T imely administration of antibiotics in the PICU can be life-saving. Recent initiatives, based on data supporting the survival benefit of antibiotic delivery within 1 hour, have addressed delays and minimized barriers (1, 2). Although controversy may remain about the strength of this evidence, I doubt few of us would object to an emphasis on early antibiotics for a patient with suspected septic shock. What to do when the initial "rule-out" is done, especially when bacterial cultures remain negative-that is a common conundrum addressed in this issue of Pediatric Critical Care Medicine, where Fontela et al (3) offer theoretical models describing clinical decision-making about antibiotic use, based on their study of 21 PICU practitioners.Using qualitative methods, the authors sampled physicians from three PICU sites in Canada, ranging in experience, gender and level of training. Using the theory of grounded behavior, they constructed models that describe the process these PICU practitioners used to make decisions about antibiotic initiation, discontinuation, and course duration. The three models they describe contain four common steps. The first step, as one might suspect, was the evaluation of objective evidence (laboratories, vital signs, imaging studies, as well as an individual patient's risk for overwhelming bacterial infection) to determine antibiotic choice and duration. If uncertainty remained after an objective evaluation, practitioners then turned to a patient safety assessment, described as a moral process incorporating the desire to avoid bad patient outcomes, potential medical-legal impacts, and was highly informed by the practitioner's risk tolerance.Step three follows if uncertainty continues-an intuitive approach that is strongly influenced by the practitioner's own experiences with overwhelming infections. The final step is the deference to expert consultants either published guidelines, if applicable, or more commonly our pediatric infectious disease colleagues (3).In addition to the four-step models, the authors found other factors, like their trust in colleagues' prior decision about antibiotics, were also influential (3). This theme echoed findings of another qualitative study of inpatient antibiotic decision-making by Livorsi et al (4), who also reported that antibiotic overuse is recognized but accepted, and that potential for adverse effects of antibiotic use has limited influence on decision-making. The profound influence of supervisors on their trainees regarding antibiotic practices, as well as reluctance to give feedback to colleagues even when they might critique their decisions, were also important themes (3,4).As the authors highlight, there is an opportunity to improve the reliability of objective data through future study, and research there remains active. Until