מילות מפתח: Antibiotic resistance, Primary care, Causal inference, Empirical antibiotic treatment

רקע מדעי ומטרה:

Optimizing antibiotic prescription in community settings requires understanding the long-term consequences of treatment choices—particularly their effect on antimicrobial resistance. Current guidelines are often based on expert opinion rather than empirical patient-level data.

שיטות:

We conducted a retrospective analysis of 23,297 adult patients in Clalit Health Services who received one of four oral antibiotics—first-generation cephalosporins (1st CEP), second-generation cephalosporins (2nd CEP), amoxicillin-clavulanate (AMC), or fluoroquinolones (FQ)—between 2017 and 2019. The primary outcome was antibiotic resistance in the first urine culture obtained within six months post-treatment. To mitigate confounding and selection bias, we applied matching, post-matching adjustment, and inverse probability of censoring weighting (IPCW). Risk differences (RDs) were estimated using standardization.

תוצאות:

1st CEP exposure was associated with the lowest risk of future resistance across all antibiotic groups. Both AMC and FQ significantly increased resistance to the same class (AMC RD=8.2%, FQ RD=22.1%). Exposure to 2nd CEP resulted in greater resistance to 3rd CEP compared to FQ (RD=-2.1%, 95%CI: -3.9 to -0.4). Nitrofurantoin, used as a negative control, showed minimal variation across groups, supporting the validity of our estimates.

מסקנות:

This study provides robust, quantitative estimates of the resistance risks associated with commonly prescribed antibiotics in primary care. The findings support preferential use of 1st-generation cephalosporins where clinically appropriate, and reinforce existing stewardship efforts to limit fluoroquinolone and AMC use. These results may direct empirical treatment decisions in family medicine.

חשיבות לרפואת המשפחה:

This study provides family physicians with patient-level, real-world data to guide empirical antibiotic choices. These insights support more informed prescribing decisions