Good reporting practices.

Abstract

One aspect of the Good Laboratory Practice (GLP) guidelines issued by the Food and Drug Administration is that they are general in nature and do not specifically state the details of laboratory practice. This generality allows laboratories to conform to existing local practice. However, general guidelines require interpretation and in interpretation lies the potential for disagreement. This article outlines an interpretation of data reporting practice, a very particular laboratory function. Specifically, it outlines some techniques found to be useful in designing and controlling computer generated reports. The vast number of complex, interacting. continuous, time varying, random events which compose the total reality of a study are sampled measured or observed -and recorded at selected points in time, yielding “raw data”. (Since not all parameters can be economically observed or recorded, the selection of the measured values and the sample times are the subject of much professional judgment and folklore). The sample constituting raw data are then collated, sorted, averaged, processed through various algorithms and finally presented in a variety of “reports.” These reports and raw data are interpreted within the context of prior experience, judgment and previous results, yielding “final documents.” In summary: raw data amounts to records of simple, direct measurements of selected processes; reports usually involve mechanical, algorithmic, rehashing or summarizing of the raw data values; and final