Effect of anticoagulation on the incidence of venous thromboembolism, major bleeding, and mortality among hospitalized COVID-19 patients: an updated meta-analysis

Objective Anticoagulation is crucial for patients hospitalized with coronavirus disease 2019 (COVID-19) due to the high risk of venous thromboembolism (VTE). However, the optimal anticoagulation regimen needs further exploration. Therefore, we evaluated the efficacy and safety of diverse anticoagulation dosage dosages for COVID-19. Methods An updated meta-analysis was performed to assess the effect of thromboprophylaxis (standard, intermediate, and therapeutic dose) on the incidence of VTE, mortality and major bleeding among COVID-19 patients. Literature was searched via PubMed, EMBASE, Web of Science, and China National Knowledge Infrastructure (CNKI) database. The odds ratio (OR) and 95% confidence interval (CI) were calculated for effect estimates. Results Nineteen studies involving 25,289 participants without VTE history were included. The mean age of patients was 59.3 years old. About 50.96% were admitted to the intensive care unit. In the pooled analysis, both therapeutic-dose and intermediate-dose anticoagulation did not have a significant advantage in reducing VTE risk over standard dosage (OR = 1.09, 95% CI: 0.58–2.02, and OR = 0.89, 95% CI: 0.70–1.12, respectively). Similarly, all-cause mortality was not further decreased in either therapeutic-dose group (OR = 1.12, 95% CI: 0.75–1.67) or intermediate-dose group (OR = 1.34, 95% CI: 0.83–2.17). While the major bleeding risk was significantly elevated in the therapeutic-dose group (OR = 2.59, 95%CI: 1.87–3.57) as compared with the standard-dose regimen. Compared with intermediate dosage, therapeutic anticoagulation did not reduce consequent VTE risk (OR = 0.85, 95% CI: 0.52–1.38) and all-cause mortality (OR = 0.84, 95% CI: 0.60–1.17), but significantly increased major bleeding rate (OR = 2.42, 95% CI: 1.58–3.70). In subgroup analysis of patients older than 65 years, therapeutic anticoagulation significantly lowered the incidence of VTE in comparation comparison with standard thromboprophylaxis, however, at the cost of elevated risk of major bleeding. Conclusion Our results indicated that for most hospitalized patients with COVID-19, standard-dose prophylactic anticoagulation might be the optimal choice. For elderly patients at low risk of bleeding, therapeutic-dose anticoagulation could further reduce VTE risk and should be considered especially when there were other strong risk factors of VTE during hospital stay. Systematic Review Registration https://www.crd.york.ac.uk/PROSPERO, identifier, CRD42023388429.


Introduction
Since December 2019, the coronavirus disease 2019 (COVID-19) pandemic, caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), has led to large-scale human transmission and caused hundreds of thousands deaths around the world (1).Due to complexity and heterogenous severity of COVID-19, large difficulties and challenges in disease management have been brought by its complications during clinical practice, among which venous thromboembolism (VTE) deserves more attention being paid to because of potential fatal events, especially in early pandemic era (2).VTE includes deep vein thrombosis (DVT) and pulmonary embolism (PE).As is known, PE is caused by an obstruction of the pulmonary arteries, most often occluded by thrombus derived from DVT of the lower extremities, and its major symptoms include dyspnea, chest pain, syncope, hemoptysis, etc.
(3) and dyspnea.In patients with COVID-19, significant abnormalities in coagulation function have been reported (4).In addition to this, vascular wall injuries, blood stream stasis, and hypercoagulable state in hospitalized COVID-19 patients increases the risk of VTE (5)(6)(7).Unpredictable deterioration and even sudden death may occur in some COVID-19 patients due to secondary VTE event during disease management (8).Thus, early recognition of risk factors and appropriate thromboprophylaxis of VTE in patients with COVID-19 are crucial for lowering in-hospital mortality and may to improve long-term prognosis.
Growing evidence has shown that prophylactic anticoagulation can effectively reduce the incidence of VTE and mortality rate in hospitalized COVID-19 patients with COVID-19, especially in critically ill patients, although at price of increased risk of bleeding (9)(10)(11).However, various dosages of prophylactic anticoagulation are used in practice to balance clinical benefit and bleeding risk.Still, no valid consensus has been reached regarding optimal anticoagulation dosage for VTE prevention in COVID-19 patients to achieve best efficacy and less hemorrhage event (12)(13)(14)(15)(16).
Although previous meta-analysis has addressed this issue, emerging new studies with diverse outcomes have been published later, and various virus strain of SARS-CoV-2 has evolved which may possess different impact on VTE risk.Therefore, we conducted this updated meta-analysis to evaluate the efficacy and safety of different prophylactic anticoagulation regimen (standard dose, intermediate dose, and therapeutic dose) on the incidence of VTE, major bleeding, and mortality, to obtain better and more detailed evidence on VTE prophylaxis for hospitalized patients with COVID-19.

Design
Low molecular weight heparins are most frequently used for thromboprophylaxis in COVID-19.Therefore, in this meta-analysis, we assessed three conventional prophylactic anticoagulation regimen with low molecular weight heparins (shown in Table 1) on the incidence of VTE, major bleeding, and mortality among COVID-19 patients.This systematic review and meta-analysis was reported in accordance with the Cochrane Handbook (17) and the guidance from Preferred Reporting Items of Systematic Reviews and Meta-Analyses (PRISMA) checklist (18).The protocol of this study has been registered on the International Prospective Register of Systematic Reviews (PROSPERO, https://www.crd.york.ac.uk/PROSPERO) with registration number of CRD42023388429.

Search strategy
We designed a high-sensitivity search strategy that combined the following search items: free-text and keyword synonyms of COVID-19 and VTE, and word clusters of prophylactic anticoagulation.Literature was searched through PubMed, EMBASE, Web of Science, and China National Knowledge Infrastructure (CNKI) database.We further searched with the keywords "standard-dose prophylactic anticoagulation", "intermediate-dose prophylactic anticoagulation", and "therapeutic-dose prophylactic anticoagulation" on bioRxiv (http://www.biorxiv.org)server, medRxiv (http://www.biorxiv.org)server and Chinaxiv (http://biotech.chinaxiv.org)server, in order to identify potential pre-publication manuscripts that met the eligibility criteria.The search spanned from January 1, 2020to October 31, 2022.The reference lists of all included articles were also reviewed for potential eligible studies.

Study selection and data extraction
Two reviewers independently performed a two-step selection, screening by title and abstract, followed by a full-text review.Studies would be included if they met the following criteria: (1) they were randomized controlled trial, observational cohort, or case-control study; (2) they enrolled hospitalized COVID-19 patients without VTE at baseline who did not receive anticoagulation in the past six months; (3) outcomes of interest were compared among patients receiving standard-dose prophylactic anticoagulation, intermediate-dose prophylactic anticoagulation and therapeutic-dose prophylactic anticoagulation; (4) outcomes of interest included one of the followings: event of VTE, major bleeding, and mortality.
Data extraction was conducted using standardized data extraction forms.The following information were collected from the retrieved literature: the first author's name, publication year, study design, research site, patient characteristics (including age, gender, and disease severity), follow-up period, incidence of VTE, major bleeding, and mortality rate.Discrepancies were solved by discussion.

Statistical analysis
Effects of 3 different dosing prophylactic anticoagulation on the incidence of VTE, major bleeding, and mortality of COVID-19 patients were presented or calculated as odds ratio (OR), relative risk (RR), or hazard ratio (HR), with 95% confidence interval (CI) from included studies.We pooled ORs across studies using inverse-variance weighted DerSimonian-Laird method to calculate effect estimate.RR and HR were considered as equivalent as OR during meta-analysis.Continuous variables were calculated as weighted mean difference (WMD) and 95% CI.The median value and interquartile range (IQR) provided from original studies were converted to mean and standard deviation (SD) according to the method by Wan et al. (19).Between-study heterogeneity was tested by Cochrane Q and I 2 statistic.I 2 > 50% or P < 0.1 was considered as significant heterogeneity and random effects model was used to combine the results.Otherwise, fixed effects model was used (17).Funnel plots and sensitivity analysis were then conducted to examine the publication bias and stability of meta-analysis result, respectively.All statistical analysis process were conducted using Review Manager 5.4 (Cochrane Collaboration, Oxford, UK) and STATA 14.0 (Stata Corporation, College Station, TX, U.S).

Literature quality evaluation
The methodological quality of included articles was evaluated using the Newcastle-Ottawa Scale (NOS), available at: https://www.ohri.ca/programs/CIinical_epidemiology/oxford.asp.The total score of NOS rangeed from 0 to 9 stars, with more stars representing higher quality.Two authors independently went through this scoring process, and discrepancies were solved by discussion.

Network meta-analysis
So far, current studies mainly compared the efficacy and safety between intermediate-dose and standard-dose anticoagulation, or between therapeutic-dose and standard-dose anticoagulation for COVID-19 patients.Fewer studies [Jonmarker et al. (20) and Blondon et al. (21)] investigated the difference between intermediatedose with therapeutic-dose prophylactic anticoagulation, making it less convincing to perform traditional meta-analysis.Thus, we chose network meta-analysis and defined standard-dose anticoagulation as plan A (plan 1 in the rank), intermediate-dose anticoagulation as plan B (plan 2 in the rank), and therapeutic-dose anticoagulation as plan C (plan 3 in the rank).The network metaanalysis was conducted using "mvmeta" package and "network" package of STATA 14.0 software.
Because the I 2 of some forest plots was greater than 50%, we continued to conduct funnel plots (Figure 4) and sensitivity analysis (Figures 5, 6   In addition, we also developed NOS for the evaluation of literature quality (Table 3), indicating that the selected articles were of good quality.
Based on the above analysis, we further con ducted a subgroup analysis to identify the source of heterogeneity in terms of elders (65 years), gender, study duration, study design, and ICU admission rate.See Table 4 for the results of the subgroup analysis.
The main purpose of network meta-analysis was to compare the difference between therapeutic-dose and intermediate-dose prophylactic anticoagulation.Further, ranking and surface under the cumulative ranking (SUCRA) probabilities were performed to carry out the recommended order of the three doses after evaluation of the incidence of VTE, major bleeding and mortality.
As shown in Figure 7, this network meta-analysis had a closedloop structure, so its results were to merge the direct and indirect comparisons and make decisions accordingly.
The adjusted funnel plot in Figure 9 pointed out no evidence of publication bias in our included articles.
Furthermore, we ranked the impact of three doses of prophylactic anticoagulation on VTE, major bleeding, and mortality in patients with COVID-19.Ranking and SUCRA were shown in Figures 10, 11, separately.

Discussion
SARS-CoV-2 infection can not only cause multiple organ damage (38), but greatly increase the risk of VTE.As early as the begin ning of the COVID-19 pandemic in 2020, studies from Wuhan, China, initially revealed that COVID-19 patients had a high risk of VTE (39), which was gradually confirmed with the outbreak all over the world.Marchandot et al. (40) summarized studies on hospitalized COVID-19 patients from different countries and found that the incidence of VTE in non-ICU and ICU patients was 3%-46% and 15%-85%, separately.Nopp et al.   42) analyzed 3 ho spitals in the Netherlands and reported that the incidence of VTE in hospitalized COVID-19 patients was 18.7%, while that in hospitalized patients with influenza from 2013 to 2018 was only 1.04%.Although the incidence of VTE varies from study to study, there is a consensus that the risk of VTE remains higher in COVID-19 patients, and the more severe the disease, the higher the risk (43).
On the other hand, the prognosis of COVID-19 patients tends to be worse if VTE occurs.A study (44) from Wuhan, China, enrolled 143 COVID-19 cases in the ICU and noted that compared with patients without DVT, the mortality of those who had comorbid DVT was significantly higher (34.8% vs.11.7%, P = 0.001).Kollias A et al. (45) developed a meta-analysis of more than 6,000 patients and revealed that the incidence of PE and DVT in hospitalized patients with COVID-19 was 32% and 27%, respectively, and the risk of death was twice higher if VTE was accompanied.Even if VTE is not the direct cause of death, it may be an important cause.Lax et al. (46) from Australia analyzed an autopsy study on 11 patients who died of COVID-19 and proved that all patients This meta-analysis included 19 studies published between January 1, 2020, and October 31, 2022.To our knowledge, this is the first meta-analysis conducting a p airwise comparison among three conventional prophylactic anticoagulations in the incidence of VTE, major bleeding, and mortality.This meta-analysis included 19 related studies with 25,289 COVID-19 patients, and the results showed that: (1) compared with standard-dose prophylactic anticoagulation, odds ratio (OR) and 95% confidence interval (CI) of VTE, major bleeding and mortality in therapeutic-dose prophylactic anticoagulation was 1.09 (95% CI: 0.58, 2.02), 2.59 (95% CI: 1.87, 3.57) and 1.12 (95% CI: 0.75, 1.67), respectively; (2) compared with standard-dose prophylactic anticoagulation, OR and 95%CI of VTE, major bleeding and mortality in intermediate-dose prophylactic anticoagulation was 0.89 (95% CI: 0.70, 1.12), 1.15 (95% CI: 0.94, 1.40) and 1.34 (95% CI: 0.83, 2.17), respectively; (3) compared with intermediate-dose prophylactic anticoagulation, OR and 95%CI of VTE, major bleeding and mortality in therapeutic-dose prophylactic anticoagulation was0.85 (95% CI: 0.52, 1.38), 2.42 (95% CI: 1.58, 3.70) and 0.84 (95% CI: 0.60, 1.17).The above results suggested that compared with COVID-19 patients receiving intermediate-dose or therapeutic-dose prophylactic anticoagulation, those who underwent standard-dose prophylactic anticoagulation had the lowest risk of bleeding events.In terms of VTE and mortality, no significant differences were found.
We further ranked the impact of the three doses of anticoagulation on VTE, major bleeding, and mortality in patients with COVID-19.Combining the results of ranking (Figure 10) and SUCRA (Figure 11), the order of probability of VTE events from high to low was: therapeutic-dose >> standard-dose > intermediate-dose.The    To verify the applicability of the above results, we conducted a subgroup analysis in terms of age, ICU admission rate, hospital stay, etc.It is worth noting that compared with the standard dose, although therapeutic-dose prophylactic anticoagulation increased the risk of major bleeding, it could significantly reduce VTE formation in patients over 65 years of age.
To sum up, consistent with ISTH guidelines and ACCP guidelines (49), we recommended a standard-dose rather than an above-standard dose (i.e., intermediate-dose or therapeutic-dose)   (56).Based on this score, the risk of VTE in COVID-19 patients can be further evaluated and guide the application of clinical anticoagulation programs.Therefore, it is scientific and reasonable to provide standard-dose prophylactic anticoagulation for all hospitalized patients in a timely manner, to increase the dose individually for elderly patients with a high risk of VTE or acceptable risk of bleeding, as well as to adjust the dose according to the patient's weight and the disease progression.It is expected that there will be a higher level of evidence to verify our conclusion in the future.There are some limitations in this study.First, the prevalence of thromboembolism in COVID-19 patients was likely to be underestimated.The possible reason was that the incidence of thrombotic events (e.g., PE, DVT, myocardial infarction, ischemic stroke, and other thromboembolism) diagnosed with routine clinical care was often less than that seen on computed   studies are required to further explore the safety and effectiveness of different treatments, so as to guide clinical practice and improve the disease status.Third, we encountered high statistical heterogeneity during the meta-analysis.Despite we conducted a prespecified sensitivity analysis, these failed to adequately explain such heterogeneity.This residual heterogeneity might derive from sources of variation between studies, most notably because of age, gender, race, lack of continuous registration, clinical measurements, nursing level, virus strains, and disease severity.Finally, most included studies were rated as having a moderate risk of bias, reflecting generally low methodological quality.The underlying explanations were the lack of control for confounders, inconsistent or unclear context in VTE evaluations, and possible selection bias due to the absence of continuous patient registration.

Conclusion
In terms of prevention and treatment of VTE, this study pointed out that COVID-19 patients in general could not benefit more from intermediate-dose or therapeutic-dose prophylactic anticoagulation than standard-dose prophylactic anticoagulation, while elderly COVID-19 patients with low bleeding risk and high VTE risk appeared to benefit more from therapeutic-dose prophylactic anticoagulation.Therefore, we suggested that individualized adjustment should be performed based on the standard-dose prophylactic anticoagulation according to the specific conditions of COVID-19 patients.At the same time, this meta-analysis further supported the expert consensus of ACCP guidelines that patients with COVID-19 should still receive standard-dose prophylactic anticoagulation, while non-critically ill patients with low bleeding risk might benefit from therapeutic-dose prophylactic anticoagulation.In summary, this meta-analysis only provided a preliminary conclusion for reference due to the objective limitations of different health service levels, types of strains, types, and doses of vaccines, presence of thromboprophylaxis, and thromboprophylaxis regimens.Further studies will still have positive clinical implications for COVID-19 patients.

FIGURE 1 PRISMA
FIGURE 1PRISMA flow chart of literature research and selection for updated meta-analysis.

FIGURE 5
FIGURE 5 Sensitivity analysis: therapeutic-dose anticoagulation versus standard-dose prophylactic anticoagulation regimen.(A) Sensitivity analysis of venous thromboembolism.(B) Sensitivity analysis of major bleeding.(C) Sensitivity analysis of mortality.
for prophylactic anticoagulation in COVID-19 patients who received no anticoagulation therapy within 6 months before admission.Only for elderly COVID-19 patients with low bleeding risk and high VTE risk, we recommended therapeuticdose prophylactic anticoagulation.In addition, the Caprini score is the most validated VTE risk assessment tool and has been used to evaluate the risk of VTE in approximately 5 million medical and surgical patients worldwide.Since COVID-19 patients are themselves at high risk for VTE, the revised Caprini Score has been tailored to the initial Caprini Score (2005 version), with the addition of a score for elevated D-dimer and a score for COVID-19 infections, specifically: asymptomatic infections are considered to be a 2-point score, symptomatic infections are considered to be a 3-point score, and symptomatic infections combined with elevated D-dimer are 5 points were considered

FIGURE 7
FIGURE 7 Network meta-analysis diagram comparing efficacy and safety of three prophylactic anticoagulation regimen for COVID-19 patients.(A) Venous thromboembolism.(B) Major bleeding.(C) Mortality.

TABLE 1
Doses of low molecular weight heparin administered in the three anticoagulation regimens.

TABLE 2 Continued
Severity case was defined as (i) need for intensive care unit admission (ii) need for mechanical ventilation with tracheal intubation (iii) CT showing severe lung invasion (iv) acute respiratory failure (v) death (vi) severe and or critical on the basis of the WHO novel grading of the severity of COVID-19.The presence of either of the above items was classified as severe case.

TABLE 4
Subgroup and sensitivity analyses for the primary outcomes.