Sepsis Contributes To as Many as Half of All Hospital Deaths in U.S., Study Says
Embargo expired: 18-May-2014 5:00 PM EDT
Source Newsroom: American Thoracic Society (ATS)
Newswise — ATS 2014, SAN DIEGO ─ Although it’s present in only a tenth of all patients who are admitted to the hospital, sepsis contributes to up to half of all hospital deaths in the U.S., according to a study presented at the American Thoracic Society’s annual conference here.
Although many studies have examined the incidence and mortality of patients diagnosed with sepsis in the U.S. over time, the study authors say that so far, the impact of sepsis on overall hospital mortality has been poorly understood.
“Our study was designed to quantify the national impact of sepsis on hospitalized patients and to highlight the importance of sepsis care on mortality at a population level,” said study lead author Dr. Vincent Liu, M.D., M.S, of the Kaiser Permanente Northern California Division of Research.
For their study, the researchers conducted a retrospective analysis of 6.5 million hospital discharge records derived from the Healthcare Cost and Utilization Project Nationwide Inpatient Sample (NIS) in 2010. The NIS is the largest publicly-available, all-payer inpatient database in the U.S., containing data from 100% of hospital discharges from a stratified sample of community hospitals.
Using diagnosis and procedure codes, the researchers identified hospital admissions and deaths of patients with sepsis and estimated the percentage of total hospital charges associated with sepsis hospitalizations.
At the end of their analysis, they found that sepsis patients had a hospital mortality rate of 10.4% compared to a rate of 1.1% in patients who did not have sepsis. They also found that of all hospital deaths nationally, as many as 52.0% were among patients diagnosed with sepsis. Sepsis hospitalizations also contributed to 21.2% of all hospital charges.
Sepsis is a serious medical issue in the U.S., affecting as many as 750,000 hospitalized patients in the U.S. annually, according to data from the Centers for Disease Control and Prevention (CDC). Between 2000 and 2010, deaths due to septicemia increased 17% despite an overall decline in overall hospitalization rates, and hospital deaths due to septicemia increased from 45,000 to 135,000 during the same period.
According to the Agency for Healthcare Research and Quality (AHRQ), sepsis is the most expensive cause of hospitalization in the United States, accounting for more than $24 billion annually.
“We were surprised to find that as many as 1 in 2 patients dying in US hospitals had sepsis. Teasing apart these findings in a similar regional study of sepsis mortality at Kaiser Permanente Northern California, we found that most patients already had sepsis at the time of hospital admission. There was also a large number of patients with less severe sepsis, a group for whom treatment guidelines are less well-defined,” Dr. Liu said. “The results of our study suggest that improved care for sepsis patients of all severity levels and in all hospital settings could result in many future lives saved.”
* Please note that numbers in this release may differ slightly from those in the abstract. Many of these investigations are ongoing; the release represents the most up-to-date data available at press time.
Sepsis Contributes To Nearly Half Of All Hospital Deaths In The US
Type: Scientific Abstract
Category: 04.12 - Sepsis (CC)
Authors: V. Liu1, G.J. Escobar1, A. Whippy2, D.C. Angus3, T.J. Iwashyna4; 1Kaiser Permanente Division of Research - Oakland, CA/US, 2Kaiser Permanente - Oakland, CA/US, 3University of Pittsburgh - Pittsburgh, PA/US, 4University of Michigan - Ann Arbor, MI/US
Rationale: Sepsis is the most expensive cause of hospitalization in the United States according to AHRQ; prior work suggests that it contributed to nearly half of hospital deaths in an integrated healthcare delivery system in Northern California. However, the national impact of sepsis on overall hospital mortality remains poorly characterized.
Methods: We conducted a retrospective analysis of data from the Healthcare Cost and Utilization Project Nationwide Inpatient Sample (NIS) in 2010. The NIS is the largest publically available all-payer inpatient database and contains data from 100% of hospital discharges in an approximately 20-percent stratified sample of US community hospitals. We identified cases of severe sepsis and septic shock with ICD-9-CM diagnosis and procedure codes based on the Angus implementation. We evaluated patient and hospital characteristics collected in the NIS sample. We estimated the fraction of all hospital admissions and deaths occurring in patients with sepsis. We also estimated the percentage of total hospital charges associated with sepsis hospitalizations.
Results: Our sample included 7.8 million hospital discharges from 1,051 hospitals in 2010. Overall, 8.3% (n = 644,570) of all discharges met criteria for sepsis. Sepsis patients were older and more likely to have chronic conditions compared with the non-sepsis population (p<0.01). Mean length of stay for sepsis hospitalizations was 10.7 ± 14.0 days versus 4.1 ± 5.6 days for others (p<0.01).
Hospital mortality was 11.0% for sepsis patients compared with 1.1% for other inpatients (p<0.01). Across hospitals, median sepsis mortality was 10.3% with an interquartile range (IQR) of 7.6% to 13.4%. In the full 2010 sample, sepsis was present in 71,141 deaths, or 48.0% of all hospital deaths. The fraction of all hospital deaths resulting from sepsis varied across hospitals (median: 42.9%; IQR: 28.6% to 53.2%). Sepsis patients who died in the hospital were older than those who survived (70.9 ± 16.9 versus 66.4 ± 18.8 years) but had slightly fewer comorbid conditions (6.6 ± 3.2 versus 6.7 ± 3.3, p<0.01 for both). In total, sepsis hospitalizations contributed to 21.2% of all hospital charges.
Conclusion: While sepsis is present in less than 1 in every 12 US hospitalizations, it contributes to nearly 1 in 2 hospitals deaths. In light of the limitations resulting from the use of administrative data, these figures may be conservative estimates.