Antibiotic treatment of staphylococcal septicaemia and endocarditis in a Swedish hospital.

In a retrospective study covering the years 1977 to 1981, the results of antibiotic treatment in 123 patients with staphylococcal septicaemia with or without endocarditis have been analysed. 80 patients (mean age 60 years) were non-drug addicts (Group I) and 43 (mean age 28 years) were drug addicts (Group II). Underlying conditions other than drug abuse were noted in 74 patients in Group I and in only 7 in Group II.S. aureus was isolated from 117 patients and S. epidermidis in 6, all of them in Group I. 91 strains were penicillinase producers, but all susceptible to isoxazolyl-penicillins. In Group I verified or highly suspected endocarditis was registered in 12 patients (15%), always left-sided, as against in 31 (72%) in Group II, of whom 25 had tricuspid valve engagement. In the multivariate pattern of antibiotic treatment 3 groups may be discerned; 1) Cloxacillin, alone (35 patients) or in a combination (57), 2) Penicillin G, alone (6) or in a combination (12), and 3) Lincomycin or clindamycin, a cephalosporin or co-trimoxazole, alone (4) or in combination (9). Additive agents were mostly an aminoglycoside or fusidic acid. Out of the 45 patients in the whole material who received single therapy 9 patients (20%) died, and out of the 78 patients who received combined therapy 13 patients (16.6%) died. In the cloxacillin group 11.8% died, compared to 35% who initially received other antibiotics. In 70 patients the initial therapy had to be changed, in 39 due to adverse drug reactions and in 31 due to therapeutic failures or for unexplained reasons. In these cases linco- or clindamycin, more rarely rifampicin or vancomycin, were used. In Group I, 20 patients (25%) died, 8 of them with endocarditis. Sequels, relapses or reinfections were noted in 21 (25%), and 39 (50%) had an uneventful course. In Group II, 2 patients (5%) died, both with endocarditis. Sequels, relapses or reinfections occurred in 11 (25%), and 30 (70%) had an uneventful course. From this unstructured material no definite conclusions can be drawn. However, the lower mortality rate in the cloxacillin group suggests this regimen to be superior. The addition of other antibiotics did not appear to influence the clinical outcome. There was a more favourable outcome in addicts than in non-addicts, despite the same general principles of antibiotic treatment. Thus, for the outcome the characteristics of the patient group seemed to have more influence than the choice of antibiotic treatment.