Inter-clinician agreement on clinical data abstracted from patients' medical charts

The purpose of this study was to determine the inter-observer repeatability of retrospective chart reviews.

Retrospective abstraction of corneal transplant patients' medical records by two independent observers was performed at the University of California, Davis, Department of Ophthalmology, Sacramento, CA. The sample consisted of 100 consecutive patients who underwent penetrating keratoplasty, and the primary outcome measure was the degree of agreement between the 2 independent observers on clinical ocular data abstracted from patients' medical charts.

Even when given specific criteria for abstraction of data from patients' medical charts, inexperienced but trained observers exhibited a small degree of disagreement (range 0 to 12%) on patient demographic variables (e.g., gender, age, and date of birth) to be used for data analysis. Larger amounts of disagreement (range 3 to 26%) were found on frequently occurring variables (e.g., predisposing ocular conditions, surgical complications, and classification of corneal staining).

Based on these data, inter-observer repeatability for retrospective chart review studies is good, but investigators using this study methodology must be quite specific about the data to be collected and should be aware of standardization techniques to enhance observer repeatability.