Statistical problems in medical research

Abstract

Many medical specialties have reviewed the drawbacks of statistical methods in medical diagnosis in specialized areas in their journals. To my knowledge this has not been done in general practice. Given
the main role of a general practitioner as a biostatistician, it would be of interest to enumerate statistical problems in assessing methods of medical diagnosis in general terms. In conducting and reporting of
medical research, there are some common problems in using statistical methodology which may result in invalid inferences being made. This paper is aimed to highlight to inexperienced statisticians, medical practitioners and personnel as well as other non-statistician some of the common statistical problem countered when using statistics to interpret data in medical research. And also comments on good practices to avoid some of these problems.