Surgeon bias in the medical record

The medical chart is presumed to be an impartial record of a patient's history, examination, and treatment. This study was designed to determine if hand surgeons introduce selective bias into the medical record. Fifty consecutive new patients were referred to a hand surgery practice for presumed carpal tunnel syndrome. Forty-four patients were diagnosed with carpal tunnel syndrome and were eligible for study inclusion. Prior to evaluation, all patients wrote in free-response fashion and diagrammed their symptom quality and location. Surgeon dictations from these visits were then evaluated against the patients' self-reported symptomatology. Statistical analysis was performed to determine if specific symptom types and locations were preferentially omitted from medical records. The surgeons' records most closely replicated patients' self-reported diminished sensibility in the thumb, index, and middle fingers (100% inclusion rate). When patients described presenting with pain, it was recorded significantly less frequently (P<.001) in the surgeon's note (16 of 27; 59%). Surgeon documentation of patient symptoms in the median nerve distribution (98% inclusion) was more likely than documentation of symptoms in the small finger (50% inclusion; P<.001), forearm (45% inclusion; P<.001), and arm (50% inclusion; P<.001). Because symptoms less readily attributed to median nerve compression at the wrist were more likely to be omitted, it is possible that surgeons introduce bias into the medical record. Although the cause of this discrepancy is likely multifactorial, the medical records often failed to document patients' pain as well as symptoms outside of the median nerve distribution.