Influenza immunization of the elderly was close to my heart during my training. I often thought our approach was too soft, resulting in a 15% immunization rate at the time. Changing the message to a hard-sell, marketing-driven and (some said) fear-mongering approach improved our center's immunization rate to 50% in one season. However the numbers were small, so I barely achieved statistical significance (after considerable torturing of the numbers, I might add) and the study was presented but not published.
Most of all, the public health folks I worked with were leery of such an aggressive message. I protested that the 30 - 50% reductions in hospitalizations and mortality spoke for themselves. How could anyone exercise a different choice when the rational conclusion was so obvious? I never understood why my approach had been so unpopular.
By the time I moved to the states, I gradually became more aware of a public health approach to health messages that resembled marketing. Most notably, CDC's Julie Gerberding seems to have established an unprecedented emphasis on "communications" within her agency. For a time, I remained convinced of the soundness of this aggressive approach.
But as my career developed I started to perceive the flip-flops in medical recommendations. We thought nopbody should eat eggs due their risk of developing cholesterol problems. Then we thought it was OK to eat eggs and meat, even in a cholesterol-reduced diet. Now we worry only a bit about dietary fat as we start to concentrate much more on carbs. Anyone care to tackle estrogen and the risk of breast of cancer or heart disease?
Of course, this dialectic is part of the scientific process. Only people who know their methods are merely an approximatiopn of the truth are willing to make absolute-sounding statements and remain prepared to revise them on a moment's notice. In fact, scientists are only being absolute about what is known. Based on the best available information, we can say things with certainty that will be revised with new information.
This ain't the Bible folks, and even then... well maybe some other time we can discuss exactly how this library of books was assembled by bishops of the early church.
The new data is that the benefit of influenza vaccine is not as outrageously high as once believed. Perhaps my residency overseers knew something about original estimates. A better method for controlling the severity of illness reduced the magnitude of the apparent effect by simply reducing confounding. We give flu shots to the frailest of the elderly, so we over-estimate the impact on a large population who may not be quite as frail overall. Good job! The vaccine is still protective, and I still think anyone who avoids does not have the right information. But clearly, it's not criminally insane to say "I'd rather not."