The Cusum Plot: Its Utility in the Analysis of Clinical Data

Abstract

Clinicians are constantly required to analyze clinical and laboratory data for trends. Is a patient's temperature really beginning to come down in response to antibiotics? Is the white-cell count or blood urea nitrogen really falling? These questions are particularly frustrating because these data often have a natural degree of scatter, which obscures early detection of trends. Discerning trends in reams of laboratory reports is extremely vexing. Flow sheets help, and plotting data on a graph helps more, although widely used only for vital signs. Other approaches are more powerful in the early detection of trends in data, and deserve recognition. . . .