Bacterial Pneumonia and Influenza Pandemics: What do the Data Tell Us?

Following the H1N1 pandemic of 2009-2010, questions were raised regarding the role of associated bacterial infections. Specifically, did influenza-associated bacterial pneumonia significantly contribute to morbidity and mortality during the last pandemic? And, if so, how did its contribution compare to previous pandemics?

"Influenza-associated bacterial pneumonia clearly contributed to morbidity and mortality during the recent pandemic," said Mathew Moore, MD, MPH, Captain, USPHS, during the 48th annual meeting of the ISDA in Vancouver. "The overall contribution of influenza-associated bacterial pneumonia was modest in comparison to the number of cases of pH1N1 infection, though limited bacterial diagnostics preclude precise estimates," he said. However, influenza-associated bacterial pneumonia had similar clinical and pathological findings among severe cases compared with previous pandemics.

Moore clarified what is meant by bacterial-associated influenza, stating that there is mixed viral bacterial pneumonia and secondary bacterial pneumonia. "On the one hand, there are the combined bacterial and viral ‘co-infections’ that have the opportunity to interact with the host as well as each other. By contrast, infection with influenza pre-disposes the host to further or secondary infection with bacteria, which can occur several days following the initial infection."

Several studies have been done to try and determine what proportion of hospitalized pH1N1 cases had evidence of any pneumonia. Jain et al (2009) showed 272 patients hospitalized between April and June 2009 in 24 states. Forty-five percent were age 18 years or less. Of the 272 hospitalizations, 249 received chest x-rays, and 40% of those had evidence of pneumonia. A study by Louie et al (2009) reported on 1,088 patients (median age 27 years) hospitalized in California, between April 23 and August 11, 2009. Chest x-rays were done in 833 patients, and 547 of those (66%) had evidence of pneumonia. "Roughly one-third to two-thirds of hospitalized pH1N1 cases had radiographic evidence of pneumonia," Moore said.

Further investigation in both studies revealed that the percentage of patients with bacterial pneumonia was extremely low. In Jain's study, only 3 of 182 had positive blood cultures, yet 206 of 260 (79%) received antibiotics.

Reviewing data sets of patients admitted to intensive care units in Mexico, Canada, Australia, New Zealand, and Argentina, Moore pointed out that the detection of bacterial pneumonia increased. "We're starting to get the sense that in patients who were really sick we were able to find evidence of bacterial infections," he said.

In the pediatric population specifically, data from the CDC show the presence of bacterial co-infection among pediatric influenza deaths between 2007 and 2009, with 32 of 84 children having a co-infection with seasonal influenza, and 46 of 157 with concomitant infection during the 2009 H1N1 pandemic. "However, the question remains, did bacterial infections cause more severe outcomes, or are we better at finding bacteria among patients with severe outcomes?" Moore said. "I think it's probably a bit of both."

The role of vaccines must also be considered; specifically, did the introduction of the 7-valent pneumococcal conjugate vaccine lessen the morbidity associated with pH1N1? Data from the South African clinical trial of the 9-valent pneumococcal conjugate vaccine show that it was efficacious in preventing 45% of influenza A-associated pneumonia. "A bacterial vaccine cannot prevent influenza infection. However, if influenza A infections are complicated by pneumococcal infections, and if pneumococcal infections can be prevented by a vaccine, then perhaps those influenza infections do not come to light in the first place," Moore said. "It may be that the implementation of the conjugate vaccine helped us out in this last pandemic. The role of antivirals, access to healthcare, and antibiotics must also be considered," he concluded.