This site is intended for health professionals only

At the heart of general practice since 1960

Read the latest issue online

CAMHS won't see you now

Is flu vaccination worth the effort?

Dr Tom Jefferson and Dr Doug Fleming take opposing sides in the flu vaccination debate

Dr Tom Jefferson and Dr Doug Fleming take opposing sides in the flu vaccination debate


There is not enough evidence to support the money and time poured into the annual effort to immunise all over-65s against flu – and a randomised controlled trial is badly needed, argues Dr Tom Jefferson

Acute respiratory infections are a GP's common lot throughout the year. They take up everyone's time and inflict short bursts of suffering on the afflicted.

We seem not to be able to do much about them. One of the problems is that ‘the flu' is not a distinct disease caused by one or two agents but a spectrum that goes from a very mild cold to a severe incapacitating and (rarely) deadly illness caused by scores of different micro-organisms.

In reality, all we can hope to achieve with vaccines is to prevent that variable amount of flu caused by the two influenza viruses, A and B. Every year a huge campaign gets under way taking more and more resources as the indications for influenza vaccine use expand.

Some years ago my Cochrane vaccines group decided to carry out systematic reviews of the evidence of the effects of influenza vaccines. Our main reasons were that we noticed considerable variability in the effectiveness estimates quoted in the literature.

Cochrane reviews entail a mind-bogglingly detailed search for, and evaluation of, every piece of available evidence – and 10 years and several reviews on, we are still at it.

We looked at the whole population of studies that compared the effects of influenza vaccines in children, adults, healthcare workers and the elderly with either placebo or no vaccination. We got a few surprises.

Sticking to the elderly, these surprises were, in order of importance:

• the paucity and relatively small size of randomised trials (5,000 participants in all)

• the large number and generally abysmal quality of studies in which participants had not been randomised

• happy-go-lucky quality reporting of methods and results.

Take the randomised trials, for example: only two of the five ever done on ‘population outcomes' (defined as cases or complications and not antibody responses) tested modern vaccines.

One was carried out over 20 years ago in Holland on a small population, and the other was done in Liverpool at the turn of the millennium.

The UK trial was made difficult, and the results impossible to interpret, by policy changes outside the researcher's control. The change in national policy from the over-75s to over-65s made recruitment into the placebo arm impossible.

Non-randomised studies were plentiful. We conducted a massive meta-analysis on hundreds of thousands of observations and outcomes.

Unfortunately, what we found did not make sense: according to our results ‘sequence', influenza vaccines prevented the omnipresent ‘flu' but not influenza proper or its complications.

They did, however, prevent deaths for all causes (cardiovascular, infectious, accidental, violent and so on – even, as one study claimed, up to 90% of them).

We thought all this did not make sense and we attributed the effects to poor study quality

Uncertainty reigns

We concluded that uncertainty on the effects of the vaccines reigned in this age group and the logical thing to do would be to conduct a large robust placebo-controlled trial.

What appeared a logical conclusion to us did not sit well with some.

We were accused, among other things, of being an out-of-control section of the Cochrane Collaboration. Yet we had assembled the evidence in a painstaking and auditable manner and published it in highly read journals as well as the Cochrane Library.

You cannot credibly refute the results of a review, and the reviewers' assessment of methodological quality of studies included in the review, without reading them. And that means all of them, not just one or two.

It seemed to us that the world was upside down. Those who reviewed scientific evidence following Cochrane rules were ‘out of control' but those who based their observations on their prejudices and personal views were reasonable people.

Thankfully, a group of American researchers has recently reached the same conclusions as we did, using completely different methods from ours. They published their findings in The Lancet Infectious Diseases last month.

So if we are uncertain if all the effort to immunise over-65s is worth it, what is to stop us from finding an answer?

I do not think there is any reason not to test the effects of current vaccines – with a large placebo-controlled trial, carried out over several seasons following correct methods, by reputable and independent researchers.

Ethics are often invoked as a show-stopper to such a trial, but for me it is unethical to continue a policy based on uncertainty.

And, no I am not volunteering to run the trial – just in case someone thinks I may be feathering my nest.

Dr Tom Jefferson is a former GP and co-ordinator of the Cochrane Vaccines Field


Rather than undermining the flu vaccination programme, it's time to focus efforts on finding an answer to RSV, argues Dr Doug Fleming

The clinical effectiveness of influenza vaccination of the healthy elderly and other risk groups has been established from randomised controlled trials.

That effectiveness is mostly seen in the impact of influenza on deaths in hospital admissions for pneumonia. The benefits are best when the vaccine is well matched to circulating strains and in those winters when influenza has a particularly severe impact on the elderly.

Judgments about the cost-effectiveness of a policy of routine annual vaccination have to be interpreted in relation to the circumstances in each year, according to the severity of the illness and extent of its impact, both of which vary.

In recent years there have been decreasing numbers of people presenting to doctors in England and Wales with flu-like illness.

This general decrease has been observed in most age groups, including those in which very few people are vaccinated. Most studies of vaccine efficacy were undertaken in the 1990s when influenza was a more serious illness (as judged by the number of people presenting with flu-like illness).

However, the clinical syndrome of flu-like illness is not specific to influenza virus infection and though incidence is well matched by virus isolate data, it may also be caused by other respiratory viruses, notably respiratory syncytial virus (RSV).

The immune response to natural influenza and to influenza vaccine deteriorates with advancing age – immune senescence. The effectiveness of influenza vaccination in the frail elderly is reduced compared with the healthy elderly.

Although vaccination may not protect all elderly people from flu virus infection, the majority of them benefit by at least having a less severe illness.

The symptoms of influenza are closely associated with the amount of virus shedding in the nose and upper airways. It is a reasonable presumption that if fewer people have flu, especially severe flu, there will be less likelihood of spread – particularly crucial in institutions and nursing homes.

The declining severity of flu in recent years has not been associated with a matching reduction in excess winter mortality, which has commonly been associated with influenza virus infection.

Other viruses

Many studies of the effect of flu vaccination have not given enough attention to the confounding effect of other viruses.

Respiratory infection in the elderly inevitably presents risks for serious outcomes.

We are beginning to appreciate more the significance of some infections, in particular RSV. Research in various parts of the world, including at the Birmingham research unit of the RCGP, increasingly points towards RSV as another major cause of winter excess mortality.

This has become increasingly evident because influenza incidence has declined, partly because of natural variation, partly because of vaccination. We also need an RSV vaccine with a comparable effect.

It may be that the benefits of flu vaccination as evident 10 to 15 years ago overstate the current situation, but not the situation that is likely to follow the spread of a novel pandemic virus. This is not the time to reduce our efforts to control flu by vaccination. It may nevertheless be appropriate to revisit the determination of policy based solely on cost-effectiveness.

I regret that greater attention has not been given to the impact of respiratory virus infections other then flu.

Most are clinically mild and self-limiting: however even though they are relatively infrequent, the serious outcomes from these apparently minor illnesses add up to a major public health problem. Routine annual influenza vaccination of the elderly and people with co-morbidities remains vital in addressing this problem.

Dr Doug Fleming is director of the RCGP Birmingham research unit and a member of the JCVI influenza sub-group

Flu vaccine

Rate this article 

Click to rate

  • 1 star out of 5
  • 2 stars out of 5
  • 3 stars out of 5
  • 4 stars out of 5
  • 5 stars out of 5

0 out of 5 stars

Have your say