In studies of Lassa fever in Sierra Leone, the prevalence of human antibody to Lassa virus ranged from 8% to 52%. Mastomys natalensis, the reservoir of Lassa virus, constituted 50%-60% of the rodents captured in houses but only 10%-20% of those captured in surrounding agriculture and bush areas (chi 2 = 90.2, P less than 10(-6), df = 1), a finding suggesting that houses are the most-important location for transmission of Lassa virus. Viral infection of Mastomys from houses ranged from 0% to 80%. The incidence of seroconversions in susceptible persons ranged from 5% to 22% per year; the ratio of illness to infection ranged from 9% to 26%, and the proportion of febrile illness associated with seroconversion was 5%-14%. Eightfold rises in titer of antibody occurred in 1%-18% of the antibody-positive population, a result suggesting reinfection. We estimate the ratio of fatalities to infection to be 1%-2%, a rate lower than estimates based on hospitalized cases. The high incidence of Lassa fever makes it a major problem in West Africa.

In a study of Lassa fever in Sierra Leone, West Africa, we identified two variables associated with a high risk of death, and we evaluated the efficacy of ribavirin and Lassa virus-convalescent plasma for the treatment of Lassa fever. A serum aspartate aminotransferase level greater than or equal to 150 IU per liter at the time of hospital admission was associated with a case-fatality rate of 55 percent (33 of 60). Patients with the same risk factor who were treated for 10 days with intravenous ribavirin, begun within the first 6 days after the onset of fever, had a case-fatality rate of 5 percent (1 of 20) (P = 0.0002 by Fisher's exact test). Patients whose treatment began seven or more days after the onset of fever had a case-fatality rate of 26 percent (11 of 43) (P = 0.01). Viremia with levels greater than or equal to 10(3.6) TCID50 per milliliter on admission was associated with a case-fatality rate of 76 percent (35 of 46). Patients with this risk factor who were treated with intravenous ribavirin within the first six days after onset of fever had a case-fatality rate of 9 percent (1 of 11) (P = 0.006), whereas those treated after seven days or more of illness had a fatality rate of 47 percent (9 of 19) (P = 0.035). Oral ribavirin was also effective in patients at high risk of death. Lassa-convalescent plasma did not significantly reduce mortality in any of the high-risk groups. We conclude that ribavirin is effective in the treatment of Lassa fever and that it should be used at any point in the illness, as well as for postexposure prophylaxis.

A prospective case-control study of Lassa fever was established in Sierra Leone to measure the frequency and case-fatality ratio of Lassa fever among febrile hospital admissions and to better delineate the clinical diagnosis and course of this disease. Lassa fever was responsible for 10%-16% of all adult medical admissions and for approximately 30% of adult deaths in the two hospitals studied. The case-fatality ratio for 441 hospitalized patients was 16.5%. We found the best predictor of Lassa fever to be the combination of fever, pharyngitis, retrosternal pain, and proteinuria (predictive value together, .81); of outcome, the best predictor was the combination of fever, sore throat, and vomiting (relative risk of death, 5.5). Complications included mucosal bleeding (17%), bilateral or unilateral eighth-nerve deafness (4%), and pleural (3%) or pericardial (2%) effusion. Lassa fever is endemic in this area and is a more-common cause of hospital admission and death than has previously been described; this disease must be considered when diagnosing febrile illness in West Africa.

We measured levels of virus in sequential specimens from 137 patients with Lassa fever. The probability of fatal disease increased significantly with the level of viremia measured either on admission or during the course of illness. The odds ratio of death in patients with viremia greater than 10 TCID50/ml was 3.7 (90% confidence interval, 1.9-7.2). The same ratio in patients with viremia greater than 10 TCID50/ml and with levels of aspartate aminotransferase greater than or equal to 150 IU/liter was 21.5 (95% confidence interval, 5.2-99.0). Virus was found in throat cultures from 39% of viremic patients, compared with 14% of nonviremic patients (P less than .002); however, the level of virus was usually less than or equal to TCID50/ml. Fewer than 3% of patients were viruric during acute illness, and virus was isolated from three of three samples of cerebrospinal fluid. On admission, 53% of patients had IgG antibodies, and 67% had IgM antibodies. Recovery was not associated with the presence of either IgG or IgM. Virus was isolated from greater than 100 serum specimens that also contained high titers of IgG. Clinical Lassa fever was shown to be a disseminated systemic, primary viral infection, with an outcome highly associated with viremia but not with development of antibody.