Mapping the Characteristics of Network Meta-Analyses on Antithrombotic Therapies: An Overview and Critical Appraisal

OBJECTIVES: A large number of network meta-analysis (NMAs) in the field of cardiac disease are available, yet the scientific literature lacks on updated straightforward synthesis of this evidence to ground decision-making process. We aimed to map and critical appraise NMAs on antithrombotic therapies used as treatment or prophylaxis of cardiac diseases and cardiac surgical procedures.

METHODS: A systematic review of systematic reviews with meta-analysis was conducted following Cochrane Collaboration and Joanna Briggs recommendations (PROSPERO-CRD2020166468). Searches to identify NMAs meeting the eligibility criteria of this this study were performed in PubMed and Scopus (Jan-2022). NMAs characteristics including metadata, statistical models’ description and main pooled results were collected. The methodological quality of NMAs was evaluated using PRISMA-NMA checklist and AMSTAR-2 tools. Descriptive statistical analyses with categorical variables reported as frequencies and continuous variables as median and interquartile range (IQR) were performed (SPSS-Statistics v.25.0).

RESULTS: Overall, n=88 NMAs published between 2007-2022 were identified. The most evaluated clinical condition was atrial fibrillation (n=57; 64.7%); around one third of studies (38.6%) assessed cardiac surgical procedures. Only 28.4% NMAs had a registered study protocol. Fifty NMAs (56.8%) were published by authors from one single country being China the most frequent. A median of 14 primary studies (IQR 5-20.75) (mostly randomized clinical trials) were included per NMA. A median of 40 (IQR 24-84.25) indirect meta-analyses per study was found. At least one network diagram for a given outcome was provided by 68 (77.2%) studies, yet only 22 (25.6%) performed a treatment ranking analyses. Conflict of interest declarations and study’s funding were informed by 34 (38.6%) and 38 (43.2%) NMAs, respectively.

CONCLUSIONS: Although there is a wide spread of NMA-type studies assessing different antithrombotic agents for different cardiac conditions, the lack of standardized conduction and reporting of NMAs (poor-moderate methodological quality) may limit their comparison and results implementation into clinical practice.