Epidemiology of tuberculosis on Gran Canaria: a 4 year population study using traditional and molecular approaches.

BACKGROUND In recent years several population based studies using restriction fragment length polymorphism (RFLP) analysis have shown a higher rate of recent transmission of tuberculosis than previously thought. This study was undertaken to determine the transmission patterns of tuberculosis and the potential causes of recent transmission on the island of Gran Canaria (Spain). METHODS The strains of all patients diagnosed with tuberculosis confirmed by culture between 1 January 1993 and 31 December 1996 were typed by RFLP using the insertion sequence IS6110. A cluster was defined as two or more isolates with an identical RFLP pattern. Epidemiological linkage through contact tracing was investigated. RESULTS Of the total of 719 patients, 153 (21.3%) were excluded because there was inadequate bacterial DNA for genotyping (n=129) or the isolates of Mycobacterium tuberculosis had less than five copies of IS6110 (n=24). The isolates from 409 patients (72.3%) were grouped into 78 different clusters with an estimated 58.5% of the cases being due to recent transmission. Young age was the only significant predictor of clustering. Only in 147 (35.9%) of the 409 patients belonging to a cluster could an epidemiological link be found. 111 patients (19.6%) were identified as having had previous contact with a tuberculosis patient and 81 of them (72.9%) belonged to a cluster. The three largest clusters included 75, 49 and 20 patients, respectively. CONCLUSION Recent transmission is frequent among patients with tuberculosis on Gran Canaria and could be associated with certain aspects of control measures. Some of the clusters described in the study could be due to the prevalence of particular strains of M tuberculosis on the island.