Critical illness from 2009 pandemic influenza A virus and bacterial coinfection in the United States

Abstract

Objectives: The contribution of bacterial coinfection to critical illness associated with 2009 influenza A virus infection remains uncertain. The objective of this study was to determine whether bacterial coinfection increased the morbidity and mortality of 2009 influenza A.

Design: Retrospective and prospective cohort study.

Setting: Thirty-five adult U.S. intensive care units over the course of 1 yr.

Patients: Six hundred eighty-three critically ill adults with confirmed or probable 2009 influenza A.

Interventions: None.

Measurements and main results: A confirmed or probable case was defined as a positive 2009 influenza A test result or positive test for influenza A that was otherwise not subtyped. Bacterial coinfection was defined as documented bacteremia or any presumed bacterial pneumonia with or without positive respiratory tract culture within 72 hrs of intensive care unit admission. The mean age was 45±16 yrs, mean body mass index was 32.5±11.1 kg/m, and mean Acute Physiology and Chronic Health Examination II score was 21±9, with 76% having at least one comorbidity. Of 207 (30.3%) patients with bacterial coinfection on intensive care unit admission, 154 had positive cultures with Staphylococcus aureus (n=57) and Streptococcus pneumoniae (n=19), the most commonly identified pathogens. Bacterial coinfected patients were more likely to present with shock (21% vs. 10%; p=.0001), require mechanical ventilation at the time of intensive care unit admission (63% vs. 52%; p=.005), and have longer duration of intensive care unit care (median, 7 vs. 6 days; p=.05). Hospital mortality was 23%; 31% in bacterial coinfected patients and 21% in patients without coinfection (p=.002). Immunosuppression (relative risk 1.57; 95% confidence interval 1.20 -2.06; p=.0009) and Staphylococcus aureus at admission (relative risk 2.82; 95% confidence interval 1.76-4.51; p<.0001) were independently associated with increased mortality.

Conclusions: Among intensive care unit patients with 2009 influenza A, bacterial coinfection diagnosed within 72 hrs of admission, especially with Staphylococcus aureus, was associated with significantly higher morbidity and mortality.

Figure 1

a. Timing of pH1N1 Cases. Number of influenza cases by testing result by month. b. pH1N1 Cases by Decades of Age. Number of cases by decade of age, divided into survivors and non-survivors.

Figure 1

a. Timing of pH1N1 Cases. Number of influenza cases by testing result by month. b. pH1N1 Cases by Decades of Age. Number of cases by decade of age, divided into survivors and non-survivors.

Figure 2. Funnel Plot of Survival and Ventilator Days in pH1N1 Patients with and without Bacterial Co-Infection

The top and bottom black lines represent survival and liberation from mechanical ventilation, respectively, in patients with pH1N1 and no co-infection. The top and bottom gray lines represent the same thing in patients with pH1N1 and bacterial co-infection. Patients still on the ventilator are censored at time of death for liberation from mechanical ventilation plots.