Methodological framework for evaluation and prevention of publication bias in surgical studies

In a modern health system, clinical and healthcare decisions are more likely to be based on evidence than individual anecdotes. Doctors, unfortunately, are no better endowed with the judicial faculty than persons in other walks of life, and it is certainly not strange that distortions of facts, put out with apparent scientific accuracy, are everywhere in evidence. Publication bias may thereby mislead healthcare providers into one particular direction of care, and here is where testing for and identification of publication bias and study validity gain utmost importance. Assessment of publication bias can be attempted by either graphical or numerical methods. All tests are rather complimentary in that each contributes to a general indication of bias in included studies within a systematic review.