Human African trypanosomiasis, or sleeping sickness, is a parasitic disease affecting rural populations in up to 36 countries in Africa. Sixty million people are at risk for the disease but surveillance is poor. An estimated 300,000 new cases occur each year with 100,000 deaths. Sleeping sickness is invariably fatal if left untreated. The medications used to treat sleeping sickness were developed more than 50 years ago and have severe adverse side effects. Few alternatives are available. From five to thirty percent of late-stage patients suffer treatment failures (relapse) and recent evidence suggests that the rate of treatment failure in many sleeping sickness endemic areas is rising. Given the lack of alternative medication, the decrease in effectiveness of first-line medications for sleeping sickness is alarming If sleeping sickness patients can be identified as likely to relapse, treatment regimens could be altered to improve outcomes. During a sleeping sickness epidemic in Tambura County, South Sudan, 2,324 cases were treated, with 180 relapses (7.8%), between 1997 and 1999. Clinical and demographic variables were examined for their predictive value for relapse in Pearson's Chi-square, Kaplan-Meier and Cox regression analyses. Relapse rates were documented for a number of sub-populations and the temporal aspects of treatment failure were explored. Finally, the geographic distribution of sleeping sickness cases was investigated for trends, an indication of drug resistant strains While controlling for the effects of other variables, trypanosomes in cerebrospinal fluid (CSF) doubled the risk of relapse (RR: 1.98, 95% CI: [1.30--3.01]) and an elevated CSF white blood cell count resulted in a two percent increased risk for every increase of 10 cells/mm3 CSF (RR: 1.002, 95% CI: [1.001--1.003]). Evidence of trypanosomes in the lymph system and gender were also predictive of relapse, but to a lesser degree. The rate of relapse in the population was 20.1/1000 person-months, or 15.3% of the population. Eighty-five percent of relapses occurred in the first year post treatment and time-to-event analyses indicated that time does not influence the probability of relapse. There was little evidence of geographic clustering of relapse cases and therefore, no indication of drug-resistant strains of trypanosomiasis
