Serological markers predicting tuberculosis in human immunodeficiency virus-infected patients.

SETTING: Human immunodeficiency virus (HIV)-positive patients retrospectively identified at the Hospital of Bari, Italy, with diagnosis of tuberculosis (TB) (n = 30) or non-tuberculous pneumonia (n = 29). Serum samples drawn at the time of diagnosis and one year before. Anti-purified protein derivative (PPD) and anti-diacyltrehalose (DAT) serum antibodies quantified by ELISA assay. OBJECTIVE: Since TB patients with HIV infection may present with elevated serum antibodies against Mycobacterium tuberculosis, we hypothesized that TB-specific antibody markers might predict TB in these subjects. DESIGN: A retrospective study was designed to assess the presence of M. tuberculosis-specific antibodies in HIV-positive patients developing TB. RESULTS: Of 30 HIV-positive TB patients, 24 (80%) had anti-PPD or anti-DAT antibodies at the time of TB diagnosis, and 20 (67%) one year before. In a sub-population of 16 of the 30 HIV-positive subjects, positivity for anti-PPD or anti-DAT antibodies one year before TB diagnosis was higher (11/16, 69%) than for the PPD skin test (4/16, 25%, P < 0.01). Antibody tests were specific for TB since positivity rates were lower both in patients with non-tuberculous pneumonia (P < 0.01) and in those with M. avium infection (P < 0.05). CONCLUSION: Antibody markers may predict TB in HIV-positive subjects, including those with negative PPD skin test.

Mathematical derivative applied to international normalised ratio and analytical variations in oral anticoagulant therapy control.

A reliable prothrombin time (PT) testing and a careful drug dosage can prevent thrombotic or bleeding complications of the oral anticoagulant therapy. The international normalised ratio (INR) as PT standardisation introduced an analytical variation that increases with higher PT measures and higher international sensitivity index (ISI) values. Our study was conducted to investigate the INR accuracy through the mathematical derivative application to reduce the analytical component of the INR uncertainty. The evaluation of accuracy among four different systems (prothrombin activity percent, PT seconds, PT ratio and INR) was determined by the simulation of a systematic error. Plasma samples were diluted 1:2; then they were compared observed with the expected values. We analysed the calculation system of the INR through the mathematical derivative in 87 PT ratio measurements. The analytical incidence of thromboplastin ISI was performed through an elaboration of INR mathematical derivative considering 10 different ISI values ranging from 1.1 to 2. The data expressed as PT ratios revealed a lower systematic error propagation suggesting that a linear system is more accurate. According to the calculation formula of INR, analytical variability increases with the PT measurements, then with the intensity of anticoagulation. Mathematical derivative suggests that the INR uncertainty due to the ISI can be reduced using a thromboplastin reagent with a low ISI or with ISI close to 1.

Prevalence of pneumonia due to Legionella pneumophila and Mycoplasma pneumoniae in a population admitted to a department of internal medicine.

We studied 177 patients with pneumonia admitted to an internal medicine department over a period of 3 years to determine the incidence of two emerging pathogens, Legionella pneumophila and Mycoplasma pneumoniae. Clinical, radiological and laboratory tests were performed and included blood cultures, serology, gram staining and sputum cultures. L. pneumophila was the agent involved in 9 patients (5.1%) and M. pneumoniae in 12 (6.8%). These prevalences were about in the middle of the range of previously published figures. Legionella pneumonia is a rare illness, which even in the absence of suggestive clinical signs must be considered because of its possibly serious course and to allow appropriate therapeutic decisions to be made.