Antibiotics in pandemic flu

With the continuing spread of avian H5N1 influenza a possible pandemic of human influenza becomes more likely. If a pandemic started soon no effective vaccine would be available and there would probably be a shortage of antiviral drugs. There is no evidence (yet) of the effectiveness of neuraminidase inhibitors in case of avian and pandemic influenza viruses,¹ and mortality among patients infected with H5N1 bird flu remains high, despite the use of neuraminidase inhibitors.² Resistance to antiviral drugs, which may even develop during treatment, might further limit the efficacy of these drugs.³ Given that secondary bacterial infection is an important and often fatal complication of influenza, antibiotics will also have a critical role in the event of a human pandemic.

In 1918-19, when antibiotics were not available, pandemic flu caused 20 million to 100 million deaths worldwide, with an estimated case fatality rate of between 2% and 4%. There are no data on the numbers of patients who died directly from influenza, or from secondary bacterial infections that might have been prevented with antibiotics.

How should antibiotics be used in the event of a flu pandemic? And how many patients with influenza will develop secondary bacterial pneumonia? In a largely healthy population of adolescents and adults who developed acute influenza (predominantly caused by the virus H3N2) the rate of respiratory events diagnosed by doctors and treated with prescribed antibiotics was around 17%, most commonly acute bronchitis and acute sinusitis. Pneumonia was diagnosed in only 1-2% of patients.⁴,⁵

Can we expect a different bacterial aetiology for cases of pneumonia complicating influenza? Staphylococcus aureus is a common cause of post-influenza pneumonia, characterised by rapid clinical deterioration, septicaemia, and high mortality. In a comparison of pathogens isolated from patients with pneumonia during the Hong Kong influenza pandemic (1968-9) with those isolated from other patients with pneumonia during a one year period outside the epidemic, the proportion of infections caused by Staphylococcus aureus was more than double during the pandemic (26% v 11%), but Streptococcus pneumoniae was even more common (48%).⁶ Carriers of Staphylococcus aureus, therefore, might be particularly at risk of catastrophic complications during a flu pandemic. Whether similar rates of complications and shifts in bacterial aetiology will occur during a H5N1 pandemic is, of course, uncertain.

Should patients with mild respiratory symptoms during a pandemic be treated empirically? General practitioners will be overwhelmed with patients who have mild symptoms of respiratory tract infection but are concerned that they might have influenza. There is no evidence to warrant deviating from current guidelines on managing influenza, in which antibiotic treatment is usually restricted to people with signs and symptoms of pneumonia, especially the very young and very old and those with underlying diseases. Wide-spread prophylactic or pre-emptive use of antibiotics could encourage antibiotic resistance and thereby counterbalance any apparent short term benefits.

Although influenza may be complicated by pneumonia in only a minority of patients, in severe cases it will be difficult to distinguish purely viral pneumonia from bacterial pneumonia.⁷ Therefore, even though most patients with severe flu-like illness will have influenza, such patients must be treated with antibiotics, especially those treated in hospital.

Should current recommendations on empirical antibiotic treatment be adjusted? Patients should have antibiotics which are effective against Staphylococcus aureus and Streptococcus pneumoniae. Although all guidelines for the empirical treatment of community acquired pneumonia cover Streptococcus pneumoniae, Staphylococcus aureus poses more of a challenge. In the United States and Europe infections caused by community associated methicillin-resistant Staphylococcus aureus (CA-MRSA) are emerging.⁸ In some urban centres, as many as half of all Staphylococcus aureus samples recovered from skin and soft tissue infections among outpatients are CA-MRSA.⁹ In such places CA-MRSA should be considered the causative pathogen in episodes of severe community acquired pneumonia that need admission to hospital. Furthermore, in areas with a high prevalence of penicillin resistant Streptococcus pneumoniae doctors should ensure that they give β-lactam antibiotics in adequate doses.

Finally, doctors might also need to consider other measures. Pneumococcal vaccination might offer some protection against secondary bacterial infections, although randomised trials do not indicate that polysaccharide pneumococcal vaccines would be protective in preventing pneumonia and death.¹⁰ Recently introduced technology now allows rapid detection of Staphylococcus aureus carriage, which could be used to identify patients at increased risk for secondary pneumonia.¹¹ Both measures would need substantial financial investments in the absence of evidence of efficacy.

Modern communication technology, rapid diagnostic testing, and better preparedness should yield real understanding of these questions in the first weeks and months of a pandemic. In the meantime we will have to rely on conventional wisdom.