A novel approach : the propensity to propagate (PTP) method for controlling for host factors in studying the transmission of Mycobacterium tuberculosis

Abstract

RATIONALE: Understanding the genetic variations among Mycobacterium tuberculosis (MTB) strains with differential ability to transmit would be a major step forward in preventing transmission. OBJECTIVES: To describe a method to extend conventional proxy measures of transmissibility by adjusting for patient-related factors, thus strengthening the causal association found with bacterial factors. METHODS: Clinical, demographic and molecular fingerprinting data were obtained during routine surveillance of verified MTB cases reported in the Netherlands between 1993 and 2011, and the phylogenetic lineages of the isolates were inferred. Odds ratios for host risk factors for clustering were used to obtain a measure of each patient's and cluster's propensity to propagate (CPP). Mean and median cluster sizes across different categories of CPP were compared amongst four different phylogenetic lineages. RESULTS: Both mean and median cluster size grew with increasing CPP category. On average, CPP values from Euro-American lineage strains were higher than Beijing and EAI strains. There were no significant differences between the mean and median cluster sizes among the four phylogenetic lineages within each CPP category. CONCLUSIONS: Our finding that the distribution of CPP scores was unequal across four different phylogenetic lineages supports the notion that host-related factors should be controlled for to attain comparability in measuring the different phylogenetic lineages' ability to propagate. Although Euro-American strains were more likely to be in clusters in an unadjusted analysis, no significant differences among the four lineages persisted after we controlled for host factors.