Evaluating the efficacy of internal teat sealants at dry-off for the prevention of new intra-mammary infections during the dry-period or clinical mastitis during early lactation in dairy cows: A systematic review update and sequential meta-analysis

A systematic review and Bayesian sequential pair-wise meta-analyses were conducted to assess the efficacy of internal teat sealants (ITS) administered at dry-off in comparison to no treatment for preventing new intramammary infections (IMI) and clinical mastitis (CM) in dairy cattle. This work updated a previous systematic review and network meta-analysis conducted in 2019 but employed a narrowed scope and eligibility. The updated eligibility included studies that used ITS without concurrent therapy compared to a no treatment control (NTC), a study population of dairy cows or prepartum heifers, controlled trial design, and assessed one of the following outcomes: incidence of new IMI at calving or CM during the first 30 days in milk (DIM). Risk of bias was assessed through the Cochrane Risk of Bias 2.0 tool. Evidence quality was assessed using Grading of Recommendations Assessment, Development, and Evaluation (GRADE). There were 141 potentially relevant records identified from the updated search conducted on April 29, 2021, with a publication date restriction of 2018 or later; one study passed full-text screening and was included. Of the 32 studies included in the previous review, 12 studies were relevant after applying the modified eligibility criteria, totaling 13 studies included in this review (12 addressing IMI at calving outcome, 4 addressing CM at 30 DIM outcome). Sequential meta-analysis was conducted for both outcomes in R 3.6.0. Decisions for stopping were assessed at each analysis for intervention effect or futility in finding an effect based on a priori minimum clinically relevant values (ORδ =0.5, 0.75). ITS at dry-off significantly reduced odds of new IMI at calving compared to NTC at the second meta-analysis (OR2 =0.27, 95% CI=0.22–0.34), and onward (OR12 =0.29, 95% CI=0.27–0.32). For CM at 30 DIM, significance was reached at the second meta-analysis (OR2 =0.59, 95% CI=0.47–0.73), and onward (OR3 =0.47, 95% CI=0.42–0.51). Stopping for effect occurred at the second analysis in both outcomes and ORδs, but low-quality evidence and heterogeneity concerns were noted. A continuity-correction to include zero-event CM studies showed significance at the third meta-analysis (OR3 =0.79, 95% CI=0.73–0.86), stopping for effect at the fourth for ORδ = 0.75 (OR4 =0.77, 95% CI=0.72–0.83), and stopping for futility at the second for ORδ = 0.5 (OR2 =0.94, 95% CI=0.75–1.20), but the main CM analysis was considered more appropriate due to the sensitivity analysi