Objective: To analyze for which reasons the certainty of evidence was downgraded in neonatal Cochrane reviews.Methods: We performed a systematic meta-epidemiological review for Cochrane neonatal reviews published in 2022-2024. The search was performed in January 2025, and all reviews were screened by two authors. We extracted the information from the summary of findings tables. As the main outcome we compared the reasons for downgrading per evidence certainty categories. Secondary outcomes included the analysis of null effects and comparison of confidence interval width. Chi2 was used to analyze the categorized variables.Results: We included 54 reviews with 467 outcomes of which evidence certainty wasrated very low (35%), low (43%), moderate (20%), and high (2%). Imprecision and risk of bias were the most frequent reasons for downgrading certainty of evidence (p<0.001). Outcomes with effect estimates including the null were more often downgraded for imprecision, whereas outcomes without null effects were more often downgraded for risk of bias. A strong association was observed between certainty level and null effects: very low certainty evidence most often included the null effect, followed sequentially by low, moderate, and high certainty evidence (p<0.001). Among dichotomous outcomes, wide confidence intervals were the predominant driver of imprecision, with CI width clearly associated both with certainty categories and with the frequency of downgrading due to imprecision.Conclusion: The neonatal evidence was mainly limited due to imprecision and risk of bias. This indicates that larger scale high-quality studies in various neonatal topics are still greatly warranted.

Objective: To analyze for which reasons the certainty of evidence was downgraded in neonatal Cochrane reviews.Methods: We performed a systematic meta-epidemiological review for Cochrane neonatal reviews published in 2022-2024. The search was performed in January 2025, and all reviews were screened by two authors. We extracted the information from the summary of findings tables. As the main outcome we compared the reasons for downgrading per evidence certainty categories. Secondary outcomes included the analysis of null effects and comparison of confidence interval width. Chi2 was used to analyze the categorized variables.Results: We included 54 reviews with 467 outcomes of which evidence certainty wasrated very low (35%), low (43%), moderate (20%), and high (2%). Imprecision and risk of bias were the most frequent reasons for downgrading certainty of evidence (p<0.001). Outcomes with effect estimates including the null were more often downgraded for imprecision, whereas outcomes without null effects were more often downgraded for risk of bias. A strong association was observed between certainty level and null effects: very low certainty evidence most often included the null effect, followed sequentially by low, moderate, and high certainty evidence (p<0.001). Among dichotomous outcomes, wide confidence intervals were the predominant driver of imprecision, with CI width clearly associated both with certainty categories and with the frequency of downgrading due to imprecision.Conclusion: The neonatal evidence was mainly limited due to imprecision and risk of bias. This indicates that larger scale high-quality studies in various neonatal topics are still greatly warranted.