Blood stream infections (BSI) lead to sepsis and septic shock, which continue to be important causes of morbidity and mortality. A vast body of literature has been devoted to this. Worldwide studies about the epidemiology, microbiological etiology and prognosis of BSI have been performed over the last six decades. These have shown changing patterns of the pathogens involved, namely the emergence of gram-negative pathogens as a major cause of bacteremia after the beginning of the antimicrobial era, 1,2 and the recent re-emergence of grampositive pathogens and fungi, especially among the growing population of immunocompromised patients. 3,4 While many studies are available from Western countries, mostly conducted in major teaching or research institutions, little work has been published from the developing countries. Extrapolation of data to hospitals in developing countries may be misleading because of the differing effects of factors such as endemic pathogens, underlying diseases, and antimicrobial prescribing habits.The present study aims to define the pattern of blood stream infections over a five-year period in a general referral hospital in a developing country, to compare it with previously published work from Saudi Arabia and other countries, to point out differences in the spectrum of micro-organisms, community-acquired and hospital-acquired BSI, and to determine the case fatality rate in relation to different micro-organisms and patient age groups.
Materials and MethodsThis study was carried out in Qatif Central Hospital (QCH). Qatif is located on the Gulf in the Eastern Province of Saudi Arabia, and the hospital is a 400-bed referral facility that provides all medical services. Data from the hospital's Department of Microbiology and from patients' records were reviewed. Patients with bacteremia and fungemia diagnosed by blood culture from May 1990 through May 1995 were identified and included in the study. The course of each episode of bacteremia or fungemia was followed until its conclusion.Ten and 3 mL of blood samples were drawn aseptically from adult and pediatric patients with suspected BSI, respectively. Each blood sample was divided into two bottles: NR6A (enriched soybean casein digest broth with CO 2 ) for aerobic cultures, and NR7A (prereduced enriched soybean casein digest broth with CO 2 ) for anaerobic cultures. Both bottles were incubated for seven days (eight weeks for brucella) at 37 C, and read daily on a Bactec NR 660 (Becton Dickinson). Subcultures on blood agar, chocolate agar and MacConkey agar were incubated aerobically with CO 2 , and subcultures on blood agar were also incubated anaerobically for 48 hours. A total of 23,704 paired bottles were processed during the study period.The significance of isolates was judged according to the growth in different blood cultures, the clinical findings and the isolates obtained from other body sites. Microorganisms isolated were identified by standard techniques 5 or by API 20 E system. An episode of bacteremia was defined by the isolation of...