Study: Flu shots may not protect elderly from pneumonia

Aug 8, 2008 (CIDRAP News) – Studies designed to tease out the benefits of seasonal influenza vaccines for elderly people have yielded conflicting results over the past few years, and now new findings suggest that the vaccine's ability to reduce the risk of pneumonia may be less than expected.

The researchers, who studied seniors enrolled in a Washington state health maintenance organization (HMO) over three flu seasons, published their findings in the Aug 2 issue of The Lancet. The lead author is Michael L. Jackson, PhD, MPH, a postdoctoral fellow at the Group Health Center for Health Studies in Seattle.

Public health experts strongly recommend flu shots for the elderly, because the disease increases the risk of problems such as cardiac events and pneumonia. However, some recent studies have cast doubt on the mortality benefits of flu shots for the elderly, and some experts have said that confounding factors in past studies led to an overestimation of vaccine benefits for this age-group.

Weeding out biases
The Lancet report describes a population-based case-control study focusing on immunocompetent people aged 65 to 94 years, identified from HMO data from the 2000, 2001, and 2002 flu seasons, when the vaccines closely matched circulating flu strains. The case-group included those who had community-acquired pneumonia between the date each year when the vaccine became available and the end of the flu season. Each case-patient was matched by age and sex with two controls.

The authors describe a variety of steps they took to control for underlying heath differences between cases and controls and other factors that might confound assessment of the vaccine's ability to protect elderly people from pneumonia. For example, the researchers:

  • Excluded people who were living in nursing homes or were immunocompromised
  • Assessed the presence and severity of various health conditions, such as asthma, chronic lung disease, congestive heart failure, dementia, stroke, alcoholism, and diabetes
  • Assessed functional limitations
  • Validated pneumonia diagnoses on the basis of two criteria: the presence of a nonchronic parenchymal infiltrate on chest radiograph or the attending physician's conclusion that pneumonia was the most likely diagnosis
  • Included people who received outpatient treatment for their pneumonia, not just those who were hospitalized

The authors used a two-step process to reduce confounding. First they looked at the relation between flu vaccination and the risk of pneumonia during "preinfluenza" periods—the time between the first availability of flu shots and the start of the flu season. They found that vaccination was linked with an apparent 40% reduction in pneumonia risk during this interval—an effect that was attributed not to vaccination but to better health among the vaccinees, since flu was not yet present in the community.

The investigators then adjusted statistically for the various potential confounding factors (differences in underlying health conditions, medication use, and functional status) until the apparent reduction in pneumonia risk for vaccinated people during the preinfluenza period was eliminated. Next, they adjusted for these confounding factors in an analysis of pneumonia risk during the flu season, when an effect of vaccination was expected.

No significant protection
The authors identified 1,173 patients who had pneumonia and 2,346 controls. Flu vaccination rates were nearly equal in the groups: 77% in the patients and 78% in the controls.

The researchers found that vaccination did not significantly reduce the risk of pneumonia in this population. The odds ratio was 0.92 (95% confidence interval [CI], 0.77 to 1.10; P=0.66), signaling an 8% reduction in risk. For individual seasons, vaccinees were found to have a 6% lower risk of pneumonia in 2000-01, a 2% higher risk in 2001-02, and a 21% lower risk in 2002-03, but none of these differences were significant.

Vaccination was found to be even less protective when the analysis was restricted to the 210 pneumonia cases that occurred during the peak of each flu season (a 5-week period with the peak-illness week in the middle). The odds ratio was 1.04 (95% CI, 0.70 to 1.55; P=0.84), signaling a nonsignificant 4% increase in pneumonia risk in those vaccinated.

One limitation of the study is that some patients could have received flu vaccines outside the managed-care setting, the report says. However, the controls had higher vaccination rates than those for Washington seniors overall, which would reduce the chance of misclassification.

Jackson and his colleagues write that their findings could mean one of two things: that influenza causes only a small proportion of pneumonia in elderly people, or that the vaccine is not very effective in reducing the risk of the influenza infection in the elderly.

The two possibilities have different implications for vaccine development and vaccination recommendations, they add. "Differentiation between them will need studies with laboratory-confirmed endpoints, such as pneumonia or serious respiratory outcomes after a confirmed influenza infection."

Answers remain elusive
It is not yet possible to say which of the two possible implications of the study is more accurate, according to two infectious disease experts, Edward Belongia, MD, and David Shay, MD, MPH, who wrote a commentary accompanying the report. Belongia is a senior epidemiologist at the Marshfield Clinic Research Foundation in Marshfield, Wis., and Shay is a medical officer in the Center for Disease Control and Prevention's influenza division.

The commentators praise the authors' efforts to purge sources of bias that may have affected previous observational studies in which vaccination was found to be protective against pneumonia. Belongia and Shay cite three factors that may explain the different results between Jackson's study and the observational studies: adjustments for confounding by differences in patients' health status, the careful validation of pneumonia diagnoses, and the inclusion of both hospitalized and outpatient subjects.

More studies are needed to explore the causes of pneumonia in seniors, particularly those who have received flu immunizations, Belongia and Shay write. Also, they said standard methods are needed to assess vaccine efficacy across different seasons and in different populations.

Keep on vaccinating
In an Aug 2 press release from the Group Health Center for Health Studies, Jackson said his findings shouldn't dampen efforts to vaccinate older people against the flu. "Despite our findings, and even though immune responses are known to decline with age, I still want my grandmother to keep getting the flu vaccine," he said. "The flu vaccine is safe. So it seems worth getting, even if it might lower the risk of pneumonia and death only slightly."

Some flu experts have called for development of new vaccines that offer more protection for elderly people, and manufacturers are developing adjuvanted vaccines in keeping with that aim, according to previous reports.

Jackson ML, Nelson JC, Weiss NS, et al. Influenza vaccination and risk of community-acquired pneumonia in immunocompetent elderly people: a population-based, nested case-control study. Lancet 2008 Aug 2;372(9636):398-405 [Abstract]

Belongia EA, Shay DK. Influenza vaccine for community-acquired pneumonia. (Commentary) Lancet 2008 Aug 2;372(9636):352-4

See also:

Aug 2 Group Health Center for Health Studies press release

Oct 9, 2007, CIDRAP News story "Fresh doubts, new support for flu shots for seniors"

This week's top reads