system at which we were able to target attempts at improvements. We would like to point out that we have a policy for telephoning all important results to the clinicians directly, so by-passing the written report.In future, the computerisation of the microbiology service with terminals in all clinical areas will lead to a more efficient turnaround time. selected patients whose blood and bone marrow were cultured for S typhi to determine the yield of S typhi from these two sites. MethodsBlood and bone marrow samples were collected from 100 patients who had fever of unknown origin. Each sample was collected in two blood culture bottles containing brain heart infusion and thioglycollate broth, respectively. About 5 ml of venous blood was inoculated in each bottle containing 45 ml of broth; similarly, 0-5 ml-1 0 ml of bone marrow was collected in each bottle containing 45 ml of broth. Bottles were sent immediately to the laboratory and incubated at 37°C for seven days. Each bottle was examined daily and subcultured on to blood agar and MacConkey's media after 24, 48, and 72 hours, and on the seventh day of incubation.Non-lactosing fermenting colonies from MacConkey's medium were tested for S typhi by slide agglutination test with specific antisera.56 Biochemical tests were performed by using API 20E strips. ResultsThe samples of blood and bone marrow were processed at The Aga Khan University Hospital's diagnostic laboratory. S typhi was isolated from the blood (n = 58) or bone marrow (n = 88). The blood and bone marrow cultures of 12 patients showed no growth for any bacteria.Thirty blood samples from 88 patients (whose bone marrow were positive) showed no bacterial growth.

No abstract

The case records of all neonates admitted to the neonatal unit at Aga Khan University Hospital (Karachi) in a 30 month period (Nov. 86-April 89) were analysed. Of 60 neonates with confirmed sepsis, 33 (55%) had non-nosocomial infection (NNC) whereas 27 (45%) had nosocomial sepsis (NC). The most common organisms causing early-onset NNC sepsis were Klebsiella species (53%) and Escherichia coli (10%), whereas the organisms causing late-onset NNC sepsis included Salmonella parathypi (21%), Group A Streptococcus (21%), Escherichia coli (14%) and Pseudomonas species (14%). Klebsiella was the most common organism causing NC sepsis, others being Staphylococcus aureus (15%) and Serratia species (15%). The mortality in NC sepsis, early-onset and late onset NNC sepsis was 44%, 26% and 43%, respectively. Risk factors associated with NNC sepsis included low birthweight, prematurity and prolonged and complicated deliveries. There was a high incidence of drug resistance to ampicillin and gentamicin among gram-negative organisms causing sepsis (mean 67%).

No abstract

Treatment of children with infections caused by Salmonella typhi strains resistant to the commonly used oral antimicrobials is a special problem. As children cannot be treated with quinolones, there is no form of oral therapy. Third generation cephalosporins, which have been shown to be effective against typhoid caused by ampicillin sensitive strains of S. typhi were effective against typhoid caused by ampicillin, chloramphenicol and sulfamethoxazole/trimethoprim-resistant strains. We treated 28 children with ceftriaxone and 8 with cefotaxime. We found ceftriaxone to be more effective than cefotaxime with significantly lower relapse rate. Antibiotic therapy of 19 other children, initially treated in a similar manner, was altered for ease of therapy or due to poor response to therapy. The high cost of this parenteral therapy and the problems in its delivery point to the need for safe, effective oral therapy.