Issue 3 / 1 February 2016

INFLUENZA vaccine may not be as effective as first thought, but remains the best preventive measure available and plays a crucial role in protecting at-risk groups from severe infection, say experts.

Writing in the Medical Journal of Australia this week, researchers estimated influenza vaccine effectiveness against laboratory-confirmed infection was 50% (95% CI, 26–66%) for general practice patients, and 39% (95% CI, 28–47%) for hospitalised patients.

The researchers reported annual case test-negative results in general practice and hospital settings in the three years to 2013.

Lead author, Professor Heath Kelly, adjunct professor of the National Centre for Epidemiology and Population Health at the Australian National University, told MJA InSight it was interesting to find that influenza vaccine was no more effective in hospitalised patients than in community patients.

“When we first started doing this work, [we thought that] although the vaccine might be only modestly effective in the community, it might be more effective for sicker people, but that seems not to be the case,” Professor Kelly said. He noted that this was consistent with findings in New Zealand, the US and Canada.

Professor Kelly emphasised, however, these findings should not have any impact on the public health messages about the importance of influenza vaccine.

“Influenza vaccine is the only effective preventive measure we have. The message is the same: vaccination is recommended and indicated for those at increased risk of severe infection – children, and those with comorbidities, such as underlying cardiorespiratory disease,” he said. “It’s just that we have to be careful with our messages that we are not exaggerating risk or benefit.” 

Related: MJA — The safety of seasonal influenza vaccines in Australian children in 2013
Related: MJA — Challenges in regulating influenza vaccines for children

Dr Alan Hampson, chairman of the Influenza Specialist Group (ISG), said there was general agreement that earlier evaluations of vaccine effectiveness had been overestimates.

“Certainly it’s an indication that we don’t have perfect vaccines, and we are working to get better vaccines,” he said, adding that efforts were underway to speed up the process of identifying and adapting viruses for vaccines, and producing vaccines that provided broader immunity.

Dr Hampson commented that the authors were “pushing the comparison a little far” by comparing an older hospital population with the younger general practice cohort.

He said the vaccine effectiveness studies also failed to take into account the protective impact of the influenza vaccine against comorbidities such as myocardial infarction.

Professor Kelly acknowledged that the two datasets were not exactly comparable. He noted, however, that the research team had tried to reduce this problem by looking only at the age groups common to both sets, and they still found no difference in vaccine effectiveness between community and hospitalised patients.

Professor Kelly said the Victorian findings also reflected those of the larger New Zealand Southern Hemisphere Influenza Vaccine Effectiveness Research and Surveillance (SHIVER) project, of which he is a member. The SHIVER project draws its community and hospital patient groups from the same areas of Auckland.

In a second paper published in the MJA, researchers countered claims in the mainstream media that 2015 had been a severe influenza season.

After analysing data from the Australian Sentinel Practices Research Network (ASPREN), the Victorian Sentinel Practice Influenza Network (VicSPN), the Sentinel Practitioners Network of Western Australia (SPNWA), and Flutracking (a national, weekly online survey of volunteers that can estimate the proportion of participants with influenza-like illness), the authors found the peak rates of influenza-like illness were not notably higher in 2015 than in the previous five years. They noted that the 2015 illness rate was similar to the 2012 rate, but the percentage of positive polymerase chain reaction test results was lower in 2015 than in 2012. This was in contrast to large increases in laboratory-confirmed influenza notified to the National Notifiable Disease Surveillance System (NNDSS), with an average annual increase of 67% between 2010 and 2014.

“This discrepancy suggests that the increase in notified cases of influenza may be attributable to increasing testing, rather than increasing morbidity,” they wrote.

Podcast: Dr James Fielding on the severity of the 2015 flu season

Dr Hampson had several concerns about these findings.

“I agree that the number of [PCR] tests is increasing, but we don’t know to what extent that that is contributing to the final numbers,” he told MJA InSight.

He noted that the NNDSS figures showed that South Australia and Queensland recorded the highest number of cases, but state surveillance data from these states was not included in the analysis. On the other hand, state data was included from WA and Victoria, which had recorded fewer cases.

Dr Hampson added that the 2015 influenza peak of positive viral samples (about 50% influenza-positive) extended over six weeks, compared with a one-week peak in 2014.

“We do need to do better in terms of assessing the impact of influenza, but I don’t think the way it’s presented in this paper is really a strong argument that 2015 was not a very significant season,” Dr Hampson said.

He said it was particularly important to keep tabs on the prevalence of influenza in young children, as, based on limited Australian data, the ISG estimates that 10 to 12 children die each year from influenza.

Asked how the 2016 influenza season was likely to unfold, Dr Hampson said it was still “too early to tell”, but there had been outbreaks of H1N1 on the Indian subcontinent and in the Middle East, and so far it has been the predominant strain in the US and Europe.

Professor Kelly commented that predicting the likely impact of influenza was a “recipe for failure”.

He said the impact was likely to be dependent on the predominant circulating strain.

“We still have problems with the H3N2 component of the vaccine, so if you get a H3N2-dominant year, that could be a problem.”

One thought on “Flu vax not perfect, but it’s the best we have

  1. Ulf Steinvorth says:

    ‘No reviews of efficacy in preventing hospital admissions have been published because there have been no trials reporting this outcome’. Enough said. (Quote from orignial study). A non-randomized, non-prospective study like this crossing the 0 in efficacy on two out of three years and failing to show improved survival or reduced severity of illness is hardly the stuff to convince sceptics or scientists that this vaccination is working.

Leave a Reply

Your email address will not be published.