LOCF
Also written as: LOCF — Last Observation Carried Forward
A statistical imputation method for handling missing data by using a participant's last recorded value.
Full Definition
Last Observation Carried Forward (LOCF) is a statistical technique used in clinical trials to handle missing data points, particularly in longitudinal studies. When a participant drops out of a study or misses scheduled assessments, their last available observation is used to fill in all subsequent missing time points until the study's end. While LOCF was historically popular in psychiatric research due to its simplicity and conservative approach, it has significant limitations including the assumption that symptoms remain stable after dropout, which may not reflect clinical reality. Modern regulatory guidelines increasingly favor more sophisticated missing data approaches such as mixed models for repeated measures (MMRM) or multiple imputation.
Usage
Usage note: Spell out on first use, then use LOCF. Note that this method is increasingly discouraged in favor of MMRM.
In Context
- "The primary analysis used LOCF to impute missing efficacy data for patients who discontinued early." — statistical analysis plan
- "LOCF analysis showed a significant treatment difference, but MMRM results were more conservative." — results comparison