Outcomes of Short Antibiotic Courses for Gram‐Negative Bloodstream Infections in Solid Organ Transplant Recipients—A Systematic Review and Meta‐Analysis

ABSTRACT

Background

Gram‐negative bloodstream infections (GN‐BSIs) contribute significantly to morbidity and mortality. Optimizing antibiotic duration is critical, particularly in immunocompromised populations, but data for solid organ transplant (SOT) recipients remain limited.

Objective

To examine the effectiveness of short versus long antibiotic courses for GN‐BSIs in SOT recipients.

Methods

This systematic review followed PRISMA guidelines (PROSPERO CRD42024620638). We searched four databases through June 2025 for randomized controlled trials (RCTs) and observational studies reporting outcomes in SOT recipients. Risk of bias was assessed, and random‐effects meta‐analyses were conducted. Primary outcome was 90‐day mortality; secondary outcomes included clinical failure.

Results

Seven studies (2 RCTs, 5 cohorts) involving 249 SOT patients were analyzed. In RCTs, 90‐day mortality (RD: −0.07; 95% CI: −0.18 to 0.05; p = 0.25) and clinical failure (RR: 0.62; 95% CI: 0.18–2.15; p = 0.45) were lower in the short‐course group. In cohort studies, 90‐day mortality (RR: 0.36; 95% CI: 0.06–2.08; p = 0.25) was also lower, but clinical failure (RR: 1.89; 95% CI: 0.81–4.43; p = 0.14) was higher in the short‐course group. None of the differences were statistically significant. Certainty of evidence was moderate in RCTs, and very low in the cohort studies.

Conclusion

Our review underscores the importance of including SOT recipients in trials comparing short versus long antibiotic therapy for GN‐BSIs. Although no statistically significant differences were observed between treatment durations, the limited sample size precludes definitive conclusions and highlights the need for larger, high‐quality studies.

In this meta‐analysis of seven studies involving 249 solid organ transplant (SOT) recipients with Gram‐negative bloodstream infections, no statistically significant differences in mortality or clinical failure were observed between short and long antibiotic courses. The limited sample size highlights the need for larger, high‐quality studies that include SOT recipients.

Abbreviations

AEs

adverse events

AHRQ

Agency for Healthcare Research and Quality

AMR

antimicrobial resistance

BSI

bloodstream infection

CI

confidence intervals

CLABSI

central line‐associated bloodstream infection

CRBSI

catheter‐related bloodstream infection

GN‐BSI

Gram‐negative bloodstream infection

GRADE

Grading of Recommendation Assessment, Development, and Evaluation

IAIs

intra‐abdominal infections

MDR

multi‐drug resistant

NOS

Newcastle‐Ottawa Scale

PRISMA

Preferred Reporting Items for Systematic Reviews and Meta‐Analyses

RCT

randomized controlled trial

RD

risk difference

RevMan

Review Manager

ROB2

risk of bias tool‐2

Robvis

Risk of Bias Visualization

RR

risk ratio

SOT

solid organ transplant

UTI

urinary tract infection

1. Introduction

Gram‐negative bloodstream infections (GN‐BSIs) remain a significant contributor to both morbidity and mortality, particularly within immunocompromised populations [1, 2]. One of the primary concerns in managing GN‐BSIs is the appropriate duration of antimicrobial therapy, as extended antibiotic courses are well‐documented to drive the development and spread of antimicrobial resistance (AMR) [3]. Despite the heightened vulnerability of certain patient groups, such as solid organ transplant (SOT) recipients, multiple studies did not demonstrate a consistent difference in clinical outcomes between SOT and non‐transplant patients with GN‐BSIs [1, 4, 5]. This raises critical questions about the optimal duration in immunocompromised populations.

Over the recent years, at least four randomized controlled trials (RCTs) investigated the efficacy and safety of shorter versus longer antibiotic courses in variable patient populations with uncomplicated GN‐BSIs [6, 7, 8, 9]. These studies concluded that abbreviated antibiotic regimens (≤ 7 days) were non‐inferior to traditional extended courses (> 7 days) in terms of patient outcomes, including recurrence rates and mortality. However, a key limitation in these trials was the underrepresentation or outright exclusion of SOT recipients, limiting the generalizability of their findings to this high‐risk group. Studies have shown that this population is significantly more likely to encounter multidrug‐resistant (MDR) pathogens, likely due to repeated healthcare exposures, immunosuppressive therapy, and frequent use of broad‐spectrum antimicrobials [5, 10, 11]. A recent systematic review and meta‐analysis [12] demonstrated that antimicrobial stewardship interventions, including short antibiotic therapy for uncomplicated graft infections in SOT recipients, do not adversely affect infection outcomes or mortality. As such, implementing precise antimicrobial stewardship strategies, including the optimization of treatment duration, is especially crucial for this group.

This review seeks to systematically examine the existing evidence comparing short and long durations of antibiotic therapy for GN‐BSIs in SOT recipients. The goal is to determine whether shorter therapies are both safe and effective, thereby informing targeted antimicrobial stewardship interventions for this specific population.

2. Materials and Methods

The protocol for this systematic review was registered on the PROSPERO database (CRD42024620638), and the review follows the guidelines from the Preferred Reporting Items for Systematic Reviews and Meta‐Analyses (PRISMA) (Table S1 for checklist). No artificial intelligence tools were used in creating this manuscript.

2.1. Eligibility Criteria

We included RCTs, case‐control, and cohort studies that compared short versus long antibiotic durations in treating GN‐BSIs. Case reports and series of less than 10 patients were excluded. No language restrictions were applied. Only studies that included adult (age 18 years or older) SOT recipients with GN‐BSI as the primary target were included. We excluded studies that focused on different types of infection (e.g., community or hospital‐acquired pneumonia, intra‐abdominal infections) and reported on patients with BSI as a subgroup.

Studies on complicated GN‐BSI were excluded. Complicated GN‐BSI was defined as (i) GN‐BSI with a source that is not controlled, that is, BSI with infected hardware, catheters, or devices that have not been removed, or any infected collections that have not been drained. (ii) GN‐BSI in patients who are hemodynamically unstable and did not show clinical improvement within 72 h of effective antimicrobial therapy. (iii) GN‐BSI with evidence of deep‐seated infection, that is, endocarditis, endovascular infection, central nervous infection, necrotizing fasciitis, or osteoarticular infection.

2.2. Outcomes

The primary outcome was 90‐day all‐cause mortality, and secondary outcomes were clinical failure, 30‐day all‐cause mortality, development of adverse events (AEs), development of AMR, and length of hospital stay.

Clinical failure was defined as (i) recurrence of BSI due to the same pathogen in the first 90 days following treatment, (ii) progression of infection with evidence of local suppurative complications or distant foci of infection that are due to the same pathogen and were not present at onset of infection within the first 30 days. Development of AMR was defined as emergence of resistance to the antibiotic used in treating the initial infection, in any positive culture of the same pathogen within 30 days following treatment.

2.3. Information Sources

A comprehensive search strategy was developed using a combination of database‐specific subject headings and text words for the main concepts of Gram‐negative bacteria, bloodstream infections, antibiotics, and short duration. Conference materials were removed from Embase. We searched Medline, Embase, Cochrane Database of Systematic Reviews, and the Cochrane Central Register of Controlled Trials from inception until June 5, 2025 (Table S5). The reference lists of included publications were also searched and considered for inclusion. All identified studies (10779) were then exported to a review management software.

2.4. Study Selection and Data Collection

After removing all duplicates (2570), (Figure 1), two independent reviewers (M.A. and A.M.) reviewed articles’ titles and abstracts for eligibility, disagreements were resolved by a third reviewer (S.H.). Pilots were run before each step to apply the screening criteria among the reviewers. All identified studies that focused on antibiotic durations for GN‐BSI on title and abstract screening were included in the full‐text review to identify studies that included SOT patients. For studies that lacked detailed information on SOT recipients, the corresponding author was contacted for details. Data on study design, inclusion and exclusion criteria, sample size, SOT population size, transplant type, infection sources, types of bacteria, proportion of MDR pathogens, and clinical outcomes were extracted.

FIGURE 1.

Preferred Reporting Items for Systematic Reviews and Meta‐Analyses (PRISMA) flowchart of the study selection process.

2.5. Risk of Bias

Two independent reviewers (M.A. and A.M.) assessed each study's risk of bias, and disagreements were resolved by a third reviewer (S.H.). We used the Cochrane Risk‐of‐Bias Tool 2 (ROB2) [13] for RCTs and the Newcastle‐Ottawa Scale (NOS) [14] for cohort studies. Observational studies were considered to have good quality, fair quality, or poor quality based on the Agency for Healthcare Research and Quality (AHRQ) standards [15]. Robvis (Risk of Bias Visualization) was used to create the risk of bias plots.

2.6. Data Synthesis

Given the difference in study designs, we analyzed RCTs and observational studies separately. A random‐effects model was employed to ensure a more conservative estimation of effect sizes. Data were synthesized using risk ratios (RR) with corresponding 95% confidence intervals (CIs). All event rates were unadjusted. In cases of zero events in one arm, a continuity correction was applied by adding a fixed value of 0.5 to enable calculation of RR. In studies with zero events in both arms, risk difference (RD) was used as the effect measure. To evaluate statistical heterogeneity and determine the suitability of conducting meta‐analyses, the I ² statistic was calculated. All statistical analyses were conducted using RevMan version 5.4.1 (Review Manager for MacOS). Assessment of publication bias using funnel plots or statistical methods could not be done due to the small number of included studies.

2.7. Subgroup and Sensitivity Analyses

To evaluate the robustness of the pooled estimates, sensitivity analyses were conducted. These included reanalyzing the primary and secondary outcomes using a fixed‐effect model and using RD in outcomes with double‐zero events. Subgroup analyses were performed based on the type of organ transplant.

2.8. Certainty of Evidence

The certainty of evidence was assessed using the Grading of Recommendation Assessment, Development, and Evaluation (GRADE) framework. This evaluation considered key domains including risk of bias, consistency of results, directness of evidence, precision of effect estimates, outcome relevance, and potential for publication bias [16]. This was done for RCTs and for observational studies separately.

3. Results

3.1. Included Studies

A total of 8209 titles were screened. From which, 152 studies were selected for full‐text review. Of these, 13 studies met the predefined eligibility criteria and were pursued for retrieval of SOT‐specific data through author correspondence if not available in the manuscript. SOT data could not be obtained from five of these studies [17, 18, 19, 20, 21], and one study included SOT patients exclusively in the long‐duration treatment arm precluding comparative analysis [22]. Consequently, seven studies were ultimately included in the final review [6, 7, 23, 24, 25, 26, 27] (Figure 1).

Among the included studies, two were RCTs, and five were retrospective cohort studies (Table 1).

TABLE 1.

Basic characteristics of the included studies.

Study (Country)	Design	Period	Definition of short versus long	Total sample size	Total SOT patients	Types of transplants	Sources of infection	Main pathogens	MDR pathogens (n, %)
Mercuro et al. [23] (USA)	Retrospective cohort	2013–2016	≤ 10 days versus > 10 days	224	11	Abdominal (kidney ± pancreas)	UTI n = 10 (91%). IAIs n = 1 (9%)	E. coli n = 7 (64%)	0 (0%)
Giannella et al. [24] (Italy)	Retrospective Cohort	2013–2016	≤ 10 days versus > 10 days	856	53	Abdominal and thoracic	UTI n = 22 (42%), IAIs n = 20 (38%), Primary n = 9 (17%), CLABSI n = 1 (1.5%), Lung n = 1 (1.5%)	E. coli n = 53 (100%)	NR
Yahav et al. [6] (Israel and Italy)	RCT	2013–2017	7 days versus 14 days	604	51	Abdominal and thoracic	UTI n = 40 (78%), IAIs n = 6 (12%), Primary n = 3 (6%), Lung n = 1 (2%), CLABSI n = 1 (2%)	E. coli n = 30 (59%)	18 (35%)
Ruiz‐Ruigómez et al. [25] (Spain)	Retrospective cohort	2012–2018	≤ 7 days versus > 7 days (following catheter removal)	54	12	Abdominal and thoracic	CRBSI n = 12 (100%)	PsA n = 5 (42%)	6 (50%)
Molina et al. [7] (Spain)	RCT	2014–2016	7 days versus 14 days	248	11	Abdominal a	UTI n = 5 (46%), IAIs n = 2 (18%), CLABSI n = 2 (18%), Primary n = 2 (18%)	E. coli n = 6 (55%)	5 (45%)
De La Villa et al. [26] (Spain)	Retrospective cohort	2014–2022	≤ 10 days versus > 10 days	105	20	Abdominal	NR	CPE n = 20 (100%)	20 (100%)
Miwa et al. [27] (Japan)	Retrospective cohort	2010–2022	≤ 10 days versus > 10 days	91	91	Abdominal	IAIs n = 60 (66%), Primary n = 16 (18%), UTI n = 13 (14%), Lung n = 1 (1%), CLABSI n = 1 (1%)	E. coli n = 26 (29%)	NR

Abbreviations: CLABSI, central‐line associated bloodstream infection; CPE, carbapenemase‐producing enterobacterales; CRBSI, catheter‐related bloodstream infection; E. coli, Escherichia coli; IAIs, intra‐abdominal infections; MDR, multidrug resistant; NR, not reported; PsA, Pseudomonas aeruginosa; RCT, randomized controlled trial; SOT, solid organ transplant; USA, United States of America; UTI, urinary tract infection.

^aTransplant type was not specified in two patients.

The definitions of short and long antibiotic duration varied across studies. The two RCTs uniformly defined short duration as 7 days and long duration as 14 days. In contrast, four retrospective cohort studies classified short duration as 10 days or fewer, and long duration as greater than 10 days. The remaining cohort study [25] was exclusively focused on Gram‐negative catheter‐related bloodstream infection (GN‐CRBSI) and defined short treatment as ≤ 7 days following catheter removal, and long treatment as > 7 days. Five studies excluded patients with polymicrobial BSIs, whereas two studies did not clearly report whether such cases were included or excluded [23, 26]. The main source of infection was urinary tract infections (UTI) in four studies, and intra‐abdominal infections (IAIs) in one study (Table 1). Further details on sources of infection are provided in Table S2.

In total, 249 SOT recipients were included across all studies. Of these, 98 patients received short‐duration antibiotic therapy, while 151 patients were treated with longer courses. In the RCTs alone, 30 patients received 7‐day courses, while 32 patients received 14‐day courses. In terms of transplant type, 233 patients had undergone abdominal transplants, comprising 139 liver, 91 kidney ± pancreas, 2 combined liver‐kidney, and 1 small bowel transplant. Fourteen patients had thoracic transplants (9 lung and 5 heart). Transplant type was not specified for two patients.

3.2. 90‐Day All‐Cause Mortality

Two RCTs and three cohort studies reported our primary outcome of 90‐day all‐cause mortality. Among the RCTs, mortality was numerically lower in the short‐duration group (0/30 [0%] vs. 2/32 [6.25%]) (Table 2). A meta‐analysis could not be performed using RR due to double‐zero events in one of the RCTs. Instead, RD was used, and the analysis yielded an RD of −0.07 (95% CI: −0.18 to 0.05; p = 0.25), which was not a statistically significant difference (Figure 2).

TABLE 2.

Event rates of primary and secondary outcomes in included studies.

Study	90‐day all‐cause mortality		Clinical failure		30‐day all‐cause mortality		AEs development		AMR Development
	E/S	E/L	E/S	E/L	E/S	E/L	E/S	E/L	E/S	E/L
Mercuro et al. [23]	0/1	0/10	0/1	3/10	0/1	0/10	0/1	3/10	0/1	0/10
Giannella et al. [24]	1/25	3/28	9/25	4/28	1/25	1/28	NR	NR	NR	NR
Yahav et al. [6]	0/25	2/26	3/25	4/26	0/25	1/26	5/25	5/26	4/25	4/26
Ruiz‐Ruigómez et al. [25]	NR	NR	1/6	0/6	2/6	0/6	NR	NR	NR	NR
Molina et al. [7]	0/5	0/6	0/5	2/6	0/5	0/6	4/5	4/6	0/5	0/6
De La Villa et al. [26]	NR	NR	0/9	0/11	1/9	0/11	1/9	1/11	NR	NR
Miwa et al. [27]	0/27	3/64	1/27	3/64	0/27	2/64	0/27	0/64	NR	NR

Abbreviations: E/L, events in the long arm; E/S, events in the short arm; NR, not reported.

FIGURE 2.

Forest plot for 90‐day all‐cause mortality in RCTs.

Among the cohort studies, there was also numerically lower mortality in the short‐duration group compared to the long‐duration group (1/53 [1.89%] vs. 6/102 [5.88%]) (Table 2). The pooled RR demonstrated a lower mortality, albeit not statistically significant, in the short‐duration antibiotic group (RR: 0.36; 95% CI: 0.06–2.08; p = 0.25) (Figure 3).

FIGURE 3.

Forest plot for 90‐day all‐cause mortality in cohort studies.

To further explore the impact of model assumptions, a sensitivity analysis using a fixed‐effect model was conducted and yielded similar results (Figures S1 and S2). Another sensitivity analysis using RD as the effect measure for the cohort studies showed similar results as well (Figure S3).

Subgroup analysis by transplant type showed that outcomes in abdominal transplant recipients were consistent with those observed in the overall SOT population (Figures S4 and S5). However, the small number of thoracic transplant recipients precluded meaningful analysis. Notably, no deaths were observed among thoracic transplant recipients or among the two patients whose transplant type was unspecified.

3.3. Clinical Failure

Among 249 SOT recipients across seven studies, 30 clinical failures were reported. In RCTs, clinical failure occurred in 3/30 (10%) of the short‐treatment group and 6/32 (18.75%) of the long‐treatment group (RR: 0.62; 95% CI: 0.18–2.15; p = 0.45). In cohort studies, failure rates were 11/68 (16.18%) and 10/119 (8.40%) in the short‐ and long‐duration groups, respectively (RR: 1.89; 95% CI: 0.81–4.43; p = 0.14). Neither analysis showed a statistically significant difference (Table 2, Figure 4A,B). These findings remained consistent across multiple sensitivity analyses (Figures S6–S9).

FIGURE 4.

(A) Forest plot for clinical failure among RCTs. (B) Forest plot for clinical failure among cohort studies.

Subgroup analyses were conducted to examine clinical failure outcomes stratified by transplant type. Among abdominal transplant recipients enrolled in the RCTs and the cohort studies, the results reflected the overall analysis (Figures S10 and S11).

Notably, no clinical failure events were reported among the four thoracic transplant recipients included in the RCTs. In contrast, among thoracic transplant recipients in the cohort studies, clinical failure occurred in 4/7 (57.14%) in the short‐duration group, compared to 0/3 (0%) in the long‐duration group with an RR of 2.65 (95% CI: 0.41–17.21; p = 0.31) (Figure S12).

3.4. 30‐Day All‐Cause Mortality

All seven studies reported 30‐day all‐cause mortality (Table 2). Among participants in the RCTs, no deaths were observed in the short‐duration group (0/30 [0%]), while only one death occurred in the long‐duration group (1/32 [3.13%]), with an RD of −0.03 (95% CI: −0.13 to 1.39; p = 0.48) (Figure S13). In the cohort studies, however, a higher number of deaths was reported in the short‐duration arm, with 4 of 68 patients (5.88%) dying compared to 3 of 119 patients (2.52%) in the long‐duration group. This translated to an RR of 1.74 (95% CI: 0.41–7.44; p = 0.45) (Figure S14).

3.5. Development of AEs, Development of AMR, and Length of Hospital Stay

AEs were reported in five of the seven included studies (Table 2), accounting for a total of 23 AEs. In the RCTs, AEs were similarly distributed between the short‐ and long‐duration groups (9/30 [30%] vs. 9/32 [28.13%]) with an RR of 1.15 (95% CI: 0.63–2.01; p = 0.65) (Figure S15). Among the cohort studies, one AE occurred among 37 patients in the short‐duration group (2.70%), while four AEs occurred among 85 patients (4.70%) in the long‐duration group. The corresponding RR was 0.97 (95% CI: 0.16–6.07, p = 0.98), showing no statistically significant difference (Figure S16).

Three studies reported on the development of AMR, however, only one had documented events, thus preventing pooled analysis (Table 2). Data on length of stay specific to SOT recipients were inconsistently reported across the included studies. As a result, this outcome could not be summarized or subjected to comparative analysis.

3.6. Risk of Bias Report

Both RCTs [6, 7] were determined to have a rating of “some concerns” on the ROB2 tool due to the lack of blinding in the second domain of the tool. Regarding the retrospective cohort studies, four of the five studies were deemed to be of good quality according to the AHRQ standards. However, one study, Mercuro et al. [23], was assessed as having poor quality due to limited follow‐up duration, and a substantial proportion of patients lost to follow‐up, which impacted the outcome domain. Detailed evaluations are illustrated in Figures S17 and S18.

3.7. Certainty of Evidence

The certainty of evidence for key outcomes comparing short‐ versus long‐duration antibiotic therapy in SOT recipients with GN‐BSI was assessed using the GRADE framework. When limited to RCTs, the certainty of evidence was moderate. When evaluating evidence derived from the cohort studies, the overall certainty of evidence was notably very low (Tables S3 and S4).

4. Discussion

This is the first systematic review and meta‐analysis to examine the comparative effectiveness of short‐ versus long‐duration antibiotic therapy specifically in SOT patients with GN‐BSI. This unique patient population, characterized by immunosuppression, high antibiotic exposure, and increased risk of MDR organisms, has been notably underrepresented in prior studies evaluating antibiotic duration for GN‐BSI. Although the small number of patients included in this review is a major limitation that reduced statistical power and precision, the synthesized evidence is novel and underscores the importance of including such patients in future research to be able to draw firm conclusions on the safety of shorter antibiotic courses in this high‐risk group.

Across both RCTs and cohort studies included in this review, no statistically significant differences were observed in all‐cause mortality at either 30 or 90 days when comparing short‐ versus long‐duration antibiotic therapy. These findings are generally congruent with the broader GN‐BSI literature in non‐transplant populations, where shorter antibiotic courses have proven non‐inferior in terms of 90‐day mortality, as shown in a recent systematic review and meta‐analysis of four RCTs encompassing more than 3000 patients [28]. In our review, RCTs showed a consistent numerical trend favoring shorter durations at both time points, whereas cohort studies demonstrated inconsistent patterns, with a slight numerical increase in 30‐day mortality in the short‐duration group. While it is difficult to interpret these trends due to the limited sample size and lack of statistical significance, the discrepancy raises critical questions about the appropriateness of short‐term mortality (i.e., 30‐day) as an outcome, particularly in this high‐risk group. A consensus definition of the optimal primary endpoint for RCTs evaluating antibiotic duration was published in 2017 [29] and proposed reporting on 90‐day all‐cause mortality as the primary endpoint for RCTs evaluating antibiotic durations for GN‐BSI.

Another observation in this review is the reversal in the direction of effect for clinical failure between RCTs and cohort studies. In the RCTs, clinical failure occurred less frequently in the short‐duration group, whereas in the cohort studies, the short‐duration group experienced a higher rate of clinical failure. Although these differences were not statistically significant, the divergence is worth highlighting. The observed discrepancies between study designs may reflect biases inherent to observational data, such as confounding and selection bias, as well as variability in mortality drivers overtime. Furthermore, the use of unadjusted data for confounding variables such as severity of illness, presence of MDR pathogens, and timing of source control may have added to these discrepancies. However, it is worth noting that the clinical failure events in the cohort studies were mainly driven by Gianella's study [24], which included a larger proportion of patients (38%) with intra‐abdominal infection as the source of GN‐BSI, this might signal the need for caution in this particular group. These factors underscore the importance of cautious interpretation of observational findings, especially when drawing conclusions about treatment efficacy in high‐risk populations.

The data on adverse events and antimicrobial resistance development were also limited. Although overall adverse events were infrequent and evenly distributed between groups, reporting was inconsistent and often lacked details. The impact of antibiotic duration on resistance development, a central tenet of antimicrobial stewardship, could not be meaningfully assessed due to sparse data and heterogeneous definitions. Future studies must prioritize standardized reporting of such outcomes, particularly in high‐risk populations. However, challenges remain when assessing AMR development in the setting of RCTs as it requires long‐term follow‐up.

Our study has several limitations. Definitions of short versus long durations varied among studies as detailed above, which might have introduced clinical heterogeneity and limited the comparability of pooled estimates. Furthermore, publication bias could not be statistically assessed due to the small number of studies; therefore, the potential for selective reporting remains, particularly given the limited inclusion of transplant‐specific data in the published GN‐BSI studies. Moreover, the lack of adjustment for confounders in the observational studies may have increased their susceptibility to selection and reporting biases. In addition, sparse and inconsistent data on AMR and development of AEs precluded definitive conclusions on these critical outcomes. Finally, the study population consisted predominantly of abdominal transplant recipients, underscoring the limited representation of thoracic transplant recipients and highlighting the need for their inclusion in future studies.

Our results support the ongoing calls for inclusion of SOT populations in clinical trials, as extrapolation from non‐SOT cohorts may not reflect real‐world outcomes in transplant patients. Important clinical factors, such as the degree of immunosuppression and differences between early and late post‐transplant periods, were not addressed in the current literature and should be systematically evaluated. Additionally, other antimicrobial stewardship strategies, such as early oral switch, have not been adequately investigated in the SOT population, limiting the applicability of evidence derived from non‐transplant cohorts. To address these gaps, future research must include adequately powered multicenter trials with standardized definitions and outcomes to inform antimicrobial stewardship strategies tailored to the transplant setting.

5. Conclusion

This systematic review and meta‐analysis aimed to evaluate short‐duration antibiotic therapy for GN‐BSI in SOT recipients in terms of mortality and clinical failure. Although no statistically significant differences were observed between treatment durations in this review, the limited number of included SOT recipients substantially limited statistical power and the ability to draw definitive conclusions. Given the growing body of evidence supporting shorter antibiotic courses in the general population, additional randomized trials in non‐transplant patients are unlikely. However, adequately‐powered, transplant‐specific studies, whether randomized or observational, are needed. Until more robust data are available, adopting short‐course therapy in transplant recipients remains a clinical judgment, requiring careful patient selection and a balance between the risks of undertreatment and those associated with prolonged antibiotic exposure.

Author Contributions

Study concept and design, protocol writing and registration, independent review of titles, abstracts, and full texts, data analysis, manuscript writing: Muath Alotaibi. Protocol writing, independent review of titles, abstracts, and full texts, data analysis, manuscript writing: Abdulellah Almohaya. Data analysis, manuscript writing, critical revision of the manuscript: Roni Bitterman. Manuscript writing, critical revision of the manuscript: José Molina. Literature search, critical revision of the manuscript: Ani Orchanian‐Cheff. Study concept and design, independent third reviewer, critical revision of the manuscript: Shahid Husain.

Funding

The authors have nothing to report.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supporting File 1: tid70209‐sup‐0001‐SuppMat.docx.

Supporting File 2: Visual Abstract.