Statistical calculations of the confidence interval and significant difference ranges of the morphologic index.

Since Hansen first noted variations in the morphology of leprosy bacilli in 1895, many investigators have given attention to such morphologic variations as revealed by staIned preparations. After the advent of antileprosy drugs, many clinicians observed rapid decrease in the proportion of uniformly stained bacilli under therapy. In 1960 Ridley first proposed a "Granularity Index" based on three forms of the bacilli (solid, fragmented and granular) for quantitative assessment of these morphologic changes (4). As the granularity index was too complex to be readily used clinically, the "Morphologic Index" ( MI ) based on two forms of the organisms, solid and nonsolid; was proposed a few years later. Since leprosy bacilli are, as yet, uncultivable in vitro, it is still useful to follow the morphologic changes in leprosy bacilli during the course of chemotherapy. The MI is thus widely used, not only in clinical practice, but also in laboratory experiments. However, many basic problems regarding the MI remain unsolved. Thus satisfactory studies on the statistical basis of the MI are not yet available. This report is an attempt to study some statistical properties of the MI as it is usually carried out from a direct smear. Morphologic Index and binomial distribution. Even in a small lepromatous lesion there is a huge set of leprosy bacilli which consists of numerous solid bacilli and numerous non-solid bacilli. The ratio of the solid bacilli to the total bacilli in the lesion is the Morphologic Index of this bacillary set, or, in statistical term, the MI of the "Population." Although we are seeking to know the real value of MI, it is impossible to know the exact value of the MI of