Survey of the seroprevalence of Bartonella quintana in homeless people.

Trench fever is caused by Bartonella (Rochalimaea) quintana, a small gram-negative rod that is transmitted by body lice. Recently, B. quintana infections in homeless patients have been reported in the United States and Europe. From October 1993 to October 1994, the seroprevalence of antibodies to B. quintana was assessed by indirect immunofluorescence in a prospective study of 221 nonhospitalized homeless people, 43 hospitalized homeless patients (cases), 250 blood donors, and 57 hospitalized matched controls. Four (1.8%) of 221 nonhospitalized homeless people tested had titers of > 1:100. Of the 43 cases, seven (16%) had serological titers of > or = 1:100. None of the 250 serum samples from blood donors contained antibodies to B. quintana. The presence of antibodies to B. quintana in cases was significantly associated with the presence of body lice, exposure to cats, headaches, eastern European origin, and pain in the legs. This study demonstrates the presence of antibodies to B. quintana in the homeless population and should alert physicians that B. quintana might be an etiologic agent of fever in homeless patients.

HIV testing, HIV infection and associated risk factors among inmates in south-eastern French prisons.

OBJECTIVES: To estimate HIV seroprevalence in the two main remand and short-stay prisons of south-eastern France and to gather linked anonymous risk-factor information. SETTING: Baumettes prison, Marseille, France between 16 November and 21 December 1992. PARTICIPANTS: Using a self-administered questionnaire about HIV testing and risk factors for HIV infection, 295 male and 137 female inmates were interviewed. The response rate was 96% (100 and 90% for men and women, respectively). At the same time, 279 of a total of 432 (65%) inmates were serologically tested for HIV; 153 (35%) declined to provide a blood sample. DESIGN: Anonymous cross-sectional and surveillance survey. RESULTS: Twenty per cent of participants (84 our of 432) were intravenous drug (heroin) users (IVDU), 51% of whom reported needle-sharing prior to incarceration; 23% reported more than two sexual partners during the last year, and 13% sexual intercourse with an IVDU during the last 5 years. HIV status was available for 356 inmates (82%; 65% from blood samples and 17% from the questionnaire); 39 were HIV-infected (10.9%; 95% confidence interval, 7.7-14.2). The inmates not tested for HIV reported proportionally less risky behaviours than non-HIV-infected inmates. HIV seroprevalence was significantly higher among recidivist inmates (19.9 versus 4.4%; P < 0.0001). The rate of HIV infection was particularly high among IVDU (34 out of 84; 40%). More female non-IVDU were HIV-infected than male non-IVDU (4.1 versus 0.6%; P = 0.04). CONCLUSIONS: This study demonstrates the high prevalence of HIV infection in south-eastern French prisons, especially among IVDU. The rates may be related to the high prevalence of risky drug practices and to delays in the development of HIV prevention programmes for IVDU in France. The higher seroprevalence rate among recidivist inmates might be the result of risk behaviours during imprisonment. Another hypothesis is that recidivist inmates are at greater risk of HIV infection because of higher levels of drug use.