PURPOSE:To assess incidence and mortality rates of severe sepsis (SS) 2004–2006 and assess the impact of statin use on SS mortality.
METHODS:We conducted a retrospective study based on hospital discharge and charge master information of ∼5 million annual discharges from ∼500 US hospitals between 2004–2006 (Premier Perspective Database). SS was defined via ICD-9 coding using the method of Dombrovskiy (Crit Care Med 2007).
RESULTS:The number of SS cases in this database ranged from 64000 in 2004 to 89000 in 2006. The percentage of SS cases among all hospitalizations increased each year from 1.4% in 2004 to 1.7% in 2006. Similarly, the proportion of SS among ICU admissions increased from 10.2% to 12.4%. In-hospital mortality among all SS patients decreased each year (35%, 33%, 31%). The percent of patients with charge-coding for dispensing a statin (statin=Yes) during the hospitalization for SS increased each year from 13% in 2004 to 18% in 2006. Each year SS patients in the statin=Yes group showed a lower mortality rate compared to those without charge for a statin (2004 28% vs 35%; 2005 26% vs 35%; 2006 23% vs 33%). The Statin=Yes group had a greater prevalence in patients with one organ dysfunction and a general negative correlation with increasing number of organ failures.
CONCLUSION:The incidence of SS increased but mortality rates decreased between 2004–2006 perhaps reflecting improvements in general sepsis treatment. In-hospital coding for a statin was consistently associated with a lower mortality rate and pattern of fewer organ failures. A cause/effect relationship cannot be confirmed from this dataset; the association may be causal but statin use may be a marker for another effect or general health status. In-hospital statin use may indicate out of hospital (pre-sepsis) statin use which could be contributing to the effects described. Nevertheless, the association is consistent in magnitude and across several years.
CLINICAL IMPLICATIONS:These data support the growing body of evidence implicating a protective effect of statins in SS and substantiates consideration of further investigation.
DISCLOSURE:Steven Simonson, Shareholder Shareholder in AstraZeneca.; Employee Employee of AstraZeneca.; No Product/Research Disclosure Information