Sir.—The recent analysis of 46 patients with the VATER association by Weaver et al1 prompts the following observations.
1. Selection for inclusion in the study by the presence of only two of the six major association defects (vertebral defects, imperforate anus, tracheoesophageal malformations, upper extremity defects, renal malformations, and cardiac anomalies) is insufficiently rigorous. When all common congenital malformations are examined, virtually all two-defect combinations occur more frequently than would be expected by chance alone.2 Twenty-eight percent of infants with cardiac lesions have noncardiac anomalies as well; skeletal anomalies are most common, and urinary tract anomalies are second.3 Do all of these infants show examples of the VATER association?
2. We are not told the number of infants with only two defects, which defects they were, or which associated "nonmajor" defects were present. As a result, it is difficult to evaluate one of the more interesting