Flu jab and pneumonia

Monday August 4 2008

“One of the main benefits of flu jabs for the elderly - protection against pneumonia - may not exist”, BBC News reported. It said that a study in thousands of vaccinated and unvaccinated people found that healthy, older people who had the jab had the same risk of pneumonia as those who had not.

This study found that flu vaccination appears to offer little protection from community acquired pneumonia (CAP), a common complication of flu, for relatively healthy, elderly people who do not live in institutions. Other studies have provided evidence to suggest that it is beneficial for vulnerable groups, such as those who are immunocompromised (from current treatments or chronic illness) and those in nursing homes.

Although the study was well-conducted, and the researchers used robust analyses to adjust for other factors that could explain the results, they admit that they might not have measured all of these. Also, it is unclear what type of CAP the participants had - viral or bacterial. Bacterial CAP is more common both as a complication of flu and as an isolated infection. Their episodes of CAP may be unrelated to flu episodes. Further research is needed; the researchers themselves call for this.

Where did the story come from?

Dr Michael L. Jackson and colleagues from the Group Health Center for Health Studies, the Fred Hutchinson Cancer Research Centre, PATH and the University of Washington carried out this study. The research was funded by the Group Health Center for Health Studies and by a fellowship grant from the Group Health Community Foundation. The study was published in the peer-reviewed medical journal: The Lancet.

What kind of scientific study was this?

The study was a nested case-control study in healthy (immunocompetent) elderly males and females between 65 and 94 years of age. The participants were enrolled with a health management organisation called Group Health (an organisation that provides medical coverage and healthcare services) in the western Washington state. They were healthy adults with no history of serious cancer, chronic renal failure, prescriptions for immunosuppressive medications during the previous two years, and no history of treatment for cancer in the previous three months. Each had also made more than two visits to Group Health in the previous two years.

For each year of the study (2000, 2001 and 2002), the researchers were interested in whether or not the participants who contracted pneumonia had been given that year's influenza vaccine. The cases in the study were those people who had an episode of CAP (either as outpatients or inpatients) during that year. Their illness was validated through assessment of their chest radiographs and medical records; the researchers who performed this task were unaware which people had been vaccinated. Each case subject was randomly matched to two control subjects from the source population for age (within one year of the case’s birth date) and gender. The controls had not had an episode of CAP before their matched case fell ill.

Researchers reviewed the participant’s medical records for the two years preceding the study start date, with particular attention to the following details: asthma, chronic obstructive pulmonary disease, congestive heart failure, alcoholism, diabetes, dementia, and stroke. The researchers also collected information on the functional status of participants; whether they needed assistance bathing, walking or eating, and whether they smoked. Other prescription data from the General Health records were also collected as indicators of other diseases. People who were living in a nursing home or another institution or who were immunocompromised (according to their records) were excluded from the study.

For each year of the study, the researchers compared the likelihood of an episode of CAP in relation to vaccination while taking into account other health factors. Their overall sample included 1,173 validated cases of CAP and 2,346 matched controls.

What were the results of the study?

Overall, people who had CAP were more likely to have chronic diseases, functional impairments, and prescriptions for lung medications and antipsychotics. The cases and controls were equally likely to have received the flu vaccination.

When the researchers took into account all the factors they believed may have influenced the outcome including age, sex, asthma, smoking, antibiotics for lung conditions and previous pneumonia, they found that there was no significant effect of vaccination on the risk of CAP in the participants. This means that people who got pneumonia were as likely to have been vaccinated as those who did not get pneumonia. Further analysis of subgroups found no effect of vaccination on infection during the peak season, on risk of hospitalisation, or on risk of infection in one of the flu seasons included in the study (2000, 2001 or 2002).

What interpretations did the researchers draw from these results?

The researchers concluded that their ‘large, population-based, nested case-control study’ found no effect of flu vaccination on the risk of community-acquired pneumonia in the elderly.

What does the NHS Knowledge Service make of this study?

In this well-conducted study, the researchers took into account as many additional factors as possible. They repeated their analyses using different statistical techniques and found that their overall conclusions did not change. In addition, they chose to analyse data from three flu seasons in which the flu vaccination was shown retrospectively to be well matched to the flu virus that ended up circulating in the population. The researchers also did what they could to validate the diagnoses of pneumonia by independently reviewing records or chest X-rays.

The main problem with studies of this design is the adjustment for, or taking into account of, other factors that might be responsible for the association between the exposure (vaccination) and the outcome (pneumonia). The efforts the researchers made in order to account for these factors increases confidence in the results.

In this study, the researchers were interested in the effect that the flu vaccination had on pneumonia in the elderly. Pneumonia is a common and serious complication of the flu virus and arises either from the influenza virus directly infecting the lungs or from a secondary bacterial infection. The researchers say that their results could mean one of two things:

  • First, that pneumonia is not an important complication of influenza in this age group, i.e. the flu vaccine may be highly effective at reducing flu in the participants, but if pneumonia is not usually associated with flu then there wouldn’t be a parallel reduction in pneumonia cases.
  • Second, that the flu vaccination is ineffective at reducing influenza in people at risk of pneumonia.

They say that these two possibilities have quite different implications for vaccine policy and development and that further research is needed to clarify the results.

Flu vaccinations are prepared before the flu season starts and are designed to protect against the strains that are expected to be predominant. They do not protect against bacterial infection. Bacterial pneumonia is common in the elderly either as a complication of flu or as an isolated infection. The results would have been clearer if the infecting cause of the pneumonia was known, e.g. from microbial reports in the laboratory, and also if it were known whether there had been an epidemic of a particular bacterial pneumonia in those seasons. Although the majority of the participants appear to have had the pneumococcal vaccine, these factors could explain in part the differences seen here.

Importantly, this study only included relatively healthy individuals (who were immunocompetent and not living in nursing homes or other institutions). The flu vaccination is recommended for a number of different groups who are considered vulnerable, such as people living in institutions, those with a chronic illness, and those who are immunocompromised for some reason (such as receiving steroid treatment or treatment for cancer). These people should continue to receive vaccinations as recommended, as there is good evidence that pneumonia, hospitalisations and possibly even death may be reduced.

Sir Muir Gray adds...

One swallow does not make a summer and we need to see these results combined with the results of other studies before older people could be advised not to bother with flu vaccination.

Analysis by Bazian
Edited by NHS Choices