Understanding and eliminating errors in perinatal research

Research studies published in the scientific literature have had a rapid impact on medical practice, especially in clinical obstetrics. Recently, incentives to control medical costs without sacrificing the quality of care have placed increased importance on studies that truly demonstrate clinical effectiveness. In addition, there are ethical concerns that improperly designed or conducted studies adversely affect medical practice and place patients at unnecessary risk. This article summarizes potential sources of error and encourages the valid interpretation of the results and conclusions of experimental studies by making four suggestions: (1) the errors and limitations that may be present in the methodology should be described and explanations given how the errors might have affected the results and why the limitations are acceptable; (2) the methods selected should allow an answer to the question proposed; (3) the investigator should verify that the appropriate statistical test is used; and (4) when unexpected or hard to explain results are found, verification before publication should be undertaken.