OBJECTIVES: Adult severe sepsis is one of the most expensive conditions treated in hospitals and represents an economic burden to western health-care systems. However, there is presently no uniform standardized ICD-9-CM algorithm to identify severe sepsis at a population-level. Our aim is to compare two established diagnostic coding strategies in terms of hospital resource utilization. METHODS:  Population-based longitudinal study using the 2006-2011 nationwide hospital discharge database. Episodes of severe sepsis were identified by two ICD-9-CM coding strategies: a) those assigned explicit ICD-9-CM codes (995.92, 785.52); and b) those assigned combined ICD-9-CM infection and organ dysfunction codes by modified Martin criteria. We calculated hospital charges based on the National Health Service charges for DRGs. Trends were assessed as to annual percent changes (APC) using joinpoint regression analysis. RESULTS:  Roughly 62% of severe sepsis episodes identified were assigned explicit ICD-9-CM codes and 38% combination codes. Trend analysis showed that despite a shorter hospital stay in the explicit code cohort (12 vs