Our guest contributor is Dr. Joshua Goldenberg, a naturopathic physician, professor, researcher, and creator of Dr. Journal Club. He has presented nationally and internationally on evidence-based medicine as well as probiotics and research methodology. He created www.DrJournalClub.com which is an online repository of educational videos serving to keep medical professionals up-to-date on current and relevant research as well as hone their evidence-based medicine skills. In his continuation of the Flu Shot Series today he discusses a Cochrane Review on using the influenza vaccines to prevent influenza and it’s complications in the healthy elderly population.
Today we take a look at a systematic review and meta-analysis on vaccinating the elderly population. The authors found 75 studies on using vaccines to prevent influenza in the elderly. However, only 5 of those were RCTs and of those really only one discussed a type of vaccine still used today and was of decent study quality.
The results of the review are as follows: It appears that flu shots are safe for the elderly. They found no evidence of severe or serious side effects from the flu shot. The RCT evidence shows that the flu shot cuts the risk of getting flu-like illness down by 41%. Non-RCT evidence suggests that it is a relative risk reduction of 24% for those in nursing homes and not statistically significant in community dwelling elderly. Flu shots decreased the risk of getting true flu by 58% according to the RCT data. This is about what we see in healthy adults as well. The non-RCT data suggests the reduction in relative risk is 35% in nursing homes and not statistically significant in community dwelling elderly.
Regarding complications for flu, there is no RCT data for pneumonia prevention though non-RCT data suggest a 47% relative risk reduction for the elderly in nursing homes and a non-statistically significant decrease in the community dwelling elderly. There is no RCT data on hospitalization but the non-RCT data show a 49% relative risk reduction in those living in nursing homes and a 27% reduction in community dwelling elderly. This is very impressive! Even more impressive seem to be the results for mortality. The RCT results show no statistically significant difference in death rate with those vaccinated compared to not vaccinated but the non-RCT data show a 54% mortality reduction for those living in nursing homes and a 29% relative risk reduction for those living in the community. This seems very impressive as well!
However, the issue of course is the difference between randomized controlled trial data and non-randomized controlled trial data. While we have RCT data on preventing flu and flu-like illness what we really care about is preventing complications and we simply don’t have any RCT evidence on this. What we do have is a lot of non-RCT evidence which suggests very impressive effects on complications but these studies are prone to have what is called selection bias or in this case, healthier vaccinator bias. Basically the concern is that the elderly who take care of their health like exercising and eating right are the ones who go out and get vaccinated because they think this is a way to stay healthy as well. So the question is, is their improved mortality because of the flu shot or because of everything else they are doing to stay healthy? There is some suggestive evidence that this is happening. For example, we see a 44% reduction in mortality during the flu season for those getting vaccinated, but if you look at mortality before the flu season you see a 61% reduction. If it is really the vaccine that is preventing death by preventing people from getting the flu how is it preventing death before the flu season even begins! In other words we should be highly suspicious that there is confounding here and we are seeing that the group of people who get vaccinated are different and healthier than the non-vaccinated groups. Additionally, the calculated mortality improvement based on the non-RCT data suggests a mortality rate decrease that far exceeds estimates of flu-caused mortality each season. So again the non-RCT evidence may be grossly overestimating the mortality effect.
In sum, a choice quote from the authors: “The available evidence is of poor quality and provides no guidance regarding the safety, efficacy or effectiveness of influenza vaccines for people aged 65 years or older. To resolve the uncertainty, an adequately powered publicly-funded randomized, placebo-controlled trial run over several seasons should be undertaken.”
I agree that more RCT evidence is needed and that the non-RCT data is probably grossly overestimating the effect of vaccination. However, this is the evidence we have and we know that flu shots are safe. Therefore, I use the precautionary principle. We know this is a safe intervention and we have evidence (albeit weak) that it has significant effects on complications and death. So until better evidence comes to light it seems reasonable to vaccinate this higher risk group and especially those community dwelling elderly that are at increased risk and any of those that live in nursing homes.