Intended for healthcare professionals

Feature Medicine and the Media

Reporting flu vaccine science

BMJ 2018; 360 doi: (Published 05 January 2018) Cite this as: BMJ 2018;360:k15

Official doubletalk hides serious problems with flu shot safety and effectiveness


After weeks of brooding about the Donahue article linking flu shots to miscarriages (Vaccine 2017;35:5314) it was with a sense of relief that I read Rob Wipond’s narrative of media attempts to sweep a serious vaccine safety issue under the rug….He points out the hypocrisy (his words were “double standard”) of authorities who dismissed the Donahue paper because it was an “observational study.” Year after year they have quoted observational studies to announce, “…80% vaccine effectiveness…60% effectiveness…40% effectiveness…” They do not mention that these studies make no effort to look for adverse vaccine effects (e.g. narcolepsy, seizures, high fever, oculorespiratory syndrome). They do not mention “negative vaccine effectiveness”, the increase in risk of illness from influenza and non-influenza viruses associated with (or caused by) the vaccines. (Cowling, Clin Inf Dis 2012;54:1778) They do not mention that a vaccine “effective” in one season may increase influenza risk in a subsequent season. (Read about “antibody-dependent enhancement” to understand one explanatory mechanism). They do not mention that the observational studies they refer to are likely to exaggerate vaccine effectiveness in the first place because of the “healthy user effect” well known to epidemiologists.

Some history: 1960 Nobel Laureate and a primary developer of today’s influenza vaccine, Macfarlane Burnet, didn’t think it was worth much. (Br J Path 1936:17:282. Natural History of Infectious Disease 1972, page212)….In 2000 Kenneth McIntosh warned that we should not routinely give influenza vaccine to healthy children until multicenter randomized trials were done over several seasons to be sure that it was safe and effective. (Editorial, NEJM 2000;342:225) His advice was ignored….In 2004 a “Seven-Step Recipe” for using the media to boost demand for the vaccine was presented to the National Influenza Vaccine Summit, sponsored by the CDC and the AMA. The recipe included, “…statements of alarm by public health authorities…prediction of dire outcomes from influenza…continued reports that influenza is causing severe illness affecting lots of people…repeated urging of influenza vaccination…” (Doshi, BMJ 2005;331:1419) Sound familiar?

Peter Collignon and his colleagues have said this: “We need much larger, independent, and better-reported prospective studies that clearly demonstrate that the benefits of influenza vaccines in children far outweigh harms…If, overall, the increased number of cases of ARI plus vaccine side effects are much larger (in vaccine recipients) than those on placebo, given the low efficacy of the vaccine, then this is a strong argument against current policies advocating routine influenza vaccination of children.” (Collignon, Clin Inf Dis 2015:60:489)

In any discussion of influenza epidemiology we should acknowledge the careful and steady (one could even say fearless) work of Danuta Skowronski and her Canadian public health colleagues. It was they who found that the 2008-9 flu shot doubled the risk of illness from the 2009 H1N1 pandemic flu. Their observations were considered important enough to alter Canadian vaccine recommendations for the 2009-10 season. However, for some reason, they had a good deal of difficulty getting their study finally published. (Skowronski, PLoS Med 2010;7(4):e1000258) This observational study was a revelation, even a shock, to many public health experts. US officials never publicly acknowledged the findings…..This is just one of a number of important papers published over the years by Skowronski and her colleagues, who have a reputation for high scientific and ethical standards.

Wipond does not mention another technique used to dismiss legitimate vaccine safety concerns, having to do with “statistical significance.” Recently, a large cohort study found that flu shots given during the first trimester of pregnancy were associated with a 20% increase in autism spectrum disorder in the offspring. P for the association was 0.01, and the authors acknowledged that, if it was causal, would mean four(4) additional autism cases for every 1000 mothers vaccinated. However, they incorrectly used a statistical manipulation to adjust the finding into “non-significance.” (Hooker, Donzelli, Zerbo. JAMA Pediatr 2017;171:600) One typical media headline about the study was, “Flu vaccine during pregnancy not linked to autism.” (Shute, NPR, 11/28/16)….This kind of thing goes on all the time with news releases for vaccine research.

The flu season in North America officially began 5 weeks ago, and the vaccine publicity juggernaut is still picking up steam. Manufacturers are hoping to sell 166 million doses in the US this season. One business group predicts an $8 billion US influenza vaccine market by 2025. (Coherent Market Insights, 1/5/18) The 2016-17 vaccine increased the risk of H3N2 illness among UK elderly by 68%, and officials are calling for better vaccines. (Osterholm, NY Times, 1/8/18) Meanwhile, in the absence of any evidence that it would help, officials and ordinary citizens in the US and UK wrangle about flu shot mandates for healthcare workers.

Wipond’s article provided new insight for an old baby doctor who, at one time, was a wholehearted vaccine advocate, but lately has been exasperated by bullying and doubletalk from vaccine authorities. Let us hope that his article helps to sustain an honest discussion about the safety and effectiveness of influenza vaccines.

Allan S. Cunningham 1/9/18

Competing interests: No competing interests

09 January 2018
Allan S. Cunningham
Retired pediatrician
Cooperstown NY 13326, USA