应当如何评价一种新疗法?接受新疗法患者的改善可能是由于其他的影响,如自然消退或附加治疗,所以印象能导致错误的结论。不过已知患者的预后,而新疗法能大大改变此预后时,就不会导致错误,如结核性脑膜炎以前是致命性的,用链霉素治疗此病无疑是一大的进步。当治疗后病情改变很大时,影响治疗组和对照组之间差别的其他因素就不突出了,疗效则压倒其他一切因素。有时医生就可宣称效果有足够的临床意义,无须统计学检验。 How should a new therapy be evaluated? The patient’s improvement in receiving the new therapy may be due to other effects such as spontaneous regression or additional treatment, so the impression can lead to erroneous conclusions. However, the prognosis of the patient is known, and new treatments can greatly change this prognosis without causing errors. For example, tuberculous meningitis was previously fatal. Streptomycin treatment is undoubtedly a major advancement. When the condition changes greatly after treatment, the other factors affecting the difference between the treatment group and the control group are not prominent, and the efficacy is overwhelming all other factors. Sometimes doctors can claim that the effect is clinically significant and does not require statistical testing.