Three new cases of reactivation of Chagas' disease in patients with AIDS, with central nervous system and/or cardiac involvement, are reported. One patient had histological evidence of acute esophageal and gastric Trypanosoma cruzi myositis, a previously unrecognized finding in patients with reactivated Chagas' disease. The patients had a low CD4 lymphocyte count and had other AIDS-defining opportunistic infections. One patient's condition improved with benznidazole therapy. Analysis of these three cases and review of the 13 others published in the literature revealed that the central nervous system is the most commonly involved site (75%), followed by the heart (44%). Early diagnosis and treatment with benznidazole or nifurtimox probably improve the survival rate. Long-term secondary prophylaxis should be recommended for patients who respond to therapy, although it is uncertain which drug to use for this purpose. T. cruzi should be included in the list of opportunistic pathogens causing infection in severely immunocompromised patients with AIDS.

Further studies are required to clarify the risk of tattoos in transmitting infectious diseases through blood transfusions. A reassessment of tattoos as a screening criterion among blood donors is justified.

Tattoos have been shown to be associated with transfusion-transmitted diseases (TTDs), particularly hepatitis B virus (HBV) and hepatitis C virus (HCV) infections. Very little is known about the association between different categories of tattoos and TTDs. In a cross-sectional study in Brazil, we studied 182 individuals with tattoos and assessed the odds of testing positive for a TTD according to tattoo type, number, design and performance conditions. Major findings were significant associations between an increasing number of tattoos and HBV infection (odds ratio (OR) of 2·04 for two tattoos and 3·48 for [ges ] 3 tattoos), having a non-professional tattoo and testing positive for at least one TTD (OR = 3·25), and having [ges ] 3 tattoos and testing positive for at least one TTD (OR = 2·98). We suggest that non-professional tattoos and number of tattoos should be assessed as potential deferral criteria in screening blood donors.

Although it is well known that the signs and symptoms of dengue fever (DF) may overlap with those of other diseases, few attempts have been made to assess quantitatively their validity in the diagnosis of DF. The clinical data recorded, in the Brazilian town of Uberlândia between 1993 and 1998, on the notification forms from 495 serologically confirmed cases of DF and 650 patients who tested negative for this disease were therefore compared. Fever, headache, myalgia, retro-orbital pain and arthralgia were the most commonly reported symptoms among the confirmed cases (each occurring in at least 92% of the cases), whereas haemorrhagic manifestations (reported in 7% of cases) were uncommon. Although comparisons between the two groups gave risk ratios (RR) very close to 1 for most signs and symptoms, rash [RR=1.86; 95% confidence interval (CI)=1.63-2.13] and petechiae (RR=1.38; CI=1.07-1.79) were more commonly found among the confirmed cases. As the positive and negative predictive values of all of the signs and symptoms investigated were low, diagnosis of DF should not be based solely on clinical grounds.

A 5-year-old girl was bitten in her left eye by a lance-headed viper identified as Bothrops moojeni, measuring 115 cm of length. There was severe facial swelling and left exophthalmus, and enucleation of the eye was necessary. The patient apparently had mild systemic envenoming, but local inflammatory signs and histological evidence of necrosis suggest that both the mechanical trauma and the local action of the venom had a role in the genesis of the eye lesion. It is arguable if the loss of the eye could be prevented even if the antivenom was administered earlier.

A retrospective survey of 473 cases of snake bite admitted to a Brazilian teaching hospital from 1984 to 1990 revealed 91 cases of bite without envenoming and/or caused by non-venomous snakes. In 17 of these cases the snake was identified, and one patient was bitten by a snake-like reptile (Amphisbaena mertensii). In 43 cases diagnosis was made on clinical grounds (fang marks in the absence of signs of envenoming). The other 30 cases were of patients who complained of being bitten but who did not show any sign of envenoming or fang mark. Most cases occurred in men (66;73%), in the 10-19 years age group (26;29%), in the lower limbs (51/74;69%), between 6 A. M. and 2 P.M. (49;61%) and in the month of April (16; 18%). One patient bitten by Philodryas olfersii developed severe local pain, swelling and redness at the site of the bite, with normal clotting time. The patient bitten by Drymarcon corais was misdiagnosed as being bitten by a snake of the genus Bothrops, was given the specific antivenom, and developed anaphylaxis. One patient bitten by Sibynomorphus mikanii presented prolonged clotting time, and was also given antivenom as a case of Bothrops bite. Correct identification of venomous snakes by physicians is necessary to provide correct treatment to victims of snake bite, avoiding unnecessary distress to the patient, and overprescription of antivenom, which may eventually cause severe untoward effects.

The rational construction of an essential drug list, considering the patient's need, drug safety, availability and the best cost-benefit ratio, is based on drug safety, efficacy and quality. However, in daily practice, the prescriber's decision is mostly influenced by drug effectiveness, following criteria that increase adherence to the treatment, such as relative drug toxicity, convenience, cost and prescriber's experience. In addition, frequent launching of new molecules for the same therapeutic indication, together with wide publicity targeting prescribers, interferes with the decision-making process. Similarly, the bonuses offered by the industry for over-the-counter drug sales interfere with the consumer's choice. The confrontation between known human biological variability and the knowledge that there is no absolute similarity between drugs of the same therapeutic class, or even generic drugs, has an impact on the prescriber's drug list, which should include the concept of first and second choice drugs. Prescribers' unfamiliarity with these subjects is a determinant factor for irrational drug use: a public health issue. The objective was to introduce to drug prescribers information that can help them building up a rational drug list for their patients, based on the National Health Surveillance Agency (Anvisa) experience of drug regulation.

No abstract