Inadequate empirical antibiotics following debridement for orthopedic infections do not increase therapy failures

<b>Introduction</b>: Empirical antibiotics should only target the most likely pathogens if antibiotic stewardship is being heeded. However, there is a drive for broader-spectrum empirical antibiotics in orthopedic infections due to the concern of therapeutic failure if a regimen fails to target subsequently identified pathogens. <b>Methods</b>: Retrospective case-control study with surgically managed orthopedic infections from July 2018 to June 2024 with a minimum follow-up of 6 months. Patients were stratified by the initial empirical treatment of either accurate empirical choice or inaccurate empirical choice. <b>Results</b>: Of 482 infection episodes, 79 antibiotic regimens (43 broad-spectrum; 9 %) were used with a median postoperative duration of 42 d (interquartile range 19-45 d); 290 infection episodes (60 %) were correctly targeted. In 192 cases (40 %), the initial empirical choice was inaccurate, with a median switching time to a targeted treatment of 4 d. There was no difference between accurate and inaccurate empirical treatment in terms of ultimate failures (18/290 vs. 15/192; Pearson <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:mrow><mml:msup><mml:mi>χ</mml:mi> <mml:mn>2</mml:mn></mml:msup> </mml:mrow> </mml:math> test, <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"><mml:mrow><mml:mi>p</mml:mi> <mml:mo>=</mml:mo> <mml:mn>0.49</mml:mn></mml:mrow> </mml:math> ), overall adverse events of therapy (15 % vs. 7 %, <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"><mml:mrow><mml:mi>p</mml:mi> <mml:mo>=</mml:mo> <mml:mn>0.11</mml:mn></mml:mrow> </mml:math> ), duration of hospital stay (median 9 d vs. 9 d, <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"><mml:mrow><mml:mi>p</mml:mi> <mml:mo>=</mml:mo> <mml:mn>0.96</mml:mn></mml:mrow> </mml:math> ), or supplementary surgical debridement (median 0 vs. 0 intervention, <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"><mml:mrow><mml:mi>p</mml:mi> <mml:mo>=</mml:mo> <mml:mn>0.58</mml:mn></mml:mrow> </mml:math> ). In multivariate logistic regression analysis, the duration of an inaccurate antibiotic treatment failed to alter the risk of "failures" (odds ratio 0.9, 95 % confidence interval 0.8-1.1). <b>Conclusions</b>: A delay in commencing targeted antibiotics does not increase the risk of a negative outcome. Narrower-spectrum empirical regimens are appropriate for clinically mild to moderate infections as a broader spectrum does not provide any clinical advantage.