Introduction

Healthcare-associated infections remain among the most frequent and consequential complications of hospital care. The problem is especially pronounced in anesthesiology, resuscitation, and intensive care units, where patients often have severe underlying disease, impaired host defenses, prolonged exposure to the hospital environment, and repeated invasive interventions. Mechanical ventilation, central venous and urinary catheters, extracorporeal support, and frequent manipulation of devices increase the number of opportunities for microbial transmission. In the EPIC III study, infection was recorded in a substantial proportion of intensive care patients and was associated with poorer clinical outcomes [1].

HAIs are generally understood as infections that arise in connection with healthcare delivery and were neither present nor incubating at the time of admission. They are associated with increased mortality, longer hospital stays, antimicrobial use, and additional costs [2]. Surveillance studies from intensive care settings consistently identify ventilator-associated pneumonia, bloodstream infection, urinary tract infection, and other device-associated infections among the main clinical forms [3,4]. Their distribution varies across hospitals and countries, but the combination of patient acuity and high device utilization makes intensive care a priority setting for prevention.

Infection prevention and control depends on reliable performance of routine measures: hand hygiene, appropriate use of personal protective equipment, aseptic technique, environmental cleaning, isolation precautions, and adherence to device-care bundles. These measures are implemented through repeated actions by healthcare workers and therefore require adequate knowledge of transmission routes, correct procedural skills, realistic risk perception, and consistent professional habits. Knowledge is a necessary foundation; however, it is not equivalent to performance. Studies repeatedly describe a gap between questionnaire-based knowledge and observed adherence to standard precautions [5,6].

This gap has practical importance. A worker may correctly identify the indications for hand hygiene yet omit the procedure during periods of high task intensity. Another may know the elements of a catheter-care bundle but be unable to complete them when staffing is insufficient or equipment is not immediately available. Workload, shift duration, interruptions, staffing levels, skill mix, access to supplies, supervisory support, and the local safety climate can therefore modify the extent to which knowledge is translated into practice [7,8]. These conditions are not secondary background variables; they shape the feasibility and reliability of preventive care.

Kazakhstan faces the same clinical and organizational challenges. Available studies from Astana and multicenter hospital surveys have documented HAIs, substantial antimicrobial use, and a prominent role of Gram-negative organisms, including infections occurring in critical care units [9,10]. At the same time, published evidence from Kazakhstan on healthcare workers’ infection-control behavior, nursing workload, staffing, and organizational safety remains limited. This creates a need to interpret local findings within the broader international evidence base while avoiding direct transfer of recommendations that may not fit staffing models, infrastructure, or resource availability.

Previous publications have often considered knowledge, hand hygiene, staffing, workload, or HAI epidemiology as separate topics. Such separation does not fully explain why educational programs may improve test scores without producing stable changes in practice. The present review addresses this limitation by examining the implementation pathway that connects cognitive readiness, organizational capacity, preventive behavior, and patient-safety outcomes. The intended contribution is not a new pooled estimate, but an integrated explanation of the knowledge–practice gap in high-risk hospital care.

The objective of this structured narrative review was to synthesize evidence on cognitive and organizational factors associated with HAI prevention. Three questions guided the analysis: (1) how are healthcare workers’ knowledge and training related to preventive practice; (2) which organizational conditions facilitate or hinder adherence; and (3) how do cognitive and organizational factors interact in intensive care and other high-risk settings?

Materials and Methods

Review design and reporting approach

This work was designed as a structured narrative review. A narrative approach was selected because the evidence base included heterogeneous study designs, settings, professional groups, measurement instruments, and outcomes. The review combined descriptive mapping with critical thematic synthesis rather than statistical pooling. Preparation and revision of the manuscript were guided by the principles of the SANRA instrument, particularly the need for a clearly stated objective, a transparent description of literature searching, appropriate referencing, scientific reasoning, and presentation of relevant endpoint data [50].

The review did not constitute a systematic review or meta-analysis. No pooled effect estimate was calculated, and no claim of exhaustive global coverage is made. Nevertheless, a structured search and explicit eligibility criteria were used to reduce arbitrary source selection and to make the basis of the synthesis clear.

Information sources and search strategy

Literature was searched in PubMed (MEDLINE) and Google Scholar. The search covered publications from 2016 to July 2026. The 2013 study by Stimpfel et al. [8] was retained outside the primary date range because it provides foundational evidence on long hospital shifts and organizational burden and was directly relevant to interpretation of workload-related findings.

The search strategy combined terms from four concept blocks: (1) healthcare-associated infection and hospital-acquired infection; (2) infection prevention, infection control, standard precautions, hand hygiene, and personal protective equipment; (3) healthcare workers, nurses, physicians, knowledge, attitude, practice, compliance, and training; and (4) workload, staffing, missed care, organizational factors, intensive care, and ICU. Boolean operators AND and OR were used to combine synonyms and concept blocks. Search terms were applied in English. Reference lists of relevant reviews and eligible articles were also examined to identify additional publications closely related to the review questions.

A representative search formulation was: (“healthcare-associated infection*” OR “hospital-acquired infection*” OR “nosocomial infection*”) AND (“infection prevention” OR “infection control” OR “standard precaution*” OR “hand hygiene”) AND (“healthcare worker*” OR nurse* OR physician*) AND (knowledge OR practice OR compliance OR training OR workload OR staffing OR “organizational factor*” OR “intensive care” OR ICU). Database syntax was adapted to the requirements of each source.

Eligibility criteria

Eligible publications examined one or more of the following: HAI epidemiology in intensive care; healthcare workers’ knowledge, attitudes, skills, or preventive practices; hand hygiene or standard-precaution compliance; education and training; staffing and nursing workload; missed care; organizational support; or associations between these factors and HAI-related outcomes. Studies involving physicians, nurses, other healthcare workers, or multidisciplinary teams were considered. Intensive care was the principal context, while studies from emergency departments and other high-risk hospital units were retained when they addressed mechanisms directly relevant to infection-prevention implementation.

Primary quantitative and qualitative studies, surveillance studies, mixed-methods studies, systematic reviews, rapid reviews, and meta-analyses were eligible. Conference abstracts, brief reports without sufficient methodological or outcome information, publications unrelated to the review questions, and records for which the full text could not be assessed were excluded. Studies focused exclusively on pathogen susceptibility without a link to prevention practice, staffing, workload, or organizational conditions were not used for the thematic conclusions.

Study selection and data extraction

Records were screened in two stages. Titles and abstracts were assessed first, followed by full-text evaluation of potentially relevant publications. The review team considered relevance to the predefined cognitive and organizational domains, the clarity of methods, and the availability of interpretable outcomes. The selection process identified 183 records, of which 65 were excluded at title and abstract screening. Full texts were assessed for 118 publications; 69 were excluded because the full text was unavailable, the content was not relevant to the review question, or the publication was a conference material. Forty-nine publications were included in the qualitative synthesis.

For each included publication, the following information was extracted where available: country and clinical context, participant group, design, sample size, method used to measure knowledge or practice, workload or staffing indicator, infection-prevention outcome, HAI outcome, principal association, and major methodological limitation. Data were organized in a working evidence matrix before thematic synthesis.

Critical appraisal and synthesis

Because the review included markedly different designs, a single formal risk-of-bias instrument was not applied across all publications. The evidence was appraised qualitatively. Greater interpretive weight was assigned to systematic reviews and meta-analyses, multicenter surveillance studies, prospective cohorts, studies using staffing or Nursing Activities Score data, and studies based on direct observation. Findings from single-center cross-sectional surveys and self-reported compliance measures were treated more cautiously because of selection, recall, and social-desirability bias.

The synthesis was conducted in four stages. First, findings were grouped into cognitive factors, observed preventive behavior, organizational conditions, and patient-safety outcomes. Second, agreement and inconsistency between studies were examined. Third, the role of measurement method was considered, particularly differences between self-report and direct observation. Fourth, an implementation pathway was developed to explain how cognitive readiness and organizational capacity jointly influence adherence. The synthesis was qualitative because definitions, instruments, populations, and outcome measures were too heterogeneous for meaningful statistical pooling.

Results

Study selection and profile of the evidence

The search and selection process resulted in 49 included publications (Figure 1). The evidence base comprised global and regional cross-sectional studies of knowledge, attitudes, and practices; studies of hand hygiene using self-report or direct observation; educational intervention studies; ICU surveillance cohorts; analyses of staffing and missed care; investigations using the Nursing Activities Score; and systematic reviews or meta-analyses of workload, staffing, and HAI outcomes.

The publications represented diverse healthcare systems and resource settings. This diversity increased the relevance of recurring patterns but limited direct comparison of percentages. Knowledge and adherence were measured using non-equivalent questionnaires, locally developed scores, checklists, direct observation, electronic monitoring, administrative staffing data, and clinical surveillance. The findings are therefore presented as thematic patterns rather than as a single summary value.

Table 1 - Main analytical domains and patterns identified in the included literature

Cognitive factors: knowledge, beliefs, and practical competence

The included studies did not support a uniform description of healthcare workers’ knowledge. In a global cross-sectional study, mean knowledge indicators were relatively high among physicians and nurses, although professional-group differences remained [11]. Other studies reported considerably lower levels, often between 40% and 60%, particularly when questionnaires assessed detailed indications, transmission mechanisms, or application of standard precautions rather than general awareness [6,12,13,18,19]. These differences probably reflect variation in training, professional composition, instruments, scoring thresholds, and clinical context.

The most consistent finding was not simply inadequate knowledge but the separation between knowing and doing. Participants frequently reported awareness of infection-prevention requirements while demonstrating lower compliance in practice [5,6,11–13]. This difference was especially apparent when self-reported behavior was compared with direct observation. Self-report is vulnerable to recall and social-desirability bias, while observation may be influenced by the Hawthorne effect; nevertheless, repeated discrepancies suggest that theoretical knowledge alone is an incomplete proxy for safe practice.

Hand hygiene illustrates this problem. Several studies found that healthcare workers recognized its importance, but correct execution of all required steps was observed in only a proportion of participants [16–18]. Knowledge of indications did not guarantee adequate duration, coverage of all hand surfaces, or performance at every opportunity. In neonatal and intensive care settings, deficits persisted in specific areas despite acceptable overall scores [18,20]. These findings indicate that cognitive assessment should distinguish declarative knowledge from procedural competence.

Training generally improved knowledge and immediate test performance. Targeted educational programs produced statistically significant gains [14], and staff with more recent training tended to perform better than those whose training had occurred earlier [15]. However, one-time instruction did not reliably produce sustained adherence. The evidence favors recurrent training combined with skills assessment, direct observation, feedback, and reinforcement within clinical workflows. Training is therefore best understood as one component of an implementation system rather than an isolated intervention.

Experience, educational level, access to protocols, and availability of protective equipment were also associated with knowledge and practice [19]. Their effects were inconsistent across studies, suggesting that individual characteristics operate within the local work environment. An experienced worker may perform poorly in an overloaded unit, while a less experienced worker may follow protocols reliably when supervision, supplies, and workflow design are adequate. The relation between cognitive factors and behavior is therefore conditional rather than linear [21,42].

Organizational factors: workload, staffing, resources, and work design

Organizational conditions determine whether preventive procedures can be completed reliably during routine care. In intensive care, workload is not captured by patient numbers alone. It also reflects patient acuity, frequency of interventions, documentation, monitoring, admissions, transfers, emergency events, isolation procedures, and the number of device-related tasks. A single critically ill patient may generate a large number of hand hygiene opportunities and aseptic procedures within a short period.

Cross-sectional evidence shows that nurses and other healthcare workers frequently identify workload and lack of time as barriers to standard precautions [22,25]. These self-reports are consistent with observational findings. Chang et al. found that hand hygiene compliance remained relatively stable up to approximately 30 opportunities per hour and then decreased by about one percentage point for each additional opportunity [7]. A separate prospective study using continuous electronic measures also found an inverse relationship between workload and hand hygiene performance even in a highly trained and well-staffed setting [23]. These studies strengthen the interpretation that workload is not merely a perceived excuse but a measurable constraint.

High patient-load periods were also associated with reduced compliance with broader infection-control protocols, including personal protective equipment [24]. The effect was not limited to a single procedure. Under time pressure, healthcare workers must prioritize urgent clinical tasks, and preventive actions may be delayed, abbreviated, or omitted. This mechanism is consistent with research on missed nursing care, in which more than 70% of nurses reported omission of at least one required activity during a shift [29]. Preventive and supportive tasks are particularly vulnerable because their consequences may not be immediately visible.

Staffing is closely related to workload but should be considered separately. A reduced number of staff members increases task density and limits the ability to redistribute work during deterioration, admission, or emergency procedures. Systematic reviews have linked lower staffing levels with hospital-acquired conditions and HAIs [27,28]. Unit-level and multisource studies similarly reported associations between staffing, missed care, safety attitudes, quality, and infection outcomes [29,30]. Short-term staffing reductions may affect infection risk within the following days, which is biologically plausible because lapses in asepsis or device care may precede clinical recognition of infection.

Workload studies using the Nursing Activities Score show that required nursing time may approach or exceed available capacity [31–33]. During admission, deterioration, or high-acuity care, direct patient-care demands increase further. When workload exceeds staffing capacity, adherence cannot be treated solely as an individual motivational issue. The problem becomes one of work-system design, including task allocation, skill mix, shift scheduling, supply placement, and protected time for infection-prevention activities.

Resource availability and organizational support modify the effect of workload. Access to hand-rub dispensers, personal protective equipment, written protocols, functioning isolation facilities, and timely replacement of supplies reduces the effort required to perform preventive actions [19,26]. Leadership, audit, and feedback also matter. Staff are more likely to treat infection prevention as a non-negotiable component of care when expectations are consistent, data are visible, and supervisors respond to system barriers rather than attributing every lapse to individual negligence [42–45].

Interaction between cognitive and organizational factors

The evidence supports an interaction model. Cognitive readiness includes knowledge of transmission, risk perception, understanding of protocols, and practical skills. Organizational capacity includes staffing, workload, time, supplies, supervision, audit, feedback, and safety culture. Reliable preventive behavior is most likely when both are adequate. Weakness in either component can disrupt implementation.

When cognitive readiness is low, staff may not recognize an indication for hand hygiene, may use personal protective equipment incorrectly, or may fail to understand the purpose of a bundle element. When organizational capacity is low, staff may recognize the correct action but omit it because several urgent tasks compete for the same time. Education primarily addresses the first pathway; staffing, workflow redesign, point-of-care supplies, and leadership address the second. A program limited to training may therefore improve knowledge scores without reducing missed care.

This interaction also explains variation between settings. The same educational intervention may be effective in a unit with stable staffing and accessible supplies but have little impact in a unit with sustained overload. Conversely, increasing staffing without addressing knowledge gaps or local norms may not ensure correct practice. Multicomponent interventions are more consistent with the causal structure suggested by the evidence than single interventions aimed only at individual behavior.

Association with HAIs and other patient-safety outcomes

HAIs remain common in intensive care populations, although reported prevalence varies widely because of differences in case mix, surveillance definitions, device utilization, and healthcare systems [38–41,46]. Device-associated infections are especially important because their prevention depends on repeated, standardized care processes. Surveillance studies show that HAIs are associated with mortality and other adverse outcomes [38,39,48].

The relationship between organizational factors and infection outcomes is supported by several lines of evidence. Reviews and observational studies associate nursing workload with infections, pressure injuries, medication errors, mortality, and poorer patient-focused outcomes [34,35]. Prospective work has linked staffing and workload with ventilator-associated pneumonia and mortality [47]. Similar relationships have been described for other device-associated infections [48]. These findings are clinically coherent, but causal interpretation should remain cautious because staffing and workload are also influenced by patient severity, unit type, resources, and case complexity.

During infectious outbreaks, workload increases because of isolation measures, additional monitoring, communication requirements, and staff absence [36]. Adherence may remain incomplete despite greater awareness of infection risk [37]. Outbreak conditions therefore reveal the limits of relying on motivation alone: heightened concern does not remove time constraints, supply problems, or staffing shortages.

Evidence relevant to Kazakhstan

Kazakhstan-specific evidence confirms the clinical relevance of HAIs but remains insufficient for a detailed analysis of cognitive and organizational mechanisms. The observational study from a critical care unit in Astana and the later multicenter point-prevalence survey documented HAI and antimicrobial-use patterns consistent with the wider international literature [9,10]. These studies establish the importance of surveillance, but they do not provide a complete national picture of healthcare workers’ knowledge, observed adherence, staffing adequacy, or nursing workload.

The main research gap is therefore not the absence of international evidence, but the limited availability of standardized local data linking staff knowledge, direct observation of practice, objective workload indicators, staffing, and HAI outcomes. Future studies in Kazakhstan should use validated instruments, clearly defined denominators, device-days, nurse-to-patient ratios, Nursing Activities Score or comparable workload measures, and prospective surveillance. Multicenter designs would allow comparison between hospitals and help distinguish local organizational effects from differences in patient severity. The design and principal analytical domain of all included publications are summarized in Supplementary Table 1 [1–49].

Supplementary Table 1 - Design and principal review domain of the included publications