[An analysis of 99 consecutive shifts in the selection of antibiotics used in therapy in a medical department].

Autor: Siboni AH; Medicinsk afdeling, Sundby Hospital, København., Holmegaard SN
Jazyk: dánština
Zdroj: Ugeskrift for laeger [Ugeskr Laeger] 1993 Nov 22; Vol. 155 (47), pp. 3840-4.
Abstrakt: During 208 days 2836 patients were admitted to Sundby Hospital, medical ward. A total of 734 antibiotic cures were initiated. About 632 (22-23%) of the patients had antibiotic treatment. Penicillin, ampicillin and sulfamethizole were the most frequently used antibiotics. The use of erythromycin was 42-50% and 64-78% of that of penicillin and ampicillin respectively. Antibiotic treatment was changed in 99 cures in 73 patients. Fifty-seven of ninety-nine (43-71%) shifts were based on culture or serology and 42/99 (29-57%) shifts were based on clinical evaluation including microscopy and urinary stix. In 26 of the latter 42 cases positive culture or serology was obtained after the antibiotic was changed. Thirteen of the 26 shifts improved treatment, six were unlucky and seven indifferent, thus giving a net advantage of 13-6 = 7 of 26 shifts. This net advantage was due to shifts from penicillin. In ten cases the antibiotic was shifted to erythromycin due to suspected atypical pneumonia, but only one case was verified. In ten antibiotic shifts in pneumonia patients the etiologic agents were not identified. In 4/99 shifts in 4/73 patients relevant specimens were not obtained. The antibiotic most often changed compared with its total use was ampicillin (32/152 congruent to 1/5 of initiated cures) and shifts from ampicillin were more often (21/32) based on culture than shifts from penicillin (16/36) (p < 0.1). Initial supplementation of ampicillin with aminoglycosides was retrospectively relevant in 13/28 cases. Cephalosporines (p < 0.001) and dicloxacillin (p < 0.02) were significantly more often used as second drug, whereas penicillin was most often used as first drug (p < 0.001).(ABSTRACT TRUNCATED AT 250 WORDS)
Databáze: MEDLINE