Culture Negative Severe Sepsis

Shipra Gupta, Ankit Sakhuja Gagan Kumar et el published a retrospective study of culture negative sepsis in December, 2016  issue of Chest.



Nationwide Inpatient Sample (NIS) database (USA) from 2000 to 2010, total of 6,843,279 admissions of patients with severe sepsis were studied, 3,226,406 (47.1%) had culture-negative results.



  1. Incidence was higher in those < 1 year or > 70 years of age and in white patients, more often admitted to nonteaching hospitals. (55% vs 50.6%; P < .001),
  2. Patients with culture negative severe sepsis had a higher comorbidity burden as evidenced by a higher CCI (28.2% vs 26.5% with CCI score > 3; P < .001) (CCI = Charlson comorbidity index score),
  3. Patients had a higher number of acute organ system dysfunctions (27.0% vs 23.1% with more than three acute organ system dysfunctions; P < .001).
  4. Patients were more likely to have acute respiratory failure (52.1% vs 46.7%; P < .001), need mechanical ventilator support (37.7% vs 35.7%; P < .001),
  5. Patients were more likely to have septic shock (41.1% vs 35.5%; P < .001), cardiovascular failure, hepatic failure, and metabolic dysfunction



Culture-negative sepsis was an independent predictor of mortality in those with severe sepsis (OR, 1.75; 95% CI, 1.72-1.77)



During the period of study, due to surviving sepsis campaign, there was improvement in the proportion of blood being drawn for cultures before administration of antibiotics.  This should have been expected to improve culture positivity rates for these admissions; however, there was  a decrease in culture positivity rates over time.



Possible reasons for the difference between these two categories.

  1. Administration of antibiotics early without performing appropriate cultures, leading to culture-negative status
  2. Delay in, or shorter duration of, antibiotic therapy or potentially inappropriate antibiotic therapy for lack of guidance from specific cultures.
  3. Severe sepsis goes unrecognized in patients with culture-negative status until later in the course of hospitalization



NIS database did not allow to  capture processes of care and details of the acute illness such as vital signs, laboratory data, drug administration, timeliness of antibiotic therapy, appropriateness of cultures, and timing of source control. Also it did not  have complete data on patient race.


Pantoprazole reduces GI bleeding in intubated patients without improving mortality.Read more