Understanding and managing sanitary risks due to rodent zoonoses in an African city: beyond the Boston Model.

The Boston Model describes a successful rodent management plan that succeeded in a first-world city in the USA. In third-world cities, which often contain informal shack settlements, it is debatable whether the Boston Model would apply. In Durban, a major harbor city of three million people on the east coast of South Africa, we investigated the sanitary risks due to rodents in both formal (residential and commercial) and informal (shacks) sectors, and we evaluated the relative merits of different management interventions suggested by the Boston Model. Blood and tissue samples of six species (Rattus norvegicus, R. tanezumi, R. rattus, Mus musculus, Mastomys natalensis, Tatera brantsi) from 262 live-trapped rodents from 54 localities were tested for antibodies or DNA for plague (n= 193: antibody test), leptospirosis (n= 221 for antibody test; n= 69 for polymerase chain reaction test for DNA) and toxoplasmosis (n= 217: antibody test). We conducted a socioeconomic survey of 90 household to determine environmental and socioeconomic disease risk factors in the shack settlement of Cato Crest. No rodents were seropositive for plague, but nine Norway rats, R. norvegicus (4.1% of the sample tested) were seropositive for toxoplasmosis, and 22 R. norvegicus (10.0% of sample tested) were seropositive for leptospirosis. Disease endemic areas were concentrated in Cato Crest and the commercial district of Durban. Serology tests of humans living in Cato Crest (n= 219) showed 0% exposure to plague, 23% to leptospirosis and 35% to toxoplasmosis. Compared with shack-dwellers, the residents of brick houses had slightly lower levels of exposure to leptospirosis and toxoplasmosis. Based on our results, environmental hygiene and rodent-trapping campaigns were launched in Cato Crest. The initiative owes much of its current success to implementation of the principles inherent in the Boston Model, even though certain elements were lacking.