The research, published in the open-access journal Nursing Open, addresses one of the most persistent and underappreciated problems in global critical care. Although intensive care units account for a small fraction of hospital beds worldwide, they generate more than 20 percent of all hospital-acquired infections. Among these, ventilator-associated pneumonia stands apart. Defined as pneumonia that develops 48 hours or more after endotracheal intubation, it strikes between 10 and 40 percent of mechanically ventilated patients globally, and in low- and middle-income countries the incidence can climb to as high as 47.9 cases per 1,000 ventilator days. Each episode prolongs intensive care stays, multiplies resource consumption, and increases the likelihood of death. Decades of clinical research have established that a constellation of simple, evidence-based interventions, often called prevention bundles, can slash these infection rates: elevating the head of the bed, meticulous oral hygiene, appropriate suctioning technique, judicious ventilator circuit management, and the use of endotracheal tubes equipped with subglottic secretion drainage. The tragedy, as the new study makes clear, is that in many resource-limited settings these interventions are simply not delivered consistently, and the reasons why are more complex than a lack of training alone.

To disentangle the relationship between nursing knowledge and systemic constraints, the research team employed an explanatory sequential mixed-methods design, one of the more rigorous architectures available for studying implementation problems in real-world clinical environments. The quantitative strand used a single-group quasi-experimental pre-test–post-test approach involving registered nurses from two intensive care units at the hospital: a four-bed General Intensive Care Unit staffed by 37 personnel and a six-bed Maternity Intensive Care Unit staffed by 27. Fifty-seven nurses completed the baseline knowledge assessment, using an eight-item multiple-choice questionnaire adapted from an established instrument developed by Labeau and colleagues, with one question about kinetic beds removed because that equipment does not exist at the study site. After the pre-test, all participants attended a structured three-hour workshop covering four thematic modules: the definition, aetiology, and pathophysiology of ventilator-associated pneumonia; evidence-based prevention bundles including patient positioning, oral hygiene, and circuit management; airway management and suctioning techniques; and intensive care-specific infection control principles. The content was validated in advance by two independent intensivists and senior nursing leadership. A post-test was administered within two weeks, completed by 56 nurses, and eight purposively selected participants then underwent semi-structured interviews designed to explain why the quantitative results looked the way they did.

The statistical results were striking. Mean knowledge scores rose from 3.63 out of 8 (standard deviation 1.57) before the workshop to 5.29 out of 8 (standard deviation 1.57) afterward, an independent-samples t-test yielding t(111) = −5.61, p < 0.001, with a mean difference of 1.65 points and a Cohen’s d of 1.06, which is conventionally classified as a large effect. A non-parametric Mann–Whitney U test confirmed the robustness of the finding (U = 748.5, p < 0.001). Perhaps more meaningful clinically was the transformation in competence categories. Before the workshop, only 24.6 percent of nurses scored in the high-competence range, while nearly a quarter fell into the low category. Afterward, the high-competence group swelled to 64.3 percent and the low group shrank to a single nurse, a redistribution that the chi-square test confirmed as highly significant (χ²(2, N = 113) = 22.43, p < 0.001). The largest item-level gains appeared precisely where baseline knowledge was weakest: correct responses about humidifier change frequency jumped by 17, open-versus-closed suction system knowledge by 16, patient positioning by 16, and awareness of endotracheal tubes with subglottic secretion drainage by 13.

Curiously, the factors one might expect to predict post-training performance did not. A multiple linear regression examining years of registered nurse experience, duration of intensive care service, and unit type found that none of these variables significantly predicted post-test scores, and the overall model explained a negligible 5.2 percent of the variance (F(3, 52) = 0.94, p = 0.428). The study’s demographic data help explain why. The nursing workforce was young, with a mean age of 31.6 years, overwhelmingly composed of registered general nurses (84.1 percent), and almost entirely lacking postgraduate critical care credentials, only 1.8 percent held such qualifications. More than three-quarters had five or fewer years of intensive care experience. The authors interpreted the null regression result through the lens of the qualitative interviews: when supervision is sparse, protocols are absent, and supplies are inconsistent, years of experience can accumulate without ever being reinforced against evidence-based standards, so experience alone stops being a reliable proxy for competence.

That interpretation was powerfully supported by the interview data. Nurses described a practice culture built on experiential learning and pragmatic task selection rather than standardized bundles. “What has helped is my experience in the ICU,” one participant explained. Another described a ward habit of routine mouth care for every ventilated patient, prioritized because it was feasible and perceived as effective. But the interviews also revealed the harsh material realities that constrain even the best-informed clinician. Nurses reported reusing suction catheters on multiple patients after washing them, hunting for scarce consumables, and losing power mid-suction. “You can even try to suction a patient then all of a sudden there’s no light,” one nurse recounted. Others described the absence of any written protocol for pneumonia prevention, with practice resting entirely on individual knowledge, and the weak standing of in-service training in the unit’s culture, where colleagues sometimes viewed education sessions as a waste of time. Supervisors, participants said, rarely came to observe practice, leaving “the system open for anyone to do what he or she thinks is the best.” Delayed laboratory feedback and families’ financial struggles to secure medications after the first week of hospitalization added further layers of difficulty.

The study was theoretically anchored in two influential frameworks from nursing and behavioral science: Patricia Benner’s Novice-to-Expert Model, which frames clinical competence as a developmental construct acquired primarily through experience rather than didactic instruction, and Bandura’s Self-Efficacy Theory, which holds that knowledge and confidence together drive the performance of specific behaviors. The findings map onto both frameworks in instructive ways. The overwhelmingly experiential nature of the nurses’ knowledge acquisition is consistent with Benner’s account of how competence develops, but it also explains why a single structured workshop could produce such large gains, because it filled gaps that years of unstructured bedside experience had never addressed. Meanwhile, the supervisory and accountability gaps described in the interviews illustrate why self-efficacy, however boosted by a mastery-oriented learning experience, cannot sustain consistent preventive behavior without institutional reinforcement.

The authors are appropriately candid about the study’s limitations. Because participant responses were not linked across assessment points, the improvement represents a group-level change rather than demonstrated within-individual gain, and the single-group design without a concurrent control group precludes strong causal inference. The single-site setting, convenience sampling, the two-week interval between training and post-test, and the low internal consistency of the adapted questionnaire (KR-20 of 0.32 pre-test and 0.42 post-test, reflecting a broad knowledge index rather than a unidimensional psychometric scale) all warrant caution. The study also measured knowledge rather than observed bedside performance or actual pneumonia incidence, leaving open the crucial question of whether the educational gains translate into fewer infections.

Nevertheless, the practical implications are difficult to ignore, and the authors frame them with unusual clarity. Education, they conclude, is necessary but insufficient. The most actionable levers for nursing leaders in similar settings are the introduction of locally feasible written prevention protocols and bedside checklists, routine audit-and-feedback cycles, structured supportive supervision, and, perhaps most fundamentally, dependable procurement of basic consumables such as suction catheters, closed-system circuits, and personal protective equipment. At the health-system level, reliable electricity and functional laboratory services emerge as prerequisites for infection prevention rather than optional luxuries. From an implementation science perspective, the study illustrates how individual capability, organizational readiness, available resources, and implementation climate interact to determine whether evidence-based practice takes hold, a dynamic well captured by frameworks such as the Consolidated Framework for Implementation Research.

In an era when global health attention has rightly focused on pandemic preparedness and antimicrobial resistance, this study is a reminder that some of the most consequential advances in patient safety may come from modest, low-cost interventions targeted at the frontline workforce, provided they are embedded within broader quality improvement efforts. A three-hour workshop, materials prepared from peer-reviewed literature and delivered with a projector, moved the majority of nurses from low or moderate competence to high competence in a setting where virtually none had formal critical care training. The next challenge, the authors argue, is to pair such education with the protocols, supervision, supplies, and infrastructure that allow knowledge to become practice, and to evaluate with paired, longitudinal designs whether combined training-plus-system-support strategies can ultimately reduce ventilator-associated pneumonia rates, shorten ventilator days, and save lives in the intensive care units that need it most.

NAVA is a form of mechanical ventilation designed to respond to a patient’s own respiratory drive. Unlike conventional modes that rely primarily on preset pressure, volume or flow targets, NAVA detects the electrical activity of the diaphragm, known as the electrical activity of the diaphragm, or EAdi. This signal is measured through specialized electrodes embedded in a nasogastric or orogastric catheter positioned near the diaphragm. The ventilator then delivers assistance in proportion to the patient’s neural breathing effort, creating a more direct connection between the patient and the machine.

That connection is intended to improve patient–ventilator synchrony. In conventional ventilation, a delay may occur between the patient’s attempt to inhale and the ventilator’s response. The machine may also continue delivering support after the patient has stopped inhaling, a mismatch known as delayed cycling. By using EAdi as a trigger and cycling signal, NAVA can theoretically respond more rapidly to changes in respiratory effort and reduce ineffective triggering, excessive assistance and other forms of asynchrony. The approach is especially relevant in intensive care, where rapidly changing physiology makes fixed ventilator settings difficult to maintain.

Yet NAVA is not simply a matter of connecting a catheter and selecting a mode on the ventilator. Clinicians must understand how the EAdi signal is obtained, how catheter positioning affects signal quality, how NAVA levels influence pressure delivery and how to interpret changes in the patient’s neural respiratory drive. They also need to distinguish an appropriate physiological response from a technical problem, inadequate sedation, worsening lung disease or patient fatigue. These requirements can make adoption challenging, even when a technology is available within an intensive care unit.

The study addressed this implementation challenge through an education bundle combining multiple forms of learning. Although education in critical care often consists of a single lecture or brief equipment demonstration, multimodal programs can reinforce knowledge through different channels, including explanation of underlying physiology, practical instruction, guided review of ventilator waveforms and opportunities to apply new skills. Such an approach is designed to reach clinicians with varied learning preferences while connecting theoretical understanding to decisions made at the bedside.

The investigators evaluated objective knowledge acquisition as well as clinicians’ subjective experiences. Objective knowledge measures can reveal whether participants understand the technical and physiological principles needed to operate NAVA safely. Comfort and confidence provide a different perspective: a clinician may answer examination questions correctly but remain hesitant when selecting settings, troubleshooting alarms or interpreting a patient’s response. Assessing both dimensions allows educators to identify whether a program produces genuine readiness or merely short-term familiarity.

The project also examined barriers to implementation, an important element often overlooked in technology adoption. Obstacles may include limited exposure to NAVA cases, uncertainty about patient selection, difficulty maintaining catheter position, inconsistent staff training, concerns about workflow and a lack of institutional protocols. Even highly motivated clinicians may struggle to use a mode reliably if equipment access, staffing patterns or team communication do not support it. Identifying these barriers can help hospitals tailor training and develop practical systems for continued competency.

A crucial feature of the evaluation was the search for clinical use after education. Demonstrating that clinicians can describe NAVA is not the same as demonstrating that it has entered routine practice. Post-education use provides a more meaningful test of whether learning has transferred to patient care. In this setting, clinical application would require teams to place and verify the specialized catheter, obtain a usable EAdi signal, select appropriate ventilator parameters and monitor both conventional respiratory measures and neural effort over time.

The findings position education as a central part of NAVA implementation rather than an optional supplement to equipment acquisition. The study suggests that successful adoption depends on more than technical availability: teams need a shared understanding of the physiology, repeated opportunities to practice, confidence in troubleshooting and a plan for addressing local workflow barriers. As intensive care increasingly adopts technologies that personalize support according to real-time biological signals, the experience described by the researchers offers a broader lesson. Translating innovation into safer clinical care requires not only advanced machines, but also education designed to make those machines understandable, usable and visible in everyday practice.

Subject of Research: Evaluation of a multimodal educational intervention for implementing neurally adjusted ventilatory assist (NAVA) in the intensive care unit, including knowledge acquisition, comfort, confidence, implementation barriers and post-education clinical use.

Article Title: From theory to practice: implementing an education bundle for neurally adjusted ventilatory assist (NAVA) in the intensive care unit

Article References: Brei, B.K., Sharma, A., Mickas, K. et al. “From theory to practice: implementing an education bundle for neurally adjusted ventilatory assist (NAVA) in the intensive care unit.” Journal of Perinatology (2026). https://doi.org/10.1038/s41372-026-02854-z

Image Credits: AI Generated

DOI: 10.1038/s41372-026-02854-z

Keywords: neurally adjusted ventilatory assist, NAVA, intensive care, mechanical ventilation, respiratory physiology, clinical education, ventilator synchrony, implementation science, critical care technology