MITCHISON D A, BHATIA A L, RADHAKRISHNA S, SELKON J B, SUBBAIAH T V, WALLACE J G
Bull World Health Organ. 1961;25(3):285-312.
A series of studies on the virulence in the guinea-pig of tubercle bacilli isolated before treatment from Indian tuberculous patients admitted to a controlled comparison of different regimens of domiciliary chemotherapy has recently been undertaken by the Tuberculosis Chemotherapy Centre, Madras. The main object of these studies was to determine whether the differences in virulence of the tubercle bacilli obtained from Indian patients before the start of chemotherapy were related to the severity or type of the patients' disease at that time and to the subsequent response to treatment. Before these relationships could be investigated, however, it was necessary to find out whether the results of the virulence tests, which were carried out over a period of two-and-a-half years at the Centre and at the Microbiological Research Establishment, Porton, England, could be considered as a unified whole-that is, as if they had all been done on the same day in the same laboratory.A proportion of the cultures was stored at - 20 degrees C for 44-78 weeks, but this did not affect their virulence. Inter-experimental variation was found to be small in the Porton series of tests and undetectable in the Madras series, and the results in the latter series could be successfully adjusted to those in the former by allowing for differences in the means and standard deviations of the distributions for the two series. The measure of virulence used was found to be reasonably acceptable for the analysis of variance technique. Suggestions are made as to ways of improving the efficiency of the experimental design in future studies.
马德拉斯结核病化疗中心最近开展了一系列研究，针对印度结核病患者在接受不同家庭化疗方案的对照比较之前分离出的结核杆菌，检测其对豚鼠的毒力。这些研究的主要目的是确定化疗开始前从印度患者身上获取的结核杆菌毒力差异，是否与当时患者疾病的严重程度或类型有关，以及与后续治疗反应有关。然而，在研究这些关系之前，有必要弄清楚在该中心以及英国波顿微生物研究机构进行的为期两年半的毒力测试结果，是否可以被视为一个统一的整体——也就是说，就好像所有测试都是在同一天在同一个实验室进行的一样。一部分培养物在-20摄氏度下保存了44至78周，但这并未影响它们的毒力。在波顿系列测试中发现实验间差异很小，在马德拉斯系列中则无法检测到，通过考虑两个系列分布的均值和标准差差异，马德拉斯系列的结果可以成功地与波顿系列的结果进行调整。结果发现，所使用的毒力测量方法对于方差分析技术来说是相当可接受的。文中还就未来研究中提高实验设计效率的方法提出了建议。
