Our systematic review and meta-analysis incorporated several studies that aimed to evaluate the efficacy of inhaled antibiotics in VAP prevention. The results demonstrated that the use of inhaled antibiotics significantly reduced the incidence of VAP with a 33% relative risk reduction compared to controls. However no significant difference in ICU mortality was found between groups, suggesting that while inhaled antibiotics may prevent infections, they do not confer a survival benefit in the general ICU population. The GRADE assessment rated the certainty of evidence as moderate, in part due to moderate heterogeneity, reflecting differences in patient risk profiles, diagnostic criteria, and antibiotic regimens. The GRADE assessment rated the certainty of evidence for VAP prevention as moderate, and low to moderate for ICU mortality (Table S1, Supplementary Material). The ROB assessment revealed an overall moderate risk, with some concerns related to blinding and allocation concealment. Despite these limitations, our findings consistently support inhaled antibiotics as an effective strategy for VAP prevention.

Ehrmann et al. conducted a large, multicenter, double-blind, randomized controlled trial evaluating the use of inhaled amikacin in mechanically ventilated patients [31]. Their findings demonstrated a significant reduction in VAP incidence in the amikacin group (15% vs. 22% in the placebo group, p = 0.004). Furthermore, lower rates of infection-related ventilator-associated complications were also observed, which reinforces the role of inhaled antibiotics as a preventive measure.

Karvouniaris et al. conducted a randomized controlled trial evaluating prophylactically nebulized colistin in an ICU setting with prevalent multidrug-resistant bacteria [25]. This study showed that while nebulized colistin did not significantly reduce the overall incidence of VAP, it did lower the incidence rate and decreased occurrences of Gram-negative and multidrug-resistant VAP. This finding suggests that in specific contexts, such as ICUs with high rates of multidrug resistance gram negative bacteria, nebulized colistin may be beneficial.

Similarly, Wood et al. investigated the use of aerosolized ceftazidime for VAP prevention in critically ill trauma patients [29]. Their randomized, double-blind, placebo-controlled trial demonstrated a significant reduction (73% lower) in VAP incidence among patients receiving aerosolized ceftazidime compared to placebo. The study also noted that aerosolized ceftazidime attenuated proinflammatory pulmonary response, which could have implications for patient outcomes beyond VAP prevention. In contrast, Lode et al. conducted a study on endotracheal gentamicin prophylaxis of VAP in ICU patients [28]. Their findings showed no significant reduction in VAP or mortality rates. This highlights the variability of outcomes on antibiotic choice and administration methods.

Some studies included in this analysis, such as those by Wood, Greenfield, and Klastersky, reported VAP incidence rates that appear considerably higher than the widely cited range of 2–30 cases per 1000 ventilator days. These elevated rates may be attributed to differences in diagnostic criteria, patient severity, or study period and design. The inclusion of such outliers likely contributed to observed heterogeneity and should be taken into account when interpreting the pooled results. It is also worth noting that several studies demonstrating the most pronounced reduction in VAP incidence also reported unusually high VAP rates in their control groups. This pattern, previously observed in other reviews on VAP prevention, may exaggerate the apparent efficacy of inhaled antibiotics due to inflated baseline risk. Readers should interpret large effect sizes with caution, particularly in studies with control group VAP rates exceeding commonly reported background incidence. This reinforces the need for consistent diagnostic criteria and real-world incidence thresholds in future trials.

While the primary focus of this meta-analysis was VAP prevention, our analysis also included data on ICU mortality. The pooled data revealed no significant reduction on overall ICU mortality between the intervention and control groups. Ehrmann et al. reported an ICU mortality rate of 23.7% in the inhaled antibiotic group versus 26.0% in the control group (p = 0.18), which indicates no statistical significance [31]. Similarly, Klick et al. and Rouby et al. also found no substantial differences in mortality rates between treatment and control groups [27, 33]. In contrast to these findings, Karvouniaris et al. reported ICU mortality rates of 7.1% versus 44% (p = 0.028), suggesting a potential mortality benefit in specific populations [25]. This variability in mortality outcomes may be attributed to differences in patient population, antibiotic selection, and study design. Notably, some of the included studies using colistin were conducted in settings with high MDR organism prevalence, although comprehensive resistance data were often not provided. Additionally, studies such as Ehrmann et al. contributed a large portion of statistical weight. Although inhaled antibiotics significantly reduce VAP incidence, the lack of corresponding mortality benefit suggests that other ICU related factors, such as sepsis progression, comorbidities, and overall antibiotic stewardship, may have a stronger impact on survival. Future studies should focus on identifying patient subgroups that may derive the most mortality benefit from inhaled antibiotic prophylaxis.

While inhaled antibiotics offer pharmacokinetic advantages, their use is not without risks. Safety concerns include potential bronchospasm, airway irritation, and ventilator filter obstruction, particularly with certain formulations or delivery devices [39, 40]. Additionally, nebulizer type and technique can significantly influence lung deposition, potentially leading to under- or overdosing [41]. Proper staff training and standardized administration protocols are essential to minimize these risks. Moreover, while inhaled antibiotics may reduce VAP incidence, their indiscriminate use poses a risk of promoting antimicrobial resistance, especially in settings with already high multidrug-resistant (MDR) organism prevalence [42]. Given that most included studies did not report follow-up cultures or resistance patterns, the long-term ecological impact remains unclear—reinforcing the need for strong antibiotic stewardship. Clinicians should consider local susceptibility data, use inhaled antibiotics selectively, and ensure that administration protocols are optimized to prevent subtherapeutic dosing. Further studies are needed to evaluate the emergence of resistance and establish safe duration thresholds for prophylactic regimens.

Trial Sequential Analysis suggested that while current evidence supports a reduction in VAP incidence, the cumulative data still fall short of the required information size for robust confirmation. In contrast, the mortality signal remains indeterminate, and more data are needed before clinical guidance can be firmly based on these endpoints. Furthermore, the stability of our findings across leave-one-out analyses confirms that no single study unduly influenced pooled results, increasing the reliability of our conclusions across both VAP incidence and ICU mortality outcomes.

While the findings are promising, several limitations should be considered. First, heterogeneity among the studies—such as patient population, antibiotic regimen, and treatment durations—may affect the generalizability of our findings. Second, different diagnostic criteria for VAP across studies could introduce potential bias and impact reported incidence rates. Third, the limited number of high-quality randomized controlled trials restricts the strength of evidence, highlighting the need for more standardized protocols in future trials. Additionally, data on long-term outcomes, antibiotic resistance patterns, and adverse effects remains scarce and necessitates further investigation.

Although sufficient data were available to categorize studies by antibiotic class, ICU population type, and delivery method, formal subgroup meta-analyses were not performed. The relatively small number of studies per subgroup and potential for underpowered comparisons may have limited interpretability; future reviews should explore these subgroup effects using formal statistical methods. Additionally, sensitivity analysis excluding large trials such as Ehrmann et al. was not conducted but may be warranted in future reviews to assess the robustness of pooled estimates.