Economic implications of optimal diagnosis and treatment of sepsis - Work in progress: Marginal penalties, antibiotic alterations, and outcome hypotheses

Our goals: (a) to design value strategies for optimal diagnosis and treatment of sepsis; (b) to assess theoretical marginal penalties of inadequate antimicrobial treatment of intensive care patients with infections (case study) ; and (c) to analyse physician antibiotic alterations in patients with blood culture-proven bloodstream infections. Marginal penalties, which reflect extra procedures and excess costs arising from uninformed treatment decisions, were transformed to subject group-distributed opportunity costs. Value analysis revealed substantial marginal penalties associated with adverse factors, such as increased ICU length of stay, procedures (catheterisation, mechanical ventilation, tracheostomy) and ultimately, higher mortality, in critically ill patients. Investigation of 66 septicaemica patients (International Classification of Disease [ICD-9] code 038 and related derivatives) hospitalised over 17 months revealed that physicians altered antibiotics extensively during three time intervals: TI₁ - collection of blood for culture to qualitative positive blood culture notification, TI₂ - notification to final MIC result, and TI₃ - final MIC result to 72 hours afterward. Empirical antibiotic alterations during TI₁ may have adversely affected survival. Alterations peaked after blood collection and after notification of qualitative positive blood culture results. Based on patterns of alterations and 28-day mortality, we hypothesize that nucleic acid testing, if used to identify organisms and rule in bloodstream infections early (4-6 h) following admission, will help facilitate diagnosis, focus antibiotic therapy, and avert dysfunctional sepsis cascades. Reduced "alteration debt" and marginal penalties should offset the costs of nucleic acid testing. The potential for enhanced survival and improved outcomes warrants clinical trials of rapid nucleic acid testing to decrease indiscriminate antibiotic alterations, evaluate proposed value strategies, and test outcome hypotheses.