider the impact of work-shift changes and schedule consistency on the health and patient safety. Reference Caruso22 It is essential to allow HCWs sufficient time to rest and recover between shifts as well as appropriate support to cope with schedule changes. Extended work hours and insufficient rest periods are known to increase AEB risk. Reference Trinkoff, Le and Geiger-Brown23 Additionally, occupational injuries can result from consecutive and cumulative shifts. Reference Hopcia, Dennerlein and Hashimoto24 Prolonged work hours can also lead to sleep disruption, negatively affecting HCW performance. Reference Geiger-Brown, Rogers and Trinkoff25 Thus, healthcare facilities should consider implementing strategies, such as shorter work hours, flexible scheduling, and regular breaks during shifts, to mitigate AEB risks associated with work schedules and shift rotation.
Insufficient support from supervisors can lead to increased stress among HCWs, negatively influencing their health, morale, and productivity. Reference Cimiotti, Aiken and Sloane26 Healthcare facilities should promote social support and safety climate among their staff Reference Gershon, Karkashian and Grosch8 through regular meetings with supervisors to discuss work-related challenges, constructive feedback, and a positive work environment. Reference Aiken, Sloane and Bruyneel27 Factors such as work environment, teamwork, burnout, and personal circumstances can influence the intent of European nurses to leave their job. Reference Estryn-Béhar, Van der Heijden and Ogińska28 Addressing these factors is essential for staff retention. By fostering a supportive culture and adequate nurse staffing, healthcare organizations can decrease AEB risk and improve overall staff safety and quality of care. Reference Needleman, Buerhaus and Mattke29 Support from supervisors is essential in reducing AEB risk because it promotes a positive safety culture and HCW adherence to safety protocols. Reference Perry, Jagger and Parker30
Using external staff in healthcare facilities can result in various issues, including increased risk of infection, accidents, and challenging work conditions. Reference Stone, Pogorzelska and Kunches31 We hypothesize that this utilization of external staff may serve as a marker for unit staffing instability and culture. Staff operating in multiple healthcare facilities may act as a vector for spreading infections between these locations. Reference Albrich and Harbarth32 Moreover, HCWs may be more susceptible to accidents given their unfamiliarity with equipment or facility layout, increasing the risk of falls, needle-stick injuries, and other mishaps. Reference Clarke33 The association between care left undone and temporary nursing staff ratios in acute-care settings underscores the need to address staffing for patient safety. Reference Senek, Robertson and Ryan34 These staff members might encounter work-related challenges such as job insecurity, dissatisfaction, and burnout. Such challenges arise due to disparities in training and support compared to permanent staff.
Despite these findings, generalizing our results to other hospitals or countries may be limited due to potential variations in organizational practices, prevention policies, cultural contexts, regulations, and available resources in different healthcare settings. Reference Braithwaite, Herkes and Ludlow35 Moreover, healthcare systems and staffing models can differ significantly between regions, potentially influencing the dynamics of AEB risks.
Underreporting of AEBs is a crucial concern with significant implications. Reference Battail, Fort and Denis36 The main causes of underreporting include fear of negative consequences such as stigmatization, legal liability, or disciplinary actions, as well as insufficient awareness, knowledge, and training on reporting procedures. Reference Gupta, Anand and Sastry37 Additionally, time constraints and complex reporting systems contribute to staff reluctance to report AEBs. Reference Perry, Jagger and Parker30 The underreporting of AEBs prevents efforts to improve healthcare worker safety and hinders the development of effective interventions to minimize the risk of infection transmission. Reference Auta, Adewuyi and Tor-Anyiin38 It also perpetuates a culture of secrecy, rather than fostering an open and transparent environment in which learning from incidents is encouraged.
The study has several strengths that enhance its validity and reliability. It was a multicenter study encompassing diverse HCWs and specialties across 4 hospitals. The longitudinal design provided a comprehensive view of the occupational blood exposure effects over time, allowing for trend analyses. Moreover, a combination of diverse metrics, including individual and organizational factors, as well as 2 levels of data (ward level and HCW level), enriched the understanding of factors influencing HCW health outcomes. We used validated scales for measuring stress, fatigue, overinvestment, social support, and human resources data (absenteeism, turnover) to facilitate the identification of contributing factors. Finally, the study was conducted in 2019, before the COVID-19 pandemic, which significantly altered healthcare organization worldwide. Reference Blumenthal, Fowler and Abrams39 Hence, our findings over the 12-month period were likely unaffected by the effects of the pandemic, an essential consideration when interpreting the results.
This study had several limitations. Data collection regarding physicians’ work characteristics, particularly work hours and shifts, was imprecise. Self-report bias and accident underreporting might have influenced our findings, with potential recall or social desirability bias. Accident underreporting could result from HCW hesitation to report incidents of exposure due to fear of retaliation or reluctance to admit mistakes. This issue may hinder supervisors from providing necessary support and resources, making it difficult to track and prevent future incidents of exposure. We were not able to measure the underreporting of AEBs. In addition, our findings did not establish a direct link between stress, fatigue, and AEBs, possibly due to the presence of confounding factors. Measurements were taken at a single time point, possibly inducing measurement bias. Although AEB data were collected over a longer 4-month period, conclusive evidence of a relationship between stress and fatigue and infectious risks did not emerge.
The study findings have important implications for healthcare organizations, clinicians, and future research in HCW safety. By understanding factors associated with AEBs, facilities can develop targeted interventions addressing risks related to occupation, age, work schedule, shift rotation, and supervisor support. These efforts could include additional training, safer work schedules, and promoting a supportive organizational culture. Healthcare facilities should consider the risks of outsourcing staff and should ensure that HCWs are trained and familiar with infection control protocols to minimize AEBs and other adverse outcomes. HCWs should be encouraged to accurately report AEBs in a supportive environment that minimizes underreporting, with anonymous systems or educational programs emphasizing the importance of reporting for safety improvement.
Future research should develop robust models integrating clinical and organizational factors to better understand the relationships between stress and fatigue and the occurrence of AEBs. Utilizing alternative analytical approaches, such as directed acyclic graphs (DAGs), Reference Textor, van der Zander and Gilthorpe40 could reveal previously unidentified relationships, guiding the development of more effective prevention strategies and enhancing HCW safety and patient care.
Acknowledgments
We acknowledge the support and funding provided by Assistance Publique – Hôpitaux de Paris (Département de la Recherche Clinique et du Développement). We also express our gratitude to all the HCWs who participated in this study.
Financial support
This study was funded by a grant from Programme de Recherche sur la Performance du Système de Soins - PREPS 2016 (Ministère de la Santé).
Competing interests
All authors report no conflicts of interest relevant to this article.
Open access
