LAUSR

WE READ WITH GREAT INTEREST COMMENTS FROM CAO ET AL ON OUR ARTICLE. WE AGREE WITH THEM THAT MULTIPLE TESTING COULD POTENTIALLY EXPLAIN SOME OF OUR FINDINGS AND SOME OF OUR RESULTS SHOULD BE INTERPRETED WITH CAUTION SINCE SOME ANALYSES WERE UNDERPOWERED AND RESIDUAL CONFOUNDING COULD NOT BE RULED OUT FOR MANY REASONS. FOR EXAMPLE, WHEN THE NUMBER OF EVENTS PER CONFOUNDERS IS SMALL, IT MAY BE DIFFICULT TO FULLY CONTROL FOR CONFOUNDING IN THE ANALYSIS. HOWEVER, ALL THESE POTENTIAL LIMITATIONS HAVE BEEN CLEARLY HIGHLIGHTED IN OUR MANUSCRIPT. IN ADDITION, OUR FINDING REGARDING THE TERATOGENICITY OF DOXYCYCLINE WAS SUPPORTED BY ANIMAL STUDIES .FURTHERMORE, OUR SENSITIVITY ANALYSIS TO ASSESS THE EXTENT OF RESIDUAL CONFOUNDING THAT WOULD BE REQUIRED TO NEGATE THE ASSOCIATION OBSERVED BETWEEN DOXYCYCLINE AND THE RISK OF VENTRICULAR/ATRIAL SEPTAL DEFECT INDICATE THAT SUBSTANTIAL UNMEASURED CONFOUNDING WOULD BE NEEDED TO REDUCE THE OBSERVED ASSOCIATION. IN CONCLUSION, WE RECOMMEND THE USE OF SAFE ANTIBIOTICS IN EARLY PREGNANCY FOR THE TREATMENT OF MATERNAL INFECTION IN EARLY PREGNANCY AFTER WEIGHING THE RISK AND BENEFITS FOR BOTH MOTHER AND FETUS.