Can procalcitonin help identify associated bacterial infection in patients with severe influenza pneumonia? A multicentre study

Cuquemelle, E.; Soulis, F.; Villers, D.; Roche-Campo, F.; Ara Somohano, C.; Fartoukh, M.; Kouatchet, A.; Mourvillier, B.; Dellamonica, J.; Picard, W.; Schmidt, M.; Boulain, T.; Brun-Buisson, C.

PURPOSE: To determine whether procalcitonin (PCT) levels could help discriminate isolated viral from mixed (bacterial and viral) pneumonia in patients admitted to the intensive care unit (ICU) during the A/H1N1v2009 influenza pandemic. METHODS: A retrospective observational study was performed in 23 French ICUs during the 2009 H1N1 pandemic. Levels of PCT at admission were compared between patients with confirmed influenzae A pneumonia associated or not associated with a bacterial co-infection. RESULTS: Of 103 patients with confirmed A/H1N1 infection and not having received prior antibiotics, 48 (46.6%; 95% CI 37–56%) had a documented bacterial co-infection, mostly caused by Streptococcus pneumoniae (54%) or Staphylococcus aureus (31%). Fifty-two patients had PCT measured on admission, including 19 (37%) having bacterial co-infection. Median (range 25–75%) values of PCT were significantly higher in patients with bacterial co-infection: 29.5 (3.9–45.3) versus 0.5 (0.12–2) μg/l (P < 0.01). For a cut-off of 0.8 μg/l or more, the sensitivity and specificity of PCT for distinguishing isolated viral from mixed pneumonia were 91 and 68%, respectively. Alveolar condensation combined with a PCT level of 0.8 μg/l or more was strongly associated with bacterial co-infection (OR 12.9, 95% CI 3.2–51.5; P < 0.001). CONCLUSIONS: PCT may help discriminate viral from mixed pneumonia during the influenza season. Levels of PCT less than 0.8 μg/l combined with clinical judgment suggest that bacterial infection is unlikely.