Search

Epidemic, louse-borne typhus persists in the rugged, mountainous areas of Ethiopia and much of northeastern and central Africa as well as in the rural highlands of Central and South America, where the conditions of living favor the harboring of body lice and where antibiotic treatment and effective louse-control measures are unavailable. The historical significance and current epidemiology of typhus, including the reservoir of Rickettsia prowazekii in flying squirrels in the United States, are reviewed, and the clinical presentation, laboratory findings, and hospital course in the cases of 60 patients admitted with epidemic, louse-borne typhus to the St. Paul's Hospital in Addis Ababa, Ethiopia, are described. Treatment of this disease with oral doxycycline, tetracycline, or chloramphenicol prevents complications and results in prompt resolution of symptoms.

One thousand six surgical patients admitted between April 1983 and January 1984 to a hospital in Addis Ababa were studied for incidence of nosocomial infections. On admission, they were carefully examined clinically to exclude community-acquired infections and to determine any underlying risk factors. All patients were closely followed up for detection of developing nosocomial infections. Specimens collected from patients with clinically diagnosed bacterial infections were processed for culture; isolates were identified and tested for susceptibility to the locally used antimicrobial agents. Nosocomial infections were detected in 165 (16.4%) patients. Wound (59%), urinary tract (26%), and respiratory tract (6%) infections accounted for more than 90% of the infections. Fourteen of 18 deaths were attributed to nosocomial infections. About 72% of the patients were given prophylaxis, which was associated with infection (p less than 0.0005). Operations and other procedures were also associated with infection (p less than 0.0005). Anemia was found as a host risk factor (p less than 0.0005). Approximately 90% of the nosocomial pathogens were gram-negative bacteria, of which 84% were Enterobacteriaceae. They were mostly resistant to the commonly used antibiotics. Our findings should create awareness of the problem in this hospital and possibly in others in Addis Ababa and hence prompt measures for effective control.

Patients admitted to a teaching hospital in Addis Ababa were studied for nosocomial infection. Of 2506 patients, 13% developed clinical infections, with the highest rate among obstetric/gynaecologic patients (17.0%). Wound infection was the most frequent type of nosocomial infection (49%) followed by urinary tract infection (25%). Antibiotic prophylaxis was given to 43% of the patients. Gram-negative bacteria comprised 88% of all isolated strains Enterobacteriaceae; 75% of all isolates were found in over 60% of the infection, Proteus 25%, Escherichia coli 20% and Klebsiella 19%. The most widely used antibiotics were ineffective against 65 to 85% of the Gram-negative strains. Cefotaxime and gentamicin were more effective. Staphylococcus aureus isolates were also similarly resistant, against which cephalothin, lincomycin and gentamicin were the more effective ones. Over 70% of the strains were multiple resistant. The findings underscore the need for a surveillance program and infection control system to reduce the high rate of infection and to institute appropriate guidelines for the use of prophylactic and therapeutic antibiotics.

Seven hundred patients admitted to the wards of the Department of Obstetrics/Gynaecology of Tikur Anbessa Hospital, Addis Ababa, between January and July 1984 were studied for the incidence of nosocomial infections. The overall hospital infection rate was 17.0%, with wound infection at 47% followed by urinary-tract infection at 15%. Over 80% of the bacterial isolates were Gram-negative bacteria, the majority of which were multiply resistant to the commonly used antibacterial drugs. Some hospital risk factors have been identified. The findings are discussed in the light of reports on nosocomial infections from hospitals of developed countries.

A total of 140 Ethiopian men with gonococcal urethritis were randomly assigned to treatment with aqueous procaine penicillin G (4.8 X 10(6) units intramuscularly [im] plus 1.0 g of oral probenicid); oral ampicillin (3.5 g plus 1.0 g of oral probenicid); spectinomycin (2.0 g im); or oral rosoxacin (Acrosoxacin; 300 mg). Failure rates were 24%, 19%, zero, and 3%, respectively. Forty-four (31.4%) patients were infected with penicillinase-producing Neisseria gonorrhoeae (PPNG) and were evenly distributed in the treatment groups. All 39 PPNG strains analyzed for plasmid content possessed a 2.6-Mdalton plasmid; 28 (71.8%) had a 3.2-Mdalton beta-lactamase-encoding plasmid, ten (25.6%) had a 4.4-Mdalton plasmid (three with and seven without a 24.5-Mdalton plasmid), and one had only a 24.5-Mdalton plasmid. Two patients were infected with N. gonorrhoeae-possessing plasmids apparently capable of encoding but not producing beta-lactamase. Both spectinomycin and rosoxacin are excellent single-dose treatment regimens for gonococcal urethritis in men. All people receiving these drugs in Ethiopia should be tested serologically for syphilis, however, as eight (11.8%) of 68 men in this study also had active latent syphilis.