Quantifying the association of sperm DNA fragmentation with assisted reproductive technology outcomes: An umbrella review

Abstract

AbstractBackgroundSystematic reviews and meta‐analyses are instrumental in shaping clinical practice. However, their findings can sometimes be marred by discrepancies and potential biases, thereby diluting the strength of the evidence presented. Umbrella reviews serve to comprehensively assess and synthesise these reviews, offering a clearer insight into the quality of the evidence presented. In the context of the relationship between sperm DNA fragmentation (SDF) and assisted conception outcomes, there is a divergence in the literature. Some reviews suggest a clear cause‐and‐effect linkage, whereas others present conflicting or inconclusive results.ObjectivesIn this umbrella review we aimed to synthesise the evidence collated in systematic reviews and meta‐analyses summarising the association of SDF with assisted reproductive technology (ART) outcomes.Search strategyAfter preregistration (https://doi.org/10.17605/OSF.IO/6JHDP), we performed a comprehensive search of the PubMed, Scopus, Cochrane Library, Web of Science and Embase databases. We conducted a search for systematic reviews on the association between SDF and ART without any restrictions on language or publication date.Selection criteriaSystematic reviews and meta‐analyses assessing the association between SDF and ART outcomes were eligible.Data collection and analysisWe assessed the quality of the included reviews using AMSTAR 2 and ROBIS, and determined the degree of overlap of primary studies between reviews estimating the corrected covered area (CCA), adjusted for structural missingness. We evaluated the most recent reviews assessing the association of SDF with live birth, pregnancy, miscarriage, implantation, blastulation and fertilisation. The synthesis of evidence was harmonised across all included quantitative syntheses, re‐estimating the odds ratio (eOR) in random‐effects meta‐analyses with 95% confidence intervals (95% CIs) and 95% prediction intervals (95% PIs). We categorised the evidence strength into convincing, highly suggestive, suggestive, weak or nonsignificant, according to the meta‐analysis re‐estimated P‐value, total sample size, I2 statistic for heterogeneity, small study effect, excess significance bias and the largest study significance.Main resultsWe initially captured and screened 49 332 records. After excluding duplicate and ineligible articles, 22 systematic reviews, 15 of which were meta‐analyses, were selected. The 22 reviews showed a moderate degree of overlap (adjusted CCA 9.2%) in their included studies (overall n = 428, with 180 unique studies). The 15 meta‐analyses exhibited a high degree of overlap (adjusted CCA = 13.6%) in their included studies (overall n = 274, with 118 unique studies). AMSTAR 2 categorised the quality of evidence in 18 reviews as critically low and the quality of evidence in four reviews as low. ROBIS categorised all the reviews as having a high risk of bias. The re‐estimated results showed that the association of SDF with live birth was weak in one and nonsignificant in four meta‐analyses. Similarly, the association of SDF with pregnancy, miscarriage, implantation, blastulation and fertilisation was also weak or nonsignificant. The association of high SDF with different ART outcomes was also weak or nonsignificant for different interventions (IVF, ICSI and IUI) and tests.ConclusionsThis umbrella review did not find convincing or suggestive evidence linking SDF with ART outcomes. Caution should be exercised in making any claims, policies or recommendations concerning SDF.