Introduction

Venous thromboembolism (VTE), encompassing pulmonary embolism (PE) and deep vein thrombosis (DVT), is a significant complication after major surgery, but also in other high-risk conditions such as prolonged immobilization, active malignancy, trauma, obesity, and critical illness requiring intensive care1,2. Thromboprophylaxis typically includes subcutaneous heparin injections or mechanical compression methods, with post-discharge duration recommendations varying by surgery type2 For patients with a history of VTE, recurrence may occur after discontinuing anticoagulant therapy, even following the recommended long-term treatment of at least three months1,2. In such cases, thromboembolic risk often persists beyond the standard period of antithrombotic therapy1,2. Low-dose aspirin could provide a simpler oral option for extended VTE prevention, exceeding the guideline-recommended thromboprophylaxis duration1,2.

However, conflicting data exist in the literature regarding the true risk-benefit profile of aspirin in VTE prevention. A study suggests that low-dose aspirin for secondary prevention can reduce the risk of PE and DVT by approximately one-third during high-risk periods for VTE3. Conversely, another study found that long-term, low-dose aspirin did not prevent VTE in initially healthy women4. While one study indicated that aspirin administered for secondary prevention reduces VTE recurrence without increasing the risk of major bleeding5, another study did not observe a significant reduction in VTE recurrence with aspirin for secondary prevention, though it noted a decrease in major vascular events (e.g., myocardial infarction, stroke, cardiovascular death)6. In contrast, another study reported that perioperative aspirin for primary prevention in noncardiac surgery increased the risk of major bleeding without reducing rates of death or nonfatal myocardial infarction, underscoring the complex risk-benefit profile of aspirin7.

This study aims to quantify the overall effect of aspirin in the extended prevention of VTE, addressing the controversy regarding its risk-benefit profile across various patient conditions, including primary and secondary prevention, as well as provoked and unprovoked VTE. This evidence is crucial for guiding decision-making by weighing the benefits of reducing VTE, mortality—particularly VTE-related mortality—and cardiovascular events against the increased bleeding risks. The study is particularly relevant for extended VTE prophylaxis in high-risk patients and examines whether aspirin use for primary and secondary prevention, provoked and unprovoked VTE can impact outcomes.

Methods

The protocol for this meta-analysis was registered in the PROSPERO database on 26 July 2024, bearing the identification number CRD42024557218. A meta-analysis of available randomised controlled trials (RCTs) was conducted with strict adherence to the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) checklist8.

The inclusion criteria for this systematic review and meta-analysis were based on the PICOS (Population, Intervention, Comparison, Outcomes, Study Design) framework. Briefly, the review focused on adult patients receiving aspirin for VTE prevention, including overall VTE as the primary outcome, with further sub-analyses assessing primary prevention (e.g., surgical patients at risk of VTE) and secondary prevention (e.g., patients with a history of VTE after discontinuation of anticoagulation), compared to those receiving placebo or no treatment.¹⁻⁷ Outcomes assessed included the incidence of provoked3,7 and unprovoked VTE4,5,6,9, major and non-major bleeding5,6,7,10, blood transfusion, cardiovascular adverse events (AEs)6,7,9,10, and mortality, which was analyzed as all-cause mortality, cardiovascular mortality, and VTE-related mortality. Only prospective RCTs in English were eligible. Full details on the PICOS criteria and definitions are provided in the Supplementary Material (SM) 1.

Search strategy

A comprehensive literature search was conducted in PubMed, Scopus, and EMBASE, covering publications up to August 19, 2024, and updated on November 10, 2024. Relevant Medical Subject Headings (MeSH) terms and keywords were combined strategically, using Boolean operators to refine results. Full search details, including term combinations and filters, are provided in SM1. To ensure comprehensive coverage, the reference lists of reviewed studies were also examined to identify any potentially overlooked studies.

Study selection, data extraction and data retrieval

Titles and abstracts of articles identified through the initial search strategy were independently evaluated by two authors (MC, ET) to filter out unrelated studies. The full texts of the remaining studies were then reviewed for compatibility with the established inclusion criteria. Data extraction was also carried out independently by the same two authors using specially designed forms for each study. Any disagreements during the study selection, data extraction, or trial evaluation were resolved by a third author (EC) who was not involved in the initial literature search. Additionally, two other authors (TP, FZ), who had not participated in the initial search or data extraction, conducted a manual review and assessment of each selected study to verify the extracted data and ensure the integrity of the final dataset.

Quality assessment and certainty of evidence assessment

The quality assessment of the included RCTs was independently conducted by two authors (MC, ET) using the Risk of Bias 2 (RoB 2) tool11. This tool evaluates potential bias across five critical domains: the randomisation process, deviations from intended interventions, missing outcome data, measurement of outcomes, and selection of reported results. Each domain is assessed using “signaling questions” designed to identify potential sources of bias. Responses to these questions are used in an algorithm to assign a risk of bias rating for each domain, categorized as “low,” “high,” or “some concerns.” An overall risk of bias assessment is then provided for each study. Disagreements in initial evaluations were resolved through consultation with a third author (EC).

The certainty of evidence was evaluated using the Grades of Recommendation, Assessment, Development, and Evaluation (GRADE) framework for meta-analyses12. Quality of Evidence (QoE) is classified into four levels: high (⊕⊕⊕⊕), moderate (⊕⊕⊕⊖), low (⊕⊕⊖⊖), or very low (⊕⊖⊖⊖). Initially rated as high due to their RCT design, evidence quality may be downgraded based on factors such as risk of bias (e.g., inadequate blinding or allocation concealment), inconsistency (assessed by variance in effect estimates using statistical measures like tau [τ], tau-squared [τ2], and I-squared [I2), indirectness (e.g., differences in study populations, interventions, or outcomes from those of primary interest), imprecision (indicated by wide 95% confidence intervals or estimates near the null effect), and publication bias (e.g., assessed through funnel plots, leave-one-out diagnostics, and Egger’s test for asymmetry).

Statistical analysis

The meta-analysis was conducted using a frequentist framework, applying both random effects and fixed effects models to calculate the relative risk (RR) and corresponding 95% confidence intervals (CI) for binary outcome data. The random effects model was preferred in this meta-analysis to account for potential variability across studies. The Mantel–Haenszel method was used to compute the fixed effects estimate for dichotomous data. For the random effects model, inverse-variance weighting was applied using the DerSimonian and Laird method to handle heterogeneity. In cases where studies reported zero events, a continuity correction of 0.5 was added to frequencies when computing the RR.

Heterogeneity was assessed using the I2 statistic, with significance interpreted at p < 0.1 for the Q test. I2 values were categorized as low (< 25%), moderate (25-50%), or high (> 50%)13. To further quantify heterogeneity, τ was computed to assess the standard deviation of effect sizes across studies, indicating variability beyond chance. Additionally, τ2 estimated the variance of true effect sizes between studies, though precise estimation was challenging due to the limited number of studies. The Q test confirmed significant heterogeneity, providing further insight into the variability of the meta-analysis results. Publication bias was assessed through visual inspection of funnel plots14. Outliers identified from leave-one-out diagnostics, such as Cook’s distances, were evaluated to determine influential studies. Egger’s test for asymmetry was performed only for analyses with 10 or more studies; a p-value < 0.1 indicated potential publication bias, while a p-value ≥ 0.1 suggested negligible risk.

Trial Sequential Analysis (TSA) was conducted to evaluate the stability and reliability of cumulative meta-analysis results on the effect of aspirin versus placebo or no treatment in preventing VTE. This analysis focused on general, primary and secondary prevention, provoked and unprovoked VTE outcomes, including low-dose aspirin. TSA helps control for random errors by applying monitoring boundaries and assessing type I and II errors15. The proportion of events in the control group, the proportion of events in the aspirin group, the highest observed relative risk reduction (RRR), and the minimum information size (MIS) from the literature3,4,5,6,7 were used to obtain the Trial Sequential Monitoring Boundary (TSMB) and Diversity Adjusted Required Information Size (DARIS). Further details are provided in SM1.

Sensitivity analyses were conducted to evaluate the robustness of results across different outcome types (primary and secondary VTE, provoked and unprovoked VTE) and aspirin dosages (low-dose).

The Number Needed to Treat for Benefit (NNTB) and the Number Needed to Treat for Harm (NNTH) were calculated to assess the net clinical benefit of aspirin in VTE prevention. NNTB was derived from the absolute risk reduction (ARR) for thrombotic outcomes, and NNTH from the absolute risk increase (ARI) for bleeding events. The NNTB/NNTH ratio was also computed to evaluate the balance between efficacy and safety.

All statistical analyses were conducted using R software version 4.1.0 (2021-05-18). Meta-analytical computations were performed using the “meta” package, including the calculation of the NNTB and NNTH using the specific “nnt()” function. Trial Sequential Analysis (TSA) computations utilized the “ldbounds” and “rpact” libraries. P-values were two-tailed, with statistical significance set at < 0.05 in accordance with standard convention.

Results

Paper selection

Of the 5527 reports initially identified, 5522 were excluded, leaving 5 RCTs with a total of 68,554 patients for inclusion.³⁻⁷ The PRISMA flow diagram is in Fig. 1.

Fig. 1
figure 1

The PRISMA flow diagram of the study selection process.

Full size image

Study characteristics

Details of the included RCTs are in Table 13,4,5,6,7. Among participants, 34,274 were assigned to treatment (aspirin), and 34,280 to control (placebo)3,4,5,6,7. Approximately 96.6% of patients were in primary prevention for overall VTE3,4,7, while only 40% were in primary prevention for DVT and PE3,7. Conversely, approximately 3.4% of patients were in secondary prevention for overall VTE4,5,6, while only 1.8% were in secondary prevention for DVT and PE3,7. Overall, 59.9% of patients received low-dose aspirin (100 mg) for VTE prevention4,5,6,7. Regarding VTE type, 25.4% of patients were treated for provoked VTE3,7, and 40% for unprovoked VTE5,6. Allocation between aspirin and placebo groups was nearly equal across all subgroups.

Table 1 Characteristics of studies considered for review and meta-analysis.
Full size table

Risk of bias assessment

The RoB 2 assessment conducted on the included RCTs3,4,5,6,7 indicated a low risk of bias across all domains (Fig. 2), with well-defined randomisation, allocation concealment, and blinding. No deviations or selective reporting were observed. Overall, the risk of bias across all studies was low, confirming their strong methodological quality. Further details on the RoB 2 assessment can be found in SM2.

Fig. 2
figure 2

Risk of bias (RoB) 2 assessment. RoB2 traffic light (A) and RoB2 summary (B) of included RCTs.

Full size image

Endpoints

The primary endpoint of this meta-analysis was the incidence of newly diagnosed symptomatic VTE, including both DVT and PE. Secondary endpoints included all-cause mortality, cardiovascular mortality, VTE-related mortality, major bleeding, and other safety outcomes such as stroke, hemorrhagic stroke, and major cardiovascular adverse events. These endpoints were pre-specified according to the study protocol and align with international guideline recommendations.

Primary endpoint

Aspirin (100–160 mg) significantly reduced VTE (RR [95%CI] = 0.80 [0.72; 0.90], P < 0.001; moderate QoE), DVT (RR [95%CI] = 0.82 [0.71; 0.95], P = 0.010; moderate QoE), and PE (RR [95%CI] = 0.79 [0.66; 0.95], P = 0.014; moderate QoE) compared to placebo (Table 2; Fig. 3)3,4,5,6,7. In the sub-analysis, aspirin significantly reduced the incidence of primary VTE, DVT, and PE compared to placebo (Table 3)3,4,7, but showed no significant benefit in preventing secondary VTE, DVT, or PE (Table 4)4,5,6.

Table 2 Efficacy and safety of aspirin compared to placebo in preventing VTE.
Full size table
Fig. 3
figure 3

Summary of main results. Aspirin reduces the risk of venous thromboembolism (VTE) by 20%, deep vein thrombosis (DVT) by 18%, pulmonary embolism (PE) by 21%, and VTE-related mortality by 56% compared with placebo. Illustration elements adapted from Depositphotos™ (licensed image provider), used under extended license permitting academic publication.

Full size image
Table 3 Efficacy and safety of aspirin compared to placebo in preventing primary VTE.
Full size table
Table 4 Efficacy and safety of aspirin compared to placebo in preventing secondary VTE.
Full size table

Aspirin was effective in reducing provoked VTE, including both its components, DVT and PE, compared to placebo3,7, and also showed a significant benefit for unprovoked VTE overall, although not for its components (DVT and PE) (Tables S1-S2 in SM3)5,6. Low-dose aspirin (100 mg) did not significantly reduce VTE, DVT, or PE compared to placebo (Table S3 in SM3)4,5,6,7.

Secondary endpoints

Aspirin (100–160 mg) did not significantly affect all-cause or cardiovascular mortality (Table 2)3,5,6,7, but significantly reduced VTE-related mortality (RR [95%CI] = 0.44 [0.26; 0.75], P = 0.002; moderate QoE) compared to placebo (Table 2)3,5,6. This reduction was consistent for primary prevention and provoked VTE (RR [95%CI] = 0.42 [0.24; 0.72], P = 0.006; moderate QoE) (Table 3 and Table S1 in the SM3)3,7, but not for secondary prevention (Table 4)5,6, unprovoked VTE (Table S2 in the SM3)5,6 or low-dose low-dose aspirin (100 mg) in VTE prevention (Tables S3 in the SM3)5,6,7.

Aspirin (100–160 mg) was associated with increased overall bleeding (RR [95%CI] = 1.13 [1.10; 1.16], P < 0.0001; moderate QoE) and major bleeding (RR [95%CI] = 1.18 [1.07; 1.30], P < 0.0001; moderate QoE)3,4,5,6,7, without increasing transfusion rates, compared to placebo (Table 2)3,4. Sub-analyses were consistent across primary prevention3,7 and provoked VTE3,7, but not for secondary prevention and unprovoked VTE (Tables 3 and 4; Tables S1-S2 in SM3)5,6. Low-dose aspirin also increased bleeding risks in the context of VTE prevention compared to placebo (Table S3 in SM3)4,5,6,7.

Aspirin did not significantly increase stroke (RR [95%CI] = 1.10 [0.86; 1.41], P = 0.417; moderate QoE)3,4,5,6,7 and, specifically, haemorrhagic stroke (RR [95%CI] = 0.95 [0.26; 3.38], P = 0.449; low QoE)3,4, and did not reduce major cardiovascular AEs (RR [95%CI] = 0.92 [0.83; 1.03], P = 0.162; low QoE)3,4,5,6,7, cardiac AEs (RR [95%CI] = 1.03 [0.79; 1.35], P = 0.773; low QoE)3,5,6,7, or myocardial infarction (RR [95%CI] = 1.03 [0.86; 1.24], P = 0.700; low QoE) compared to placebo (Table 2)3,5,6,7. Findings were consistent in primary3,7 and secondary prevention (Tables 3 and 4)5,6, provoked3,7 and unprovoked5,6 VTE (Tables S1-S3 in SM3), and with low-dose aspirin (100 mg) used in VTE prevention (Tables S3 in the SM3)4,5,6,7. However, for secondary prevention and unprovoked VTE, aspirin was linked to lower major cardiovascular AEs (RR [95%CI] = 0.71 [0.57; 0.90], P = 0.004; moderate QoE) compared to placebo (Table 4)5,6.

Forest plots are provided in SM4, while funnel plots and leave-one-out diagnostic charts identifying highly influential studies are in SM5. These plots assess aspirin’s efficacy and safety compared to placebo in VTE prevention. The funnel plots indicate no significant publication bias, while the leave-one-out diagnostics confirm that the exclusion of influential studies does not substantially alter the overall results, supporting the robustness of the meta-analysis. The Quality of Evidence (QoE) for the outcomes is summarised in Tables 2, 3 and 4 and in SM3 [Tables S1–S3]).

In the TSA, cumulative z-scores crossed the monitoring boundaries for several VTE types, providing firm evidence of aspirin’s benefit in general, primary, provoked, and secondary VTE prevention (overall outcome), while evidence for unprovoked VTE and low-dose aspirin (100 mg) were conclusive but not firm (Table 5 and SM3 [Table S4]). Cumulative z-scores for DVT, PE, and VTE-related mortality indicated firm or conclusive evidence for general and primary prevention, and for provoked VTE. In contrast, evidence was inconclusive for some outcomes in secondary and unprovoked VTE, and in low-dose aspirin treatment (Table 5 and SM3 [Table S4]). TSA graphs supporting these findings are provided in SM6.

Table 5 Trial sequence analysis to detect relative risk reduction in VTE, DVT, PE and VTE-related mortality, with α = 5% and power = 80%.
Full size table

The NNTB/NNTH ratio analysis across low, moderate, and high baseline risk levels confirms a favorable benefit–risk profile of extended aspirin therapy in VTE prevention. This finding was consistent in subgroup analyses for primary and provoked VTE prevention, particularly in higher-risk populations, while the benefit was limited or uncertain in secondary, unprovoked, and low-dose aspirin settings (SM7).

Discussion

Our meta-analysis assessed the overall role of aspirin in VTE prevention, with additional analyses for primary and secondary prevention. Due to the limited number of RCTs, we conducted a pooled analysis but performed subgroup analyses where feasible to distinguish provoked from unprovoked VTE. Our meta-analysis demonstrates that extended aspirin therapy (100 to 160 mg) significantly reduces the risk of VTE, DVT, and PE compared to placebo, especially in primary prevention in surgical patients. Aspirin is effective in reducing the risk of provoked VTE, including both DVT and PE, and also shows a significant benefit for unprovoked VTE overall, although not for its individual components. Low-dose aspirin (100 mg) does not significantly reduce the risk of VTE, DVT, or PE. While aspirin does not impact overall or cardiovascular mortality, it significantly reduces VTE-related mortality. Aspirin use at any dose increases the risk of bleeding but does not elevate blood transfusion requirements. These effects were consistent for primary prevention and provoked VTE but not for secondary prevention, unprovoked VTE, or low-dose aspirin use. Importantly, aspirin does not increase the risk of stroke or haemorrhagic stroke, nor does it reduce major cardiovascular events, except in secondary prevention and unprovoked VTE, where a significant reduction in major cardiovascular events was observed.

While the POISE-2 trial found no significant reduction in VTE for primary prevention in noncardiac surgery7, pooled analyses indicate a one-third reduction in symptomatic provoked VTE in noncardiac surgery16. Extending aspirin (100–160 mg) after stopping antithrombotic prophylaxis may be appropriate for patients with risk factors (e.g., history of VTE, thrombophilia, obesity, malignancy), particularly for those unable or unwilling to use anticoagulants2, which remain superior in reducing VTE occurrence, making aspirin a secondary option in such cases1,2 Postoperative VTE events in high-risk surgical patients17, such as those with obesity18, often occur weeks after surgery17,18, typically after stopping prophylaxis18, underscoring the potential benefit of aspirin use beyond the initial short-term prophylaxis period (4–14 days) or even the extended antithrombotic prophylaxis phase (19–42 days)2 Aspirin’s low cost and convenient oral administration make it an appealing option, especially as early ambulation, while crucial, is insufficient to prevent VTE,19 necessitating a careful balance between the benefits of aspirin and its associated bleeding risks2. Recent multicenter trial evidence shows aspirin prescribed at discharge is noninferior to low-molecular-weight heparin in preventing mortality and major thromboembolic events in extremity fractures, highlighting its role in thromboprophylaxis20. Previous studies reported no significant difference in overall mortality between extended- and standard-course antithrombotic prophylaxis2. Our meta-analysis demonstrated that prolonged aspirin, administered beyond the 19–42 days recommended by guidelines after major surgery2, significantly reduced VTE, DVT, PE, and VTE-related mortality in primary prevention and provoked VTE. However, the duration of aspirin therapy varied across studies. In the PEP Trial, aspirin was administered for 35 days, but most VTE events were recorded within the first 16 days of hospitalization3. In the POISE-2 Trial, aspirin was given for 30 days7. While these findings support aspirin use beyond the initial prophylaxis period, further studies are needed to assess its efficacy beyond 42 days in primary prevention.

Regarding secondary prevention, the INSPIRE analysis showed a 42% reduction in recurrent VTE risk after adjusting for treatment adherence9 However, our findings did not confirm a statistically significant benefit of aspirin in reducing overall VTE risk in secondary prevention, although some effect cannot be excluded due to wide confidence intervals. This is consistent with guidelines that consider aspirin as an option for patients who refuse or cannot tolerate oral anticoagulants for extended prophylaxis21 Aspirin use should also be reconsidered when anticoagulant therapy is discontinued, as it may have been stopped when anticoagulation began, requiring a balance between its benefits and bleeding risks1,2. In line with our results, aspirin showed a significant benefit in preventing unprovoked VTE overall, although not for its individual components (DVT and PE), suggesting that aspirin may have a limited role in this setting. Nevertheless, anticoagulation remains the standard of care for secondary prevention2.

Consistent with other meta-analyses20,22,23,24, we found no benefits in reducing overall or cardiovascular mortality, which could have meaningful implications for clinical practice. In the context of VTE prophylaxis, aspirin may offer survival benefits, particularly for patients at high VTE risk20,22,23,24, warranting its careful consideration despite the associated bleeding risks. Although caution is advised for healthy individuals without atherosclerosis—where aspirin may slightly reduce myocardial infarction risk but increase major bleeding23,—this finding underscores its potential value in targeted high-risk populations.

While effective in reducing VTE, aspirin’s bleeding risk is a concern. The INSPIRE analysis showed no significant bleeding difference between aspirin and control groups,9 but other evidence suggests caution. A pooled analysis of two RCTs found an 18% higher bleeding risk with aspirin in non-cardiac surgery patients,16 and recent meta-analyses for primary prevention of cardiovascular events show a 40% increase in major bleeding risk22 Our meta-analysis identified a 13% increase in major bleeding with aspirin, though not linked to stroke, specifically haemorrhagic stroke, nor to all-cause or cardiovascular mortality. Nevertheless, caution around intracranial haemorrhage remains, with meta-analyses reporting a 30%22, 33%23, or 34% increased risk of haemorrhagic stroke24 These findings underscore the importance of balancing the benefits of VTE prevention with the associated bleeding risks. The NNTB/NNTH ratio supports a favorable benefit–risk profile, particularly in high-risk patients undergoing primary prevention or with provoked VTE. Nonetheless, careful patient selection remains essential.

Comprehensive data on aspirin’s impact on major cardiovascular events indicate potential benefits in reducing cardiovascular incidents by 9%22 or 11%24, with a decrease in myocardial infarction by 13%20, 15%24, or 18%23. Additionally, aspirin has been associated with a 6%23 or 12%22 decrease in ischaemic stroke. While data from individual RCTs suggest potential benefits3,7, our analysis found that aspirin reduced major cardiovascular adverse events only in the context of secondary prevention and unprovoked VTE. A pooled analysis indicated a substantial 34% reduction in these events with aspirin9, and our findings showed a 29% reduction. However, no significant benefit was observed in primary prevention and provoked VTE. Furthermore, within the context of VTE prevention, aspirin did not demonstrate any effect in reducing the risk of cardiac adverse events or myocardial infarction across any of the conditions explored. While our analysis did not show significant cardiovascular benefits of low-dose aspirin (100 mg/day) in VTE prevention, evidence from sensitivity analyses indicates that aspirin at doses of 100 mg or less daily is associated with reductions in composite cardiovascular outcomes, including a 13% reduction in ischemic stroke, an 11% reduction in total stroke, and a 13% reduction in myocardial infarction24. These benefits appear more pronounced in specific populations, such as individuals under 65 years of age and those with a BMI ≥ 25, with similar bleeding risks22,24. Additionally, patients with a low 10-year MACE% risk saw more cardiovascular benefits from aspirin use than those at high risk22.

Strengths and limitations of the study

By conducting multiple sensitivity analyses, this meta-analysis tested the robustness of the results, reinforcing their clinical relevance. A comprehensive literature review contributed to a robust dataset, and focusing exclusively on adult RCTs improved the applicability of the findings to this specific population.

However, the study has limitations. It did not assess varying aspirin dosages or long-term adherence, which could impact the risk-benefit balance. Important factors such as comorbidities and postoperative care variations were inconsistently reported, possibly influencing outcome interpretation. While major bleeding definitions align with existing literature10, minor bleeding requires broader criteria to meet recent definitions of clinically relevant non-major bleeding25. Many RCTs predate this consensus25, limiting exact conformity. Additionally, only one study reported gastrointestinal bleeding outcomes4, constraining the assessment of GI bleeding risks with aspirin for VTE prevention.

Although statistical heterogeneity was acceptable, clinical heterogeneity remains a limitation of our study. Variability in follow-up duration, aspirin regimens, and clinical settings among the included trials may affect the generalizability of our findings. Nevertheless, the consistency of results across studies supports the overall robustness of our conclusions.

Higher RRR thresholds in TSA, based on the maximum reported values, ensure robustness but may overlook some significant results. Lower RRR thresholds could provide a more nuanced view of aspirin’s effects.

The NNTB and NNTH estimates are based on aggregated trial data and may not fully capture differences across patient subgroups. Further studies are needed to refine patient selection criteria.

Excluding non-RCT studies and grey literature may risk publication bias despite the benefits of focusing on peer-reviewed research. Current guidelines recommend DOACs for prolonged VTE treatment. Lastly, our meta-analysis aimed to address a gap in the literature by evaluating aspirin as an alternative in patients unable or unwilling to take anticoagulants. Since this was not a network meta-analysis, direct comparisons with DOACs were not included.

Conclusion

Our meta-analysis suggests that extended aspirin therapy (compared to placebo) may reduce the risk of VTE, DVT, and PE, particularly in primary prevention and provoked VTE, and possibly in unprovoked VTE overall, while also highlighting an increased risk of bleeding. These findings support aspirin’s potential role in thromboprophylaxis as an alternative in selected patients, especially when anticoagulation is contraindicated or refused. However, the heightened bleeding risk underscores the importance of careful patient selection and monitoring. Future studies should further explore the risk–benefit balance and optimal patient selection.