[Prognosis and treatment of bacterial endocarditis. Review of 70 cases]

The long term results of medical and surgical treatments have been reviewed in 70 cases examined during the last 8 years. In 24 of them, bacterial endocarditis developed after valvular replacement. Antibiotic treatment was administered during at least 6 weeks. Determinations of the minimum inhibitory concentration of the antibiotic and of the bactericidal power of the serum were considered as the most important laboratory controls. In the post-operative cases with prosthetic valves, numerous pathogenic organisms were found, including gram (minus) bacilli and monilia, which may partly account for the high mortality in this group. Despite surgical reinterventions, 20 of the 24 patients died. For the gram (+) cocci, prognosis is however statistically worse (at the 99% confidence limit) when prosthetic valves are present. Out of the 46 patients without prosthetic valves, 31 (68%) had a favourable bacteriological and haemodynamic evolution; in this group, gram (minus) bacilli and monilia were never found to be the pathogenic organisms. For streptococci infections, recovery was obtained in 73% of the cases (24 out of 33). Percentages of long term recovery are very similar in mitral (67%) and aortic (69%) valvular involvment, but the frequency of surgical indication was very different in either group. In cases of aortic valvular disease, valvular replacement was indicated for bacteriological or haemodynamic reasons in 19 of the 20 patients who recovered. Amongst the 9 patients who died, 6 were not operated upon. Valvular replacement is by far less frequently necessary in cases of bacterial endocarditis on the mitral valve : one prosthetic valve amongst 8 patients who recovered. Finally, the authors discuss the place of surgery in the treatment of bacterial endocarditis. According to their experience, recent improvement in the prognosis of the disease seems to depend more on the surgical approach of the problem than on the discovery of new antibiotics.