Thursday, October 15, 2009

B - Does analysis using "last observation carried forward" introduce bias in dementia research?

Does analysis using "last observation carried forward"introduce bias in dementia research?

CMAJ • October 2008;179 (8). doi:10.1503/cmaj.080820.

A very critical standpoint on a widely used analytical technique in dementia research drug trials, called " last observation carried forward". Patients affected by dementia who are on drug trials are followed over a period of time. When there is a drop-out from the evaluation trial, the so called "last observation carried forward" (statistical technique) is applied; taking into account the last result of the observation period (which eliminates the actual state of the patient's progress or decline after the interruption of the evaluation test) creating, therefore, an artificial analytical result. When the trial stops, the data obtained is, so to say, "outstretched" to obtain what in the end is a fake outcome of a drug trial that should be intended to realistically measure the ongoing test to its completion and for its eventual benefits to such patients.

http://www.cmaj.ca/cgi/content/full/179/8/751

No comments: