Revising the Decision Analysis for Febrile Children at Risk for Occult Bacteremia in a Future Era of Widespread Pneumococcal Immunization

Publisher Website
Google Scholar
Abstract
Previous decision analyses (DA), which have attempted to determine a superior management option for febrile children at risk for occult bacteremia, have come to different conclusions based on their underlying assumptions. Most DAs have recommended a laboratory evaluation with antibiotic treatment or have concluded that all management options are roughly the same. The purpose of this revised DA is to determine which management strategy will be superior in an anticipated era of widespread pneumococcal immunization. Decision analysis methodology comparing the following 3 management options: Obs: observation, no tests, no antibiotics; CBC+: Complete blood count (CBC) first; if the white blood cell count (WBC) >15,000, then blood culture (BC) plus antibiotics (Abx); BC+Abx: BC+Abx for all patients. Assumptions include a current bacteremia rate of 2% to 4%, and 95% of this bacteremia is caused by S. pneumococcus (which has a 15% persistent bacteremia rate) with the remainder caused by virulent bacteria (which has a 100% persistent bacteremia rate), and other assumptions made by previous decision analysis publications. Pneumococcal vaccine efficacy rates of 0%, 50%, 75%, and 100% in preventing pneumococcal bacteremia were analyzed. Overall death and neurologic sequelae rates are lower with efficacious pneumococcal vaccine. The Obs strategy is superior for all efficacy rates (including 0%) if the negative consequences of treatment are high. If the negative consequences of treatment are low, the CBC+ strategy is superior at pneumococcal vaccine efficacy rates of 50% or greater. The absolute differences between the outcomes of the 3 strategies are small, making them similar. This decision analysis indicates that widespread efficacious pneumococcal vaccine will reduce the overall morbidity and mortality associated with pneumococcal bacteremia. This favors the Obs strategy, but the superior management strategy is still dependent on an assumption of the negative consequences of treatment. Since the outcome measures for each of the management strategies are similar numerically, the strategy chosen in clinical practice may consider other factors associated with the clinical encounter such as those described.