Diazepam is still one of the most used of the benzodiazepine group of drugs. Extensive studies over 10 years have defined a fairly complete profile of its kinetics. Minor aspects relating to patterns of its metabolism and excretion in certain age groups and in some disease states remain to be described quantitatively. However, there is more than sufficient kinetic information available for the requirements of good clinical practice. For optimum clinical benefit with minimum side-effects the following kinetic properties should be borne in mind: (a) there is a large interindividual variation (up to 30-fold) in dose/blood level ratios, especially when treatment is short-term; (b) the elimination half-life is prolonged in the elderly and the newborn and in some cases of liver disease; (c) there is accumulation of the active N-desmethylated metabolite during long-term treatment; and (d) administration of diazepam to pregnant women leads to rapid distribution from the maternal to fetal compartment: accumulation of both diazepam and desmethyldiazepam could cause prolonged sedation in the newborn. As there does not appear to be any clear relationship between the concentration of diazepam in the plasma and clinical effect, measurement of blood levels, other than for research purposes, is unnecessary. Based on kinetic data, a single administration of diazepam at night should be adequate for hypnotic and anxiolytic effects in most patients.

A prospective multicenter study concerning the incidence, onset time, risk factors and mortality of pneumonia was carried out by the Intensive Care Units Collaborative Group for Infection Control in Lombardy, Northern Italy. Out of 1304 patients admitted over 3 months in 16 intensive care units (ICUs), 441 met the criteria for the protocol (no previous pulmonary infection or irreversible terminal illness, ICU stay greater than 48 h). The incidence of acquired pneumonia was 21.3% (94/441), with 54.2% of cases diagnosed within 4 days of admission (early onset pneumonia). Impairment of airway reflexes on admission and more than 24 h respiratory assistance were shown as significant risk factors (RR) for early onset pneumonia (respectively RR = 12.4, with 95% confidence interval (CI) = 5.3-28.9 and RR = 3.3, with 95% CI = 1.8-5.9). A suggested pathogenetic mechanism is aspiration of oropharyngeal contents at the onset of acute illness, due to depression of protective reflexes with delayed clearance of bacterial contamination. No protection was offered by routinely applied prophylactic antibiotic therapy.

In 724 critically ill patients who had received prolonged (greater than 24 h) ventilatory assistance since admission to the ward, we analyzed the relationship between artificial ventilatory support and pulmonary infection. Two different approaches were used. The first, plotting the incidence of pneumonia versus the duration of ventilatory support, confirms previous data: the incidence rises from 5% in patients receiving one day of respiratory assistance to 68.8% in patients receiving more than 30 days. In the second approach, the same data were computed as an actuarial life table with the day of pneumonia onset as terminal event. This different approach, focusing on the onset of infection rather than on incidence, allows a new insight into the problem of nosocomial infection: pneumonia in patients with respiratory support is an early occurrence with a high and constant rate of acquisition and, therefore, a high risk of infection in the first eight to ten days. Later pneumonia acquisitions were rare, and the risk after ten days of ventilatory support is low.

Diazepam (DZ) placental transfer in pregnant women at term, following single or repeated drug administration by various routes, was evaluated. DZ and its metabolite N-demethyldiazepam (NDZ) were constantly present in umbilical cord plasma at concentrations comparable to the mother's shortly after drug administration. N-methyloxazepam (MOX) was detected in cord plasma only in a limited number of cases following chronic DZ treatment. Postmortem analysis of fetal tissue concentrations showed accumulation of NDZ in heart and lungs. Differences in NDZ concentrations between venous cord (VC) and arterial cord (AC) plasms suggest metabolic degradation of DZ in the fetus. The DZ apparent plasma half-life in the newborn was found to be longer (31 plus or minus 2 hr) than previously observed in infants and children. The low drug clearance appears to be linked to reduced urinary excretion of hydroxylated metabolites, suggesting limited capability to dispose of DZ in the newborn.

The study indicates that alcohol can be considered an independent risk factor for stroke in Italy.

Risk factors for pneumonia were analysed in a large population of critically ill patients, collected in two prospective multicentre "pneumonia studies" in Italy. Twenty-three intensive care units were involved and the study time was 150 unit months. Only patients without previous pulmonary infection, with intensive care unit stay greater than or equal to 48 hours and no rapidly irreversible illness at admission were included. The incidence of pneumonia in the 1475 selected patients was 15% (220 cases). 239 patients died in ICU; the mortality rate was significantly higher in patients developing pneumonia (p less than 0.0001); pneumonia was found to be an independent highly significant risk factor for death in critically ill patients (OR = 3.88; p less than 0.0001). Multivariate analysis of seven risk factors for pneumonia showed a significantly higher risk in patients with neuromuscular disease (OR = 3.8, p less than 0.002), impairment of airway reflexes at admission (OR = 2.93, p less than 0.0001), and greater than or equal to 24h respiratory assistance (OR = 3.05, p less than 0.0001). Impairment of airway reflexes at admission to the emergency room or intensive care unit identifies the population who will experience 3/4 of the overall lower respiratory tract infections. Rapid recognition of at-risk patients seems clinically important and may improve awareness programs and preventive approaches.

Diazepam (D Z) placental transfer in pregnant women at term, following single or repeated drug administration by various routes, was evaluated. DZ and its metabolite N-demethyldiazepam (NDZ) were constantly present in umbilical cord plasma at concentrations comparable to the mother's shortly after drug administration. N-methyloxazepam ( MO X) was detected in cord plasma only in a limited number of cases following chronic DZ treatment. Postmortem analysis of fetal tissue concentrations showed accumulation of NDZ in heart and lungs. Differences in NDZ concentrations between venous cord (VC) and arterial cord (AC) plasma suggest metabolic degradation of DZ in the fetus. The DZ apparent plasma half-life in the newborn was found to be longer (31 ± 2 hr) than previously observed in irifants and children. The low drug clearance appears to be linked to reduced urinary excretion of hydroxylated metabolites, suggesting limited capability to dispose of DZ in the newborn.

The pharmacokinetics of aztreonam were studied in ten critically ill intubated patients with lower respiratory tract infections. Serum and urinary concentrations of the drug and its penetration into bronchial secretions after a 2-g intravenous bolus were measured. Using a two-compartment linear model a terminal half-life of 1.87 + 0.46 h was determined. No interpatient differences were found for half-life values, AUC or volume of distribution, except in the case of one obese patient. The greatest variability was observed in the clearance values, and in particular the extrarenal clearance which ranged from 0.3 to 9.6 1/h. Maximum concentrations in bronchial secretions were reached very quickly in the 2 h after drug administration, with a range of 4.8-18.7 mg/l. No accumulation of aztreonam after repeated doses was detected.