Wei S C, Tsai J J
Department of Neurology, China Medical College, Taichung, R.O.C.
Zhonghua Yi Xue Za Zhi (Taipei). 1994 Jun;53(6):331-7.
Critical assessment of diagnostic accuracy is indispensable to resident training and medical education, especially in regard to our situation of the extremely low rate of autopsy. Recent setup of advanced technology for neurological diagnosis has resulted in a tendency to overuse these tools at the expense of using classic bedside diagnostic approach. Disclosing the common errors of daily practices among neurological residents necessitates implementing this pioneer study.
The tentative diagnosis of each patient, as performed by neurological residents after finishing bedside diagnostic processes, was assessed by final diagnosis concluded from clinical syndrome and/or the results of laboratory studies. The overall rate of diagnostic error and the frequency of diagnostic inaccuracy in various disease entities were evaluated. The reason for the diagnostic error was determined by reviewing the records regarding the entire diagnostic process.
1336 consecutive patients from January 1990 to December 1990 were recruited for this study. The initial bedside diagnosis was correct in 901 (67%) patients. No definite final diagnosis could be obtained in 169 (13%) patients. The diagnoses were incorrect in 266 (20%) patients. The highest rate of inaccuracy was found in the diagnosis of subdural hematoma (56%). The other common diseases of high rate of diagnostic inaccuracy were myasthenia gravis (50%), subarachnoid hemorrhage (42%), Guillain-Barré syndrome (40%), traumatic disorders (39%), herniation of intervertebral disc (33%), metabolic encephalopathy (30%), infection of central nervous system (30%), intracranial neoplasm (24%), drug overdose or intoxication (22%), and mixed neurological and metabolic encephalopathy (21%). The explored reasons for diagnostic inaccuracy were errors of reasoning (38%), inadequate data base (35%), and inadequate fund of knowledge (27%).
This study confirmed that the basic methods of bedside diagnosis with a standardized sequence of history taking, physical examination and diagnostic reasoning are currently still the most fundamental process of achieving an accurate diagnosis in neurological emergencies. No shortcut or mechanical substitute is available for clinical diagnosis.
对诊断准确性进行严格评估对于住院医师培训和医学教育至关重要，尤其是在我国尸检率极低的情况下。近期神经诊断先进技术的设立导致了过度使用这些工具的趋势，而牺牲了经典的床边诊断方法。揭示神经科住院医师日常实践中的常见错误需要开展这项开创性研究。
神经科住院医师完成床边诊断流程后对每位患者的初步诊断，通过根据临床综合征和/或实验室研究结果得出的最终诊断进行评估。评估了诊断错误的总体发生率以及各种疾病实体中诊断不准确的频率。通过审查整个诊断过程的记录来确定诊断错误的原因。
本研究招募了1990年1月至1990年12月期间连续的1336例患者。901例（67%）患者的初始床边诊断正确。169例（13%）患者无法获得明确的最终诊断。266例（20%）患者的诊断错误。硬膜下血肿诊断的不准确率最高（56%）。诊断不准确率较高的其他常见疾病包括重症肌无力（50%）、蛛网膜下腔出血（42%）、格林-巴利综合征（40%）、创伤性疾病（39%）、椎间盘突出症（33%）、代谢性脑病（30%）、中枢神经系统感染（30%）、颅内肿瘤（24%）、药物过量或中毒（22%）以及神经和代谢性混合性脑病（21%）。诊断不准确的原因经探究为推理错误（38%）、数据库不充分（35%）以及知识储备不足（27%）。
本研究证实，采用标准化的病史采集、体格检查和诊断推理顺序进行床边诊断的基本方法目前仍是在神经科急症中实现准确诊断的最基本过程。临床诊断没有捷径或机械替代方法。
