BACKGROUND: Fever of unknown origin (FUO) has been insufficiently studied in patients infected with HIV. The aim of this study was to determine the incidence, mean hospital stay, etiology, diagnostic methods and clinical course in patients infected with HIV and FUO. METHODS: Descriptive prospective study of patients infected with HIV and with FUO for a period of 24 months (February 1993-February 1995). Out of a total of 1,202 admissions, 100 were prompted by FUO in 95 patients (61 males, 34 women). Sixty-six per cent were parenteral drug abusers or had been so and the mean T CD4+ lymphocyte count was 0.061 x 10(9)/1 +/- 0.075 x 10(9)/1. RESULTS: The incidence of FUO was 8.3 every 100 patients/year. The mean hospital stay was 31.3 +/- 17.8 days, compared with an overall mean stay for HIV-positive patients of 14.6 +/- 8 days (p < 0.001). An etiologic diagnosis was achieved in 90% of cases. Pulmonary or extrapulmonary tuberculosis accounted for approximately 50% of cases, whereas non tuberculous mycobacteria (MAI and M. kansasii) for 20%. The other etiologies (CMV, visceral leishmaniasis, PCP, toxoplasmosis, cryptococcosis, lymphoma) accounted each for less than 5% of diagnoses. In 14 occasions there were two concomitant diagnoses. Mortality rate was 22%, which was higher when two simultaneous diagnosis were present [RR: 3.17 (1.5-6.6)]. In 45% of fatal cases one of the diagnoses went undiagnosed premortem. The highest diagnostic yield was obtained with sputum culture (34.2% of diagnosis) and blood culture (21.5%). CONCLUSIONS: FUO in common in patients infected with HIV, with important health resources consumption associated. The obtention of diagnosis was possible in a high percentage of patients. It is associated with a high hospital mortality rate. The most common diagnoses were tuberculosis and infections by other mycobacteria. It is possible to obtain the diagnosis with non-invasive procedures in 75% of cases.
BACKGROUND: Fever of unknown origin (FUO) has been insufficiently studied in patients infected with HIV. The aim of this study was to determine the incidence, mean hospital stay, etiology, diagnostic methods and clinical course in patients infected with HIV and FUO. METHODS: Descriptive prospective study of patients infected with HIV and with FUO for a period of 24 months (February 1993-February 1995). Out of a total of 1,202 admissions, 100 were prompted by FUO in 95 patients (61 males, 34 women). Sixty-six per cent were parenteral drug abusers or had been so and the mean T CD4+ lymphocyte count was 0.061 x 10(9)/1 +/- 0.075 x 10(9)/1. RESULTS: The incidence of FUO was 8.3 every 100 patients/year. The mean hospital stay was 31.3 +/- 17.8 days, compared with an overall mean stay for HIV-positivepatients of 14.6 +/- 8 days (p < 0.001). An etiologic diagnosis was achieved in 90% of cases. Pulmonary or extrapulmonary tuberculosis accounted for approximately 50% of cases, whereas non tuberculous mycobacteria (MAI and M. kansasii) for 20%. The other etiologies (CMV, visceral leishmaniasis, PCP, toxoplasmosis, cryptococcosis, lymphoma) accounted each for less than 5% of diagnoses. In 14 occasions there were two concomitant diagnoses. Mortality rate was 22%, which was higher when two simultaneous diagnosis were present [RR: 3.17 (1.5-6.6)]. In 45% of fatal cases one of the diagnoses went undiagnosed premortem. The highest diagnostic yield was obtained with sputum culture (34.2% of diagnosis) and blood culture (21.5%). CONCLUSIONS: FUO in common in patients infected with HIV, with important health resources consumption associated. The obtention of diagnosis was possible in a high percentage of patients. It is associated with a high hospital mortality rate. The most common diagnoses were tuberculosis and infections by other mycobacteria. It is possible to obtain the diagnosis with non-invasive procedures in 75% of cases.