Influence of additional information on interrater reliability in the neurologic examination

We performed a routine neurologic examination on 200 patients who were reexamined by a second physician under three different conditions: (1) reexamination of only one item, (2) complete reexamination without knowledge of the patient's history and complaints, and (3) complete reexamination with such knowledge. We used changes in reliability as an index of the effect of additional information on basic data collection. Additional information increased reliability, indicating that diagnostic hypotheses obtained mainly from the history improve the examiner's precision. When a complete reexamination was performed with knowledge of the patient's history, reliability was substantial (kappa > 0.6). More experienced examiners do not perform the examination more reliably, but do utilize the patient's history differently.