Articles

Yes, But. The Annotated Atlantic.

Reposted on 11/8 with multiple typo corrections.

The Atlantic recently published an article called “Does the Vaccine Matter?.” The quick answer is “yes”. If you want to know more, keep reading. They concluded, based on a narrow interpretation of a small subset of the data, that vaccines probably do not matter. The tone suggests that the vaccine is a vast boondoggle perpetuated on the American people by frightened doctors and greedy pharmaceutical companies. At least that is my take on the article, your mileage may vary. Lets look at that article, and its review of the influenza vaccine, and see whatthe authors  say, how they say it, and, perhaps more importantly, what they don’t say.

Unfortunately, I do not have a good story to tell with protagonists and antagonists and lone voices protesting the evil medical industrial complex. I don’t have a morality tale to tell, with good guys and bad guys. I have the medical literature, with its numbers and uncertainties and nuance. I also have patients I have to treat and have to apply the medical literature to as best I can.
This entry may be a bit of a repetition for those who read my previous entry on vaccine efficacy, but my entry hit the blogosphere a few days before the Atlantic article, so I did not get a chance to incorporate it into my entry.

Let us go through the article paragraph at a time, with commentary as needed. The bold is from the Atlantic. There is a logical fallacy called Argue By Demanding Impossible Perfection. I wonder why I mention that…

Drive too fast along Red Lion Road, beside Philadelphia’s Northeast Airport, and you will miss the low-rise cement building where the biotech company MedImmune has been quietly pumping out swine flu vaccine at about a million doses a week. Through the summer and fall, workers wearing protective gear that covered them from head to toe brewed up batches of live, genetically modified flu virus. Robots then injected tiny doses of virus-laden fluid into glass vials, which were mounted into nasal spritzers, labeled, and readied for shipment at the direction of the Centers for Disease Control and Prevention, in Atlanta, which is helping to coordinate the nation’s pandemic-preparedness plan. In the most ambitious vaccination program the nation has mounted since the anti-polio campaign in the 1950s, the federal government has commissioned MedImmune and four other companies to produce enough vaccine to cover the entire U.S. population.

Vaccination is central to the government’s plan for preventing deaths from swine flu. The CDC has recommended
that some 159 million adults and children receive either a swine flu shot or a dose of MedImmune’s nasal vaccine this year. Shots are offered in doctors’ offices, hospitals, airports, pharmacies, schools, polling places, shopping malls, and big-box stores like Wal-Mart. In August, New York state required all health-care workers to get both seasonal and swine flu shots. To further protect the populace, the federal government has spent upwards of $3 billion stockpiling millions of doses of antiviral drugs like Tamiflu–which are being used both to prevent swine flu and to treat those who fall ill.

Paragraphs one and two set the scene. Vaccines are important. Already I am uncertain as to the content. Yes, the US wants to dispense 159 million
H1N1 vaccines to decrease the morbidity and mortality from swine flu. But swine flu shots are not yet available at the time this was published, just seasonal flu. Watch as you read the article as the authors bounce from swine flu to seasonal flu to treatment to death to prevention and often fail to be precise as to  what they are referring to. And, as I tell the residents, precision of thought manifests as precision of speech and writing, he says setting himself up for a huge fall in the comments. I now wait for that shoe to drop. At least I am not going to make a grammar complaint. That would lead to certain sarcasm in the comments.

But what if everything we think we know about fighting influenza is wrong? What if flu vaccines do not protect people from dying–particularly the elderly, who account for 90 percent of deaths from seasonal flu?

What if the 14,400 plus influenza vaccine articles on Pubmed are wrong, all the biology and virology and pharmacology and clinical trials about influenza are wrong?  What if every brick in the wall was an illusion and the edifice of flu treatment is wrong. What if a few brave souls can see the real truth. I look forward to a review of the 50 years of influenza research in all its complexity. It’s a huge literature, with multiple lines of evidence all converging on the conclusion that vaccines and antivirals are effective against influenza. Because I would hate to make decisions based on the opinions of a few people reading a narrow sampling of the literature.

And which is it? Is EVERYTHING wrong?  Everything? Or the narrow issue of mortality prevention. The articles starts with everything, but focuses on vaccine preventable mortality and, to a lessor extent, the efficacy of Tamiflu. Does the vaccine prevent death in the elderly? That is not as simple a question to answer as you would think. Here is where the nuance comes in.

Here are possible outcomes of receiving the flu vaccine:

Patient A doesn’t get flu. Life continues.

Patient A gets the flu, but has a milder case and doesn’t die directly of flu.

Patient A gets the flu, but has a milder case and doesn’t go on to die of complications of flu, like a secondary infection or worsening
heart failure.

But perhaps of equal importance, under all three possibilities, patient A is less likely to pass on the disease to the next person who may not have immunity and then could die of flu or its secondary complications.

Key point: with highly contagious diseases, you try and prevent or ameliorate disease in individuals, but one of the effects is the prevention or amelioration of disease in populations. You try for herd immunity, which, because you need a vaccination every year, we are not even close to achieving. So, for example, you have a large population that does not get a reliable response to the vaccine. Like the elderly. And the elderly are constantly exposed to a population that has the flu because of lack of vaccination: the community,  for them (the elderly)  the effect of flu vaccination will be less than one might like. And it is an important issue in evaluating flu vaccine efficacy in the elderly and makes the epidemiologic data always less robust than we would like.

“And what if the expensive antiviral drugs that the government has stockpiled over the past few years also have little, if any, power to reduce the number of people who die or are hospitalized.”

What if? Well,

“Two data sets were presented at the second European Influenza Conference in Malta (www.eswi.org) in September. A large
retrospective cohort study of patients with influenza-like illness (n = 176001) taken from a US health database showed that oseltamivir (75 mg twice daily, n = 39 202) significantly reduced the risks of pneumonia by 32% (P <0.001) and of death by 91% (P < 0.05) (Nordstrom et al).

In Canadian patients with laboratory-confirmed influenza (A or B) requiring hospital admission (< 15- > 64 years; n = 359), oseltamivir reduced the risk of death by 68% (McGeer et al). Treatment with oseltamivir therefore statistically and meaningfully reduces the risk of death in patients of all ages and from all walks of life, infected with influenza A or B. (16293852).”

Whew. At least we do not have to worry about oseltamivir efficacy. It is important to remember that antivirals are nowhere near as good as antibacterials and efficacy is always less than 100%. But decreasing risk of death by 68%? When I have a patient being admitted with severe influenza, 68% is a lot better than nothing. Of course, one of the thrusts of the article, as we will see, is that the only trial upon which we can make decisions is the randomized, placebo controlled trial. It is the best data, but not the only data.

The U.S. government–with the support of leaders in the public-health and medical communities–has put its faith in the power of vaccines and antiviral drugs to limit the spread and lethality of swine flu.

It is not faith, it is a reliance on the data, with all its problems and uncertainties. But as we go through the paper, remember in complex diseases and their treatment, it is the preponderance of data that guides what we do. Medical knowledge is cumulative and changing and rarely has a simple binary, yes/no, black/white, answer. Almost always the answer starts with a “it depends on.”

Other plans to contain the pandemic seem anemic by comparison.

Plans like let people get the flu and suffer and die? Shut down society for three months? For pandemic H1N1, short of telling the population to stay home and don’t go out for the next three months, or not inhaling for the next three months, there are not a lot of other good options. One of the issues we worry about in pandemic flu is how the hospital is going to function if there is no one to make and deliver all the supplies it takes to keep our doors open. Your household probably has the same issues. When a third of everyone stays home, just who is going to get the food to the store, much less to your larder? What happens to essential services? Police, fire, ambulance drivers, bartenders and brewers? We need these people.

“Yet some top flu researchers are deeply skeptical of both flu vaccines and antivirals. Like the engineers who warned for years about the levees of New Orleans, these experts caution that our defenses may be flawed, and quite possibly useless against a truly lethal flu.”

I am already confused. Is this about seasonal flu? H1N1 pandemic flu? Or a repeat of the highly virulent H1N1 flu of 1918 that killed maybe 5% of the world, which we appear to have avoided this time around? Remember the “it depends on” mentioned a few paragraphs ago? Already the “it depends on” is being obscured. The response needed for seasonal flu is not the same as needed for a highly lethal pandemic, and, as H1N1 progresses, we are learning the differences between it and seasonal flu. There is a difference in the response to a flu that kills 36,000 elderly each year, a flu that kills 30,000 (maybe) children and pregnant women, and a flu that kills 5% of everyone.

And that unless we are willing to ask fundamental questions about the science behind flu vaccines and antiviral drugs, we could find ourselves, in a bad epidemic, as helpless as the citizens of New Orleans during Hurricane Katrina.

I am getting increasingly pumped. A good review of the entire flu literature appears to be in the offing. Because if you cherry pick and focus on one or two papers, no matter how well done, you can make any argument you choose. And that would be a disservice to everyone.

Then as now, flu seemed to appear out of nowhere each winter, debilitating or killing large numbers of people, only to vanish in the spring. Today, seasonal flu is estimated to kill about 36,000 people in the United States each year, and half a million worldwide.

Yet the flu, in many important respects, remains mysterious. Determining how many deaths it really causes, or even who has it, is no simple matter.

No simple matter. Watch for the later, “Demonstrating the efficacy (or lack thereof) of vaccine and antivirals during flu season would not be hard to do, given the proper resources. Take a group of people who are at risk of getting the flu, and randomly assign half to get vaccine and the other half a dummy shot. Then count the people in each group who come down with flu, suffer serious illness, or die.

Rather than not being ‘simple’, such a determination ‘would not be hard to do’, which is, I guess, different than simple. It is not simple, I agree, which is why I discuss it on this site.

We think we have the flu anytime we fall ill with an ailment that brings on headache, malaise, fever, coughing, sneezing, and that achy feeling as if we’ve been sleeping on a bed of rocks,

I do like the phrase sleeping on a bed of rocks and I fully intend to steal it for talks. I usually say you feel like you have been pummeled all over by a small hammer, but I prefer their description.

but researchers have found that at most half, and perhaps as few as 7 or 8 percent, of such cases are actually caused by an influenza virus in any given year. More than 200 known viruses and other pathogens can cause the suite of symptoms known as “influenza-like illness”; respiratory syncytial virus, bocavirus, coronavirus, and rhinovirus are just a few of the bugs that can make a person feel rotten. And depending on the season, in up to two-thirds of the cases of flu-like illness, no cause at all can be found.

No problems with this. It is the frustrating problem for those of us in the medical trenches: we can rarely diagnose quickly and inexpensively, those self-limited infections that are life-inconveniencing, but not-life threatening. Is the flu? Which can kill and infected others? Or a less potentially fatal disease?

Nobody knows precisely why we are much more likely to catch the flu in the winter months than at other times of the year. Perhaps it’s because flu viruses flourish in cool temperatures and are killed by exposure to sunlight. Or maybe it’s because in winter, people spend more time indoors, where a sneeze or a cough can more easily spread a virus to others. What is certain is that influenza viruses mutate with amazing speed, so each flu season sees slightly different genetic versions of the viruses that infected people the year before. Every year, the World Health Organization and the Centers for Disease Control and Prevention collect data from 94 nations on the flu viruses that circulated the previous year, and then make an educated guess about which viruses are likely to circulate in the coming fall. Based on that information, the U.S. Food and Drug Administration issues orders to manufacturers in February for a vaccine that includes the three most likely strains.

This, of course, is one of the three problems with the influenza vaccine that make determining efficacy difficult. The match between the circulating strain and what is in the vaccine is often less than perfect. The better the match between the vaccine and strain, the better the vaccine protection. And, as the virus mutates further and further away from the vaccine strain as the flu season progresses, it becomes more likely that the effect of the vaccine will be to cause a milder disease in the vaccinee but still, hopefully, lead to less likely spread of the disease.

Another problem with the vaccine is that those who need it most are least likely to not respond to the vaccine. More on that later.

The third problem is that the flu vaccine is, in part, a vaccine that gets more bang for the buck as the vaccination rates increase in populations. Unfortunately, we never get a large percentage of the population vaccinated. The best results, as the Ontario experience demonstrates, show that as vaccination rates increase for everyone, morbidity and mortality declines in the population. A rising tide, as they used to say, lifts all boats.

Every once in a while, however, a very different bug pops up and infects far more people than the normal seasonal flu variants do. It is these novel viruses that are responsible for pandemics, defined by the World Health Organization as events that occur when “a new influenza virus appears against which the human population has no immunity” and which can sweep around the world in a very short time. The worst flu pandemic in recorded history was the “Spanish flu” of 1918-19, at the end of World War I. A third of the world’s population was infected, with at least 40 million and perhaps as many as 100 million people dying–more than were killed in World Wars I and II combined. (Some scholars suggest that one reason World War I ended was that so many soldiers were sick or dying from flu.) Since then, two other flu pandemics have occurred, in 1957 and 1968, neither of which was particularly lethal.

Public Health and infectious disease specialists have been worrying a repeat of  the 1918 influenza pandemic. It is why the concern with the bird flu, which, while lethal in humans (66% mortality), is not easily spread human to human. When H1N1 hit in Mexico last spring, it wan not  known if the flu would be mild or fatal, although the preliminary Mexican mortality rates looked worrisome, those rates have not been repeated in the industrialized world.

Those in charge of preparing for pandemics or outbreaks always have two choices: over prepare and look like a fool if the pandemic does not occur or under prepare and look incompetent if the pandemic does occur. And make no mistake, if there is a fatal pandemic the repeats the  1918 experience, there will be no way to prepare, and those in charge, no matter what they do, will look like incompetent fools.

In August, the President’s Council of Advisors on Science and Technology projected that this fall and winter, the swine flu, H1N1, could infect anywhere between one-third and one-half of the U.S. population and could kill as many as 90,000 Americans, two and a half times the number killed in a typical flu season. But precisely how deadly, or even how infectious, this year’s H1N1 pandemic will turn out to be won’t be known until it’s over.

So what should they do? Prepare for worst case? Prepare for best case and hope that H1N1 isn’t all that bad? Something in between? I always have sympathy for those in charge of pandemic planning. No matter what decision is made, it will probably be the wrong one.

Most reports coming from the Southern Hemisphere in late August (the end of winter there) suggested that the swine flu is highly infectious, but not particularly lethal. For example, Australian officials estimated they would finish winter with under 1,000 swine flu deaths–fewer than the usual 1,500 to 3,000 from seasonal flu. Among those who have died in the U.S., about 70 percent were already suffering from congenital conditions like cerebral palsy or underlying illnesses such as cancer, asthma, or AIDS, which make people more vulnerable.

For example, the US has 13x the population of Australia, so that would be, hmmmmm, maybe 13,000 deaths in the US. As I write this it is healthy children and pregnant woman who have the greatest odds of dying of H1N1.

There is always that implication that those who died, well, were not worth saving. They were due to die. Culling the herd, they were sick and their time was up. Cancer, cerebral palsy, AIDS, well, they were supposed to dance with the reaper. They had a reason to die. I know the authors did not mean it that way, but it still rankles. I hear the same reasoning when we review hospital acquired infections and deaths: the patient often has numerous comorbid conditions that made the infection or death more likely. That does not stop us from asking what could we have done differently to have prevented the death. That the patient had an increased risk of death is not an excuse for a preventable illness or death.

It’s the same when people say flu kills ‘only’ x number of people. I had a flu-related death in a young person with no underlying medical problems this week. That patient was not an “only”. But as someone who is often partially  responsible for the critically ill, I always have a habit of personalizing deaths.

“The moral test of a government (and a health care system I will add) is how it treats those who are at the dawn of life, the children; those who are in the twilight of life, the aged; and those who are in the shadow of life, the sick and the needy, and the handicapped.” – Hubert Humphrey.

If this was a multiple choice test, I would say we might have pulled a D.

Public-health officials consider vaccine their most formidable defense against the pandemic–indeed, against any flu–and on the surface, their faith seems justified.

That’s right. Over 14,000 references on the flu vaccine covering the all aspects of the disease and forming multiple lines of evidence all converging on the efficacy of the vaccine is a superficial, surface reality. But underneath the surface we have to look for the truth THEY don’t want us to know. Sorry. That’s Kevin Trudeau’s line.

Back to pandemic, i.e. swine, flu,  although we are going to jump to seasonal flu in a moment. What is different about H1N1 is that no one has immunity. There is a huge population of susceptible people, maybe 60% of the population, that can get this disease, and influenza often has a 20-30% attack rate for susceptible people. So it should go through the country like grass through a goose. From a disease causing/biologic/immunologic point of view, it’s the same old same old. It is just another strain of flu, we just have an enormous number of people who can get this strain. There is no reason to suspect that if we can induce a good response to the viral hemagglutinin (the H) and the neuraminidase (the N) with the vaccine that people should not be protected from H1N1. Historically, when we have a good match with the flu and vaccine we  get great protection from acquiring the flu.

Vaccines developed over the course of the 20th century slashed the death rates of nearly a dozen infectious diseases, such as smallpox and polio, and vaccination became one of medicine’s most potent weapons. Influenza virus was first identified in the 1930s, and by the mid-1940s, researchers had produced a vaccine that was given to soldiers in World War II. The U.S. government got serious about promoting flu vaccine after the 1957 flu pandemic brought home influenza’s continuing potential to cause widespread illness and death. Today, flu vaccine is a staple of public-health policy; in a normal year, some 100 million Americans get vaccinated.

Yep. Less than a third. Not enough. That is part of the problem. Herd immunity depends on the vaccine and the organism, but as a rule of thumb when vaccination rates fall below 90% herd immunity fails.

But while vaccines for, say, whooping cough and polio clearly and dramatically reduced death rates from those diseases, the impact of flu vaccine has been harder to determine.

No, it hasn’t. As we will see,  the impact of flu vaccine has been demonstrated in numerous studies. If the end point is death, and that is the only end point you are interested in, then the data is not as clear cut as one would like. But the beneficial impact of flu vaccination is there, in the literature (the scientific literature, not a novel, which is a different kind of literature, not usually used as evidence).

If you only looked at deaths that air bags prevented, it would also not be as impressive than if you included the injuries and expense benefits of air bags. Vaccines have multitudinous benefits. Most of all the vaccine prevents people from getting the flu.  There are numerous placebo controlled trials in various populations to demonstrate decreased acquisition of flu by those who receive the vaccine. I reviewed this in detail in the vaccine efficacy post, complete with references.

Flu comes and goes with the seasons, and often it does not kill people directly, but rather contributes to death by making the body more susceptible to secondary infections like pneumonia or bronchitis. For this reason, researchers studying the impact of flu vaccination typically look at deaths from all causes during flu season, and compare the vaccinated and unvaccinated populations.

Such comparisons have shown a dramatic difference in mortality between these two groups: study after study has found that people who get a flu shot in the fall are about half as likely to die that winter–from any cause–as people who do not. Get your flu shot each year, the literature suggests, and you will dramatically reduce your chance of dying during flu season.

Yet in the view of several vaccine skeptics, this claim is suspicious on its face. Influenza causes only a small minority of all deaths in the U.S., even among senior citizens, and even after adding in the deaths to which flu might have contributed indirectly. When researchers from the National Institute of Allergy and Infectious Diseases included all deaths from illnesses that flu aggravates, like lung disease or chronic heart failure, they found that flu accounts for, at most, 10 percent of winter deaths among the elderly. So how could flu vaccine possibly reduce total deaths by half?

Welcome to the difficulties of clinical medicine. Conflicting studies. Different methodologies in different populations studied at different times have, what the?, different results. You mean there is variability in clinical trials? Water is wet? Fire is hot? What next? Crap. I guess I will have to read  all the studies, compare and contrast them,  and come to some conclusion. Or I could ask, say, Tom Jefferson.

Tom Jefferson, a physician based in Rome and the head of the Vaccines Field at the Cochrane Collaboration, a highly respected international network of researchers who appraise medical evidence, says: “For a vaccine to reduce mortality by 50 percent and up to 90 percent in some studies means it has to prevent deaths not just from influenza, but also from falls, fires, heart disease, strokes, and car accidents. That’s not a vaccine, that’s a miracle.”

Well, I doubt fires were a major cause of death in the elderly, but stroke and heart attacks are a well described long term complication of infections, perhaps due to inducing a long term pro-thrombotic effect, as I discussed in the Boost your Immune system post. If people get the flu vaccine they will be healthier and less likely to fall, break a hip, and die. They will be less likely, since they are healthy, to fall asleep with a lit cigarette and die in a house fire. If people get the flu vaccine and they have less complications and exacerbations of underlying diseases, and as a result they will be driving to their doctors less often, so  yes, he continues only a little sarcastically, they would be less likely to die in car accidents. Aspirin is the wonder drug that works wonders, perhaps the vaccine is the miracle drug that works miracles. The point, of course, is there are a lot of indirect benefits to being healthy. There are always numerous confounders that are hard to control for.

The estimate of 50 percent mortality reduction is based on “cohort studies,” which compare death rates in large groups, or cohorts, of people who choose to be vaccinated, against death rates in groups who don’t. But people who choose to be vaccinated may differ in many important respects from people who go unvaccinated–and those differences can influence the chance of death during flu season. Education, lifestyle, income, and many other “confounding” factors can come into play, and as a result, cohort studies are notoriously prone to bias. When researchers crunch the numbers, they typically try to factor out variables that could bias the results, but, as Jefferson remarks, “you can adjust for the confounders you know about, not for the ones you don’t,” and researchers can’t always anticipate what factors are likely to be important to whether a patient dies from flu. There is always the chance that they might miss some critical confounder that renders their results entirely wrong.

And there is always the chance that you DID NOT miss an unknown confounder that renders the results entirely wrong. There is the chance that “study after study” is wrong. Which is why you need to rely on multiple lines of evidence for vaccine efficacy.

What is the source of the figure of  50% mortality reduction? As written it sounds impressive.  I am sure I am about to get my ass handed to me as I discuss statistics, but I read Atlantic quote as an absolute decrease in mortality. But the numbers, when you look at the NEJM article, are a relative risk reduction to get the 50% decrease in mortality.

Of those who died, there were fewer vaccinated than unvaccinated who died. That is a flu of a different color.

From the NEJM,  the most reliable source of medical dogma,

“There were 713,872 person-seasons of observation. Vaccinated subjects were slightly older and had higher prevalence rates of all the baseline medical conditions except dementia or stroke “

“During the 10 influenza seasons, there were 4599 hospitalizations for pneumonia or influenza and 8796 deaths. The observed hospitalization rates for unvaccinated and for vaccinated participants were, on average, 0.7% and 0.6% per season, respectively, with corresponding death rates of 1.6% and 1.0% per season.

“Influenza vaccination was associated on average with substantial reductions in hospitalizations for pneumonia and for influenza (vaccine effectiveness, 27%; adjusted odds ratio, 0.73; 95% confidence interval [CI], 0.68 to 0.77) and in death (vaccine effectiveness, 48%; adjusted odds ratio, 0.52; 95% CI, 0.50 to 0.55). Estimates varied from season to season and across the 18 cohorts (Fig. 2). In the two seasons with a poor match between the vaccine and the virus strain, vaccine effectiveness was lower for reducing death (in seasons with a poor match, vaccine effectiveness was 37% [ad- justed odds ratio, 0.63; 95% CI, 0.57 to 0.69]; in seasons with a good match, vaccine effectiveness was 52% [adjusted odds ratio, 0.48; 95% CI, 0.46 to 0.51]) but not for reducing hospitalization.”

The study points out vaccine efficacy is variable and that

“Our inclusion criteria were designed to minimize the possibility of survivor bias, and we did not find evidence for a healthy-vaccinee effect in our analyses. Nevertheless, residual confounding may have influenced our results, and our sensitivity analyses indicate how our estimates of vaccine effectiveness would be lower, though still significant, after adjustment for the effect of a strong hypothetical unmeasured confounder.”

So anyone who reads this literature reads it with a grain of salt substitute. Part of the problem, I would suspect, is 50% is just such a nice number. I would speculate that it became part of the medical culture, like 98.6 being normal temperature, easy to remember, but not absolutely true. You know, “it depends”.

I will confess that until I read the Atlantic, I had never heard of 50% as an absolute number. I recognize, as do my colleagues, that all flu vaccine efficacy  statistics depends on “it depends” and are therefore not clear cut.

The most recent edition of the standard ID text, Mandell, Douglas and Bennett, has this to say about the flu vaccine in the elderly:

“Relatively few prospective trials of protective efficacy have been conducted in high-risk populations. In one placebo-controlled prospective trial in an older adult population, inactivated vaccine was approximately 58% effective in preventing laboratory-documented influenza. In addition, numerous retrospective case-control studies are available that have documented the effectiveness of inactivated influenza vaccines in older adults. Vaccine is protective against influenza- and pneumonia-related hospitalization in older adults, and it is accompanied by a decrease in all-cause mortality… Vaccine has also been shown to be protective in limited studies in other high-risk groups, including those with HIV infection.

It has recently been shown that inactivated vaccine administered to older adults and persons with coronary artery disease can reduce the rates of coronary events or stroke during the influenza season.”

No 50% there.

I also went to the CDC site, and no 50% there either.

At IDSA national meeting last week I went to a talk on communicating vaccine safety and one of the panel members mentioned the importance of not qualifying the issue of vaccine safety with phrases like the ‘data show no association between vaccine and autism’ which is a different message than ‘vaccines do not cause autism’. Give, we were told, clear cut, declarative statements without ambiguity or qualifiers. Which I find hard to do. Medicine is all about ambiguity and qualifiers.

I am struck by the ‘damned if you do damned if you don’t’ disadvantage that we SBM-oriented folks have. The anti-science goofs can simply repeat lies over and over. Vaccines cause autism. If I use the qualifiers and look at the nuance, I seem like an equivocating weasel. If I decide to use a simple message, like flu vaccines prevent death by 50%, I look like a lying shill for the medical industrial complex. But I doubt that the authors would resort to the ‘doctors are evil shills of the medical industrial complex’ gambit; that kind of characterization would be so cheesy in a reputable periodical.

But if you have a story tell, it is best to avoid the known qualifiers, nuance, and subtleties and set up a straw man you can beat like a mule. So, I suppose, it would be a straw mule argument.

When Lisa Jackson, a physician and senior investigator with the Group Health Research Center, in Seattle, began wondering aloud to colleagues if maybe something was amiss with the estimate of 50 percent mortality reduction for people who get flu vaccine, the response she got sounded more like doctrine than science. “People told me, ‘No good can come of [asking] this,'” she says. “‘Potentially a lot of bad could happen’ for me professionally by raising any criticism that might dissuade people from getting vaccinated, because of course, ‘We know that vaccine works.’ This was the prevailing wisdom.”

Nonetheless, in 2004, Jackson and three colleagues set out to determine whether the mortality difference between the vaccinated and the unvaccinated might be caused by a phenomenon known as the “healthy user effect.” They hypothesized that on average, people who get vaccinated are simply healthier than those who don’t, and thus less liable to die over the short term. People who don’t get vaccinated may be bedridden or otherwise too sick to go get a shot. They may also be more likely to succumb to flu or any other illness, because they are generally older and sicker. To test their thesis, Jackson and her colleagues combed through eight years of medical data on more than 72,000 people 65 and older. They looked at who got flu shots and who didn’t. Then they examined which group’s members were more likely to die of any cause when it was not flu season.

Jackson’s findings showed that outside of flu season, the baseline risk of death among people who did not get vaccinated was approximately 60 percent higher than among those who did, lending support to the hypothesis that on average, healthy people chose to get the vaccine, while the “frail elderly” didn’t or couldn’t.

Or it could be that the pleomorphic benefits of flu vaccination last longer than we suppose and extend beyond the flu season.

In fact, the healthy-user effect explained the entire benefit that other researchers were attributing to flu vaccine, suggesting that the vaccine itself might not reduce mortality at all. Jackson’s papers “are beautiful,” says Lone Simonsen, who is a professor of global health at George Washington University, in Washington, D.C., and an internationally recognized expert in influenza and vaccine epidemiology. “They are classic studies in epidemiology, they are so carefully done.”

The results were also so unexpected that many experts simply refused to believe them. Jackson’s papers were turned down for publication in the top-ranked medical journals. One flu expert who reviewed her studies for the Journal of the American Medical Association wrote, “To accept these results would be to say that the earth is flat!” When the papers were finally published in 2006, in the less prominent International Journal of Epidemiology, they were largely ignored by doctors and public-health officials. “The answer I got,” says Jackson, “was not the right answer.”

The Jackson article is a good and interesting paper. If the JAMA reviewer rejected it solely said the reason iof the statement quoted above, that would indeed be appalling. I hope that the editors of JAMA  not use her services again to review a paper. If it was rejected on that basis, then JAMA has some serious editorial issues. It would be of interest to see the entire text from all the reviewers. As a rule, reviewers are a wee bit more detailed in the reasons behind their rejection. Of course, if it were rejected by JAMA based on a detailed critique that concluded with the flat world statement, that would not make for a story of good and bad. But I am old school. I like to see the full text from which quotes are taken. I doubt a reputable periodical like the Atlantic would take quotes out of context to mislead readers. That would not be responsible.

There are  problems with the Jackson study but it is not the results.

One, it is an outlier, and outliers need confirmation. The preponderance of all the literature suggests that influenza vaccine prevents disease and death. If you do not get flu, you can’t die from flu or flu related illnesses. When outliers are published, people read them, think “huh, that’s interesting”, but there is going to have to be more than one contradictory study to change my practice. But if “study after study” shows mortality benefit, and one study does not, it is food for thought, but not necessarily the basis of changing practice. The results, above all, needs to be repeated by others on a different population with the same methods. In medicine we tend to be conservative about changing practice unless there is a preponderance of data to suggest a change is reasonable. Except, of course, if our big pharma overlords take us to a good streak house.

As a fellow, I read the reports of H. pylorii causing gastritis and ulcer that were starting to trickle in and, contrary to popular myth, I can remember discussing these article with my attending, a diarrhea expert. The attitude was, interesting, lets see if future data supports it. The myth that it was rejected and scorned has been discussed elsewhere.

The other annoying aspect is the division of mortality effect into a red state/blue state divide. Again with the binary. Medical results, especially epidemiological data, is always moving as data collection and analysis are refined. At 25 I was told that half of everything I learned in medical school would be proven wrong, the problem is that I would not know which half.

When I read the Jackson article and the NEJM article back to back, I am left  with the question not of yes/no as to mortality, but instead wondering what the magnitude of the effect is. Given all the other data on influenza vaccine efficacy, I have trouble abandoning vaccination of the elderly based on one epidemiologic study. Some day it may come to that, but not yet. Not enough data.

As Jackson said in the discussion, “We found the greatest reductions in the risk of death and of pneumonia hospitalization in the period before influenza season, when there should be no true vaccine effect.”

I am also not so sure that is true. The pro-thrombotic effects of infections can last for a year, as I discussed in the boost your immune system post, with a resultant increase in stroke, heart attack and pulmonary embolism, the first two accounting for significant mortality in the elderly. The effects of not getting the flu because you were vaccinated could persist a year, so I am not so sanguine the mortality effects of the vaccine would not be demonstrable longer than anticipated. That is one alternative explanation for the results and there may be others. Which means, of course, we need more studies. And it means the answer is not as clear as one would like.

The history of flu vaccination suggests other reasons to doubt claims that it dramatically reduces mortality. In 2004, for example, vaccine production fell behind, causing a 40 percent drop in immunization rates. Yet mortality did not rise. In addition, vaccine “mismatches” occurred in 1968 and 1997: in both years, the vaccine that had been produced in the summer protected against one set of viruses, but come winter, a different set was circulating. In effect, nobody was vaccinated. Yet death rates from all causes, including flu and the various illnesses it can exacerbate, did not budge. Sumit Majumdar, a physician and researcher at the University of Alberta, in Canada, offers another historical observation: rising rates of vaccination of the elderly over the past two decades have not coincided with a lower overall mortality rate. In 1989, only 15 percent of people over age 65 in the U.S. and Canada were vaccinated against flu. Today, more than 65 percent are immunized. Yet death rates among the elderly during flu season have increased rather than decreased.

This does not surprise me, as people get older and more frail they are less likely to respond to the vaccine. Unfortunately, depending on the host and the vaccine match, getting the vaccine does not mean the patient will be protected from disease. The vaccine mediated mortality effects on the elderly may not be as robust as we would like, depending on the subset of elderly vaccinated.

One issue not addressed is whether the deaths are shifted. Everyone dies eventually, and I have always wondered if the the vaccine  just shifted when everyone dies. But the rates stay the same, they mortality wave has moved downstream. I am sure I will get smacked for that statement.

Vaccine proponents call Majumdar’s last observation an “ecological fallacy,” because he fails, in their view, to consider changes in the larger environment that could have boosted death rates over the years–even as rising vaccination rates were doing their part to keep mortality in check. The proponents suggest, for instance, that influenza viruses may have become more contagious over time, and thus are infecting greater numbers of elderly people, including some who have been vaccinated. Or maybe the viruses are becoming more lethal. Or maybe the elderly have less immunity to flu than they once did because, say, their diets have changed.

Or maybe vaccine just doesn’t prevent deaths in the elderly. Of course, that’s the one possibility that vaccine adherents won’t consider. Nancy Cox, the CDC’s influenza division chief, says flatly, “The flu vaccine is the best way to protect against flu.” Anthony Fauci, a physician and the director of the National Institute of Allergy and Infectious Diseases at the NIH, where much of the basic science of flu vaccine has been worked out, says, “I have no doubt that it is effective in conferring some degree of protection. To say otherwise is a minority view.”

The flu vaccine IS the best way of protecting yourself against the flu. Except for living in a bubble with no human contact. And it is effective in conferring SOME degree of protection. Both statements are as true as a scientist can make them. I think it is something we all consider: what is the degree of effectiveness and in what subgroups. Any thoughtful physician, when taking care of a patient, asks how best to apply population data to the individual sitting in the office. Note, by the way, none of the vaccine adherents are saying a 50% decrease in mortality.

Majumdar says, “We keep coming up against the belief that we’ve reduced mortality by 50 percent,” and when researchers poke holes in the evidence, “people pound the pulpit.”

I can just see the preacher up there extorting the congregation: “I have no doubt that it is effective in conferring some degree of protection. To say otherwise is a minority view.” Ooohhhhhhh. That’s what I call fire and brimstone preaching. I have admit that in 25 years of infectious disease, I have not seen any pulpit pounding over the flu vaccine. Perhaps I have attended the wrong talks, and I suspect that PowerPoint, the preferred method of education, prevents pulpit pounding.

The most vocal–and undoubtedly most vexing–critic of the gospel of flu vaccine is the Cochrane Collaboration’s Jefferson, who’s also an epidemiologist trained at the famed London School of Tropical Hygiene, and who, in Lisa Jackson’s view, makes other skeptics seem “moderate by comparison.” Among his fellow flu researchers, Jefferson’s outspokenness has made him something of a pariah. At a 2007 meeting on pandemic preparedness at a hotel in Bethesda, Maryland, Jefferson, who’d been invited to speak at the conference, was not greeted by any of the colleagues milling about the lobby. He ate his meals in the hotel restaurant alone, surrounded by scientists chatting amiably at other tables. He shrugs off such treatment. As a medical officer working for the United Nations in 1992, during the siege of Sarajevo, he and other peacekeepers were captured and held for more than a month by militiamen brandishing AK-47s and reeking of alcohol. Professional shunning seems trivial by comparison, he says.

Yep. People are jerks. However, this is a story of a few lone, brave voices in the wilderness, their voices unheard by the Man. It really is not about the literature, so I suppose the Man is going to overreact.

“Tom Jefferson has taken a lot of heat just for saying, ‘Here’s the evidence: it’s not very good,'” says Majumdar. “The reaction has been so dogmatic and even hysterical that you’d think he was advocating stealing babies.” Yet while other flu researchers may not like what Jefferson has to say, they cannot ignore the fact that he knows the flu-vaccine literature better than anyone else on the planet.

It is not a matter of like, as if it were personal preference, like choosing a style of beer. I disagree with Dr. Jefferson’s conclusions from reading the literature. I appreciate his analysis, agree with his criticisms, but I, and many others, think his conclusions are wrong. The Atlantic  makes it sound as if people agree with his conclusions yet ignore them. If I disagree with someones conclusions, I am hardly going to act on them.

He leads an international team of researchers who have combed through hundreds of flu-vaccine studies. The vast majority of the studies were deeply flawed, says Jefferson. “Rubbish is not a scientific term, but I think it’s the term that applies.”

Rubbish? No. Not perfect? Yep. Welcome to the real world. Again this binary approach: either the data is perfect or its rubbish. I am just a clinician who spends most of my time taking care of people with infections. Trying to decide how much morbidity and cost you want to inflict on a patient before you have enough data to make a diagnostic and therapeutic decision is not easy. I could, as  an example, probably diagnose every pneumonia by sending every patient off for an open lung biopsy and combining it with 1000’s of dollars of diagnostic testing. But to what end? In the real world one has to be satisfied with diagnostic uncertainty, especially if you feel that you are not missing a diagnosis that could hurt or kill someone if left untreated.

Only four studies were properly designed to pin down the effectiveness of flu vaccine, he says, and two of those showed that it might be effective in certain groups of patients, such as school-age children with no underlying health issues like asthma. The other two showed equivocal results or no benefit.

All that time and money spent on trials and every almost every study was incompetent and flawed. If only they had asked Dr. Jefferson first as only he knows how to do a proper clinical study. Must be hard to be as knowledgeable as Dr. Jefferson, with all that knowledge and expertise, and to have to be ignored. I mention this with the sarcasm as in the Atlantic Dr. Jeffersonis represented as the be all and end all of designing studies, yet, as best I can tell from Pubmed, he has never participated in a clinical trial. He usually does reviews of studies done by others. Of course, theoretical expertise in an area does not necessarily invalidate criticisms; the Pope’s opinions on contraception may be valid despite practical experience.

What is the topic? Preventing flu or preventing death?  My entry on flu efficacy discusses the data and difficulty with deciding exactly what are the benefits of the flu vaccine. The problem with popular articles is the lack of references, so it is hard to comment. Effective how? Endpoints? What population were studied? There is so little real substance in the last paragraph that I cannot tell if it is rubbish or not. But it sure sounds like someone does not know what they are doing.

Flu researchers have been fooled into thinking vaccine is more effective than the data suggest, in part, says Jefferson, by the imprecision of the statistics.

Huh? Jackson uses statistics to show the vaccine doesn’t work. I wish they would explain why one set of statistics is good and those of others are not. It would be nice to know which statistics are statistically appropriate. Perhaps they have been ‘fooled’ by multiple lines of convergent evidence that consistently demonstrates efficacy of the influenza vaccine on multiple endpoints. Damn, maybe all 14,000 references on the  influenza vaccine have fooled me.

The only way to know if someone has the flu–as opposed to influenza-like illness–is by putting a Q-tip into the patient’s throat or nose and running a test, which simply isn’t done that often.

That is simply not true, although the recent NEJM reference that looked at vaccine efficacy was probably not published before the Atlantic was written. However, the direct inoculation studies did stick a swab in the patient for culture after giving a nose full of flu after placebo or vaccine. So in the best of all possible worlds, we know that vaccine can prevent flu. Problem is, of course, translating the perfect world to the real world.

Likewise, nobody really has a handle on how many of the deaths that are blamed on flu were actually caused by a flu virus, because few are confirmed by a laboratory. “I used to be a family physician,” says Jefferson. “I’ve never seen a patient come to my office with H1N1 written on his forehead.When an old person dies of respiratory failure after an influenza-like illness, they nearly always get coded as influenza.”

Again, no argument with this. I wonder if the coding issue is true, it is a reference free statement, but coding data, which by the way both the NEJM and Jackson used to calculate their results, are known to be unreliable. So the data from Jacksons codes are good and the NEJM are not? Again, the problem of how deaths from flu are calculated is discussed on a blog entry. Real world = messy.

There’s one other way flu researchers may be fooled into thinking flu vaccine is effective, Jefferson says. All vaccines work by delivering a dose of killed or weakened virus or bacteria, which provokes the immune system into producing antibodies. When the person is subsequently exposed to the real thing, the body is already prepared to repel the bug completely or to get rid of it after a mild illness. Flu researchers often use antibody response as a way of gauging the effectiveness of vaccine, on the assumption that levels of antibodies in the blood of people who have been vaccinated are a good predictor–although an imperfect one–of how well they can ward off the infection.

Antibodies are a surrogate for efficacy and, as a rule, antibodies levels are predictive for vaccine efficacy. Fooled by statistics, fooled by antibodies, we docs sure are a gullible bunch of rubes who just fell off the turnip truck. It fools no one. I fail to follow how an antibody response fools people into thinking the flu vaccine is effective. As the Cochrane review points out, the best effect of the flu vaccine occurs when there is a good match between the vaccine and flu, and the only way that is going to happen that I can think of  is that the vaccine generated a good antibody response.

A meta analysis of the vaccine efficacy in 66,248 people (almost Cardiology level of patients involvement) from 2004:

“Inactivated parenteral vaccines were 30% effective (95% CI 17% to 41%) against influenza-like illness, and 80% (95% CI 56% to 91%) efficacious against influenza when the vaccine matched the circulating strain and circulation was high, but decreased to 50% (95% CI 27% to 65%) when it did not. Excluding the studies of the 1968 to 1969 pandemic, effectiveness was 15% (95% CI 9% to 22%) and efficacy was 73% (95% CI 53% to 84%). Vaccination had a modest effect on time off work, but there was insufficient evidence to draw conclusions on hospital admissions or complication rates. Inactivated vaccines caused local tenderness and soreness and erythema. Spray vaccines had more modest performance. Monovalent whole-virion vaccines matching circulating viruses had high efficacy (VE 93%, 95% CI 69% to 98%) and effectiveness (VE 66%, 95% CI 51% to 77%) against the 1968 to 1969 pandemic.

AUTHORS’ CONCLUSIONS: Influenza vaccines are effective in reducing cases of influenza, especially when the content predicts accurately circulating types and circulation is high. However, they are less effective in reducing cases of influenza-like illness and have a modest impact on working days lost. There is insufficient evidence to assess their impact on complications.”

There’s some merit to this reasoning. Unfortunately, the very people who most need protection from the flu also have immune systems that are least likely to respond to vaccine. Studies show that young, healthy people mount a glorious immune response to seasonal flu vaccine, and their response reduces their chances of getting the flu and may lessen the severity of symptoms if they do get it. But they aren’t the people who die from seasonal flu. By contrast, the elderly, particularly those over age 70, don’t have a good immune response to vaccine–and they’re the ones who account for most flu deaths. (Infants with severe disabilities, such as leukemia and congenital lung disease, and people who are immune-compromised–from AIDS, or diabetes, or cancer treatment–make up the rest. As of August8, only 36 deaths from swine flu had been confirmed among children in the U.S., and the overwhelming majority of those children had multiple, severe health disorders.)

Again, I have issues with that ‘only’. Tell 36 mothers that their child was an ‘only’. I know. Little things like compassion should not be part of the equation, but I am a clinician, not a scientist. We shouldn’t make policy on the basis of a few children dying. Now its ‘only’ about 100, mostly healthy, dead children as of this posting. That is a question that I will not even try an answer. At what point do the number of deaths go from ‘only’ to ‘worrisome’? What is the acceptable death rate in a population? Got me. In the 80’s, if it were AIDS in gay men, any death rate was acceptable, but no death in a Legionnaire was acceptable. Me bitter much?

In Jefferson’s view, this raises a troubling conundrum: Is vaccine necessary for those in whom it is effective, namely the young and healthy? Conversely, is it effective in those for whom it seems to be necessary, namely the old, the very young, and the infirm? These questions have led to the most controversial aspect of Jefferson’s work: his call for placebo-controlled trials, studies that would randomly give half the test subjects vaccine and the other half a dummy shot, or placebo. Only such large, well-constructed, randomized trials can show with any precision how effective vaccine really is, and for whom.

It is not radical but unethical to do these studies. Some basic principals of human research include:

Research should be based on a thorough knowledge of the scientific background (Article 11), a careful assessment of risks and benefits (Articles 16, 17), have a reasonable likelihood of benefit to the population studied (Article 19)

We have so much data that show beneficial effects of influenza vaccine and it would be difficult to find a physician to participate. As I have mentioned before, not unless you really want to have a death panel for the elderly. It would also make for interesting informed consent.

There is a third option discussed later, universal vaccination.

In the flu-vaccine world, Jefferson’s call for placebo-controlled studies is considered so radical that even some of his fellow skeptics oppose it. Majumdar, the Ottawa researcher, says he believes that evidence of a benefit among children is established and that public-health officials should try to protect seniors by immunizing children, health-care workers, and other people around them, and thus reduce the spread of the flu. Lone Simonsen explains the prevailing view: “It is considered unethical to do trials in populations that are recommended to have vaccine,” a stance that is shared by everybody from the CDC’s Nancy Cox to Anthony Fauci at the NIH. They feel strongly that vaccine has been shown to be effective and that a sham vaccine would put test subjects at unnecessary risk of getting a serious case of the flu. In a phone interview, Fauci at first voiced the opinion that a placebo trial in the elderly might be acceptable, but he called back later to retract his comment, saying that such a trial “would be unethical.” Jefferson finds this view almost exactly backward: “What do you do when you have uncertainty? You test,” he says. “We have built huge, population-based policies on the flimsiest of scientific evidence. The most unethical thing to do is to carry on business as usual.”

Flimsiest of scientific evidence? Rubbish? The man is good with hyperbole. There are now 5 studies that demonstrate that vaccinating those around the elderly for influenza decreases death or heart attacks in the elderly.

The Ontario experience, which I am sure  is deeply flawed to Dr. Jefferson, is still compelling: the greater the number of people vaccinated, the greater the decrease in mortality in the community. Most studies show benefit from the vaccine.

When you have uncertainty, which is always the case in medicine, you try and decrease the uncertainty with tests. But not by potentially killing people. And that potential is not trivial.

It would be an interesting informed consent: a large number of studies suggests that the vaccine is protective for morbidity and mortality but we want to prove it with a placebo controlled trial. Are you going to sign up? And you have to ask: if the test proposed by Dr. Jefferson, a large trial of vaccine in the elderly with death as  an endpoint is accomplished and it shows that the elderly do have increased mortality if they are not vaccinated, which is strongly suggested by the preponderance of data we already have, what are you going to tell the next of kin of the dead? We were right, thanks for the sacrifice.

Such a trial would have been doable 30 years ago, but not now.

All epidemiologic studies have issues:

“Recent studies questioning the plausibility of reported mortality benefits among vaccinated elderly persons may themselves be based on assumptions that are susceptible to important limitations and multiple biases. Future studies that incorporate prospectively collected information on functional status, life expectancy, and other types of data may provide additional insights into these concerns. At present, even after taking into account the potential for residual bias and confounding, most studies confirm the benefits of vaccination among the elderly for reducing hospitalization and death (19840665).”

Of course, to discuss that one would have to abandon the good/bad dichotomy that makes for a good story. But that is a reasoned review, with qualifiers and nuance. Doesn’t make for a good story.

Just after 6 p.m. on a warm Friday evening in July, Dr. David Newman is only minutes into a 10-hour shift in the emergency room of New York City’s St. Luke’s Hospital, and already he has assumed responsibility for 11 patients. The young Italian tourist sitting on the bed in front of the doctor has meningitis, and through an interpreter, Newman tells him he almost certainly has the viral form of the disease, which will do nothing more than make him feel ill for a few days. There is a tiny chance, says Newman, that the illness is caused by a bacterium, which can be deadly, but he is almost positive that’s not what the tourist has. He says to his patient, “I can’t tell you with 100 percent certainty that you don’t have it, but if you do, you’ll begin to feel worse and you’ll need to come back.” The tourist, on learning that he might be infected with a potentially lethal disease, looks down at his feet and confesses that he is much more worried about another illness: swine flu. Newman smiles patiently. “It would be nice if you had swine flu,” he says. “Compared to bacterial meningitis, swine flu is safe.”

Late last spring, as headlines and airwaves warned of a possible pandemic, patients like Newman’s began clogging emergency rooms across the country, a sneezing, coughing, infectious tide of humanity more worried than truly sick, but whose mere presence in the emergency room has endangered the lives of others.

“Studies show that when there is ER crowding, mortality goes up, because patients who need immediate attention don’t get it,” says Newman, the director of clinical research in the Department of Emergency Medicine at the hospital, which is affiliated with Columbia University. In an average year the ER at St. Luke’s, a sprawling 1,076-bed hospital on 113th Street, takes in 110,000 patients, some 300 a day. At the height of the summer swine flu outbreak, that number doubled. The vast majority of panicky patients who came in the door at St. Luke’s and other emergency departments didn’t actually have the virus, and of those who did, most were not sick enough to need hospitalization. Even so, says Newman, when patients with even mild flu symptoms show up in the hospital, they vastly increase the spread of the virus, simply because they inevitably sneeze and cough in rooms that are jammed with other people.

Hmmmm. Maybe, the tide of humanity could benefit from, hmmmm, a vaccine. To prevent H1N1 and decrease spread. Given the good match from the vaccine and the current strain, one would expect an excellent immune response in the young and if enough people get vaccinated then herd immunity may kick in, protecting those who can’t respond to the vaccine. It would keep the ER from being a cesspool of contagion.

Many of the worried sick come to St. Luke’s and other hospitals in search of antiviral drugs. The CDC recommends the use of two drugs against H1N1: oseltamivir and zanamivir, better known by their brand names, Tamiflu and Relenza, which together form the second pillar of the government’s anti-pandemic-flu strategy. Public-health officials at the state and local levels are also recommending the drugs. Guidelines issued by the New York City Department of Health, says Newman, “encourage us to give a prescription to just about every patient with the sniffles,” a practice that some experts worry will quickly lead to resistant strains of the virus.

We do not have the resources to test everyone for H1N1, so currently we are testing those who are admitted to the hospital. And we were never “treating anyone with the sniffles”. Dr. Newman exaggerates.

The first New York State Public Health communication says “Antiviral treatment for confirmed or suspected ill cases of swine influenza virus infection may include either oseltamivir or zanamivir, with no preference given at this time. Initiate treatment as soon as possible after the onset of symptoms.” Most of the CDC guidelines have said pretty much the same thing since the onset of the pandemic. What we have learned is that the lethality of H1N1 has been less than feared and as a result recommendations as to when to start antivirals has changed over time. But sniffles? Puh leze.

Remember that when H1N1 started in Mexico it had an high mortality rate (maybe 4%, we didn’t know with certainty, but 4% is 1918 level mortality) that was subsequently not replicated in the US. Under those circumstances, early treatment was reasonable when there was no vaccine to prevent the disease. Current recommendations are to treat those at risk of dying of flu. As the pandemic has progressed we have learned and adapted to the situation.

Indeed, that’s already happening. Daniel Janies, an associate professor of biomedical informatics at Ohio State University, tracks the genetic mutations that allow flu virus to develop resistance to drugs. Flu can become resistant to Tamiflu in a matter of days, he says. Handing out the drug early in the pandemic, when H1N1 poses only a minimal threat to the vast majority of patients, strikes him as “shortsighted.”

That is the question and problem with all antibiotics and antivirals: the trade off for the short term use and benefit is long term resistance and loss of drug efficacy. It just happens faster with influenza. Certainly  now that we know it is a mild illness in most people  the drug should be reserved for those at risk of death and the very ill. Hindsight is always 20:20 and people forgets the uncertainty associated with the beginning of the outbreak.

Indeed, samples of resistant H1N1 were cropping up by midsummer, increasing the likelihood that come late fall, many people will be infected with a resistant strain of swine flu. Alarmed at that prospect, the World Health Organization issued an alert on August 21, recommending that Tamiflu and Relenza be used only in severe cases and in patients who are at high risk of serious complications. By mid-August, two U.S. swine flu patients had developed Tamiflu-resistant strains.

The Borg did have it partially right. Resistance is inevitable and we try to be judicious with the use of antivirals. As is often the case with novel infections, at first the response is to play it safe and be aggressive with treatment and isolation until you have more data and know it is reasonable to be less aggressive. Those of us in the trenches, who get to watch people suffer and die from these diseases and are responsible for their lives, cannot take that responsibility lightly. I have to bet your life that I make the right decision, and at the beginning of the H1N1 pandemic, the feeling was to play it safe.

The U.S. first began stockpiling Tamiflu and Relenza back in 2005, in the wake of concern that an outbreak in Southeast Asia of bird flu, a far more deadly form of the disease, might go global. On November 1, 2005, President George W.Bush pronounced pandemic flu a “danger to our homeland,” and he asked Congress to approve legislation that included $1billion for the production and stockpiling of antivirals. This was after Congress had already approved $1.8billion to stockpile Tamiflu for the military, a decision that was made during the tenure of Defense Secretary Donald Rumsfeld. (Before joining the Bush Cabinet, Rumsfeld was chairman for four years of Gilead Sciences, the company that holds the patent on Tamiflu, and he held millions of dollars’ worth of stock in the company. According to Roll Call, an online newspaper covering events on Capitol Hill, Rumsfeld says he recused himself from all government decisions involving Tamiflu. Gilead’s stock price rose more than 50 percent in 2005, when the government’s plan was announced.)

The perfidious influence of big pharma. Even if the right decision is made, trying to prepare for a repeat of the 1918 pandemic, it is tainted by Rumsfeld.

As with vaccines, the scientific evidence for Tamiflu and Relenza is thin at best.

The evidence isn’t thin, but the effects of the drug in healthy people are modest. But that is not why Tamiflu was being stockpiled. Bird flu, H1N5, has a 66% death rate. But for a variety of reasons is not very infectious. What would happen if there was another re-assortment in a pig somewhere with the current H1N1 and the current bird flu and the new strain became both contagious and fatal and there was no vaccine? The best you could do was decrease morbidity and mortality with early treatment with Tamiflu. Early on it appeared that H1N1 was going to be lethal.

In its general-information section, the CDC’s Web site tells readers that antiviral drugs can “make you feel better faster.” True, but not by much. On average, Tamiflu (which accounts for 85 to 90 percent of the flu antiviral-drug market) cuts the duration of flu symptoms by 24hours in otherwise healthy people. In exchange for a slightly shorter bout of illness, as many as one in five people taking Tamiflu will experience nausea and vomiting. About one in five children will have neuropsychiatric side effects, possibly including anxiety and suicidal behavior. In Japan, where Tamiflu is liberally prescribed, the drug may have been responsible for 50 deaths from cardiopulmonary arrest, from 2001 to 2007, according to Rokuro Hama, the chair of the Japan Institute of Pharmacovigilance.

Such side effects might be worth risking if the antivirals prevented serious complications of flu, such as pneumonia, hospitalization, and death. Roche Laboratories, the company licensed to manufacture and mark et Tamiflu, says its drug does just that. In two September 2006 press releases, the company announced, “Tamiflu significantly reduces the risk of death from influenza: New data shows treatment was associated with more than a two third reduction in deaths,” and “Children with influenza [are] 53 percent less likely to contract pneumonia when treated with Tamiflu.”

I addressed this at the top of the post. Again with the binary approach. Antivirals have always had modest effects in treating infections. But during a pandemic, having  30% of the work force potentially out sick with flu, is also going to have adverse effects. It would be nice when I am in a major car accident and I am rushed to the trauma center that the staff is actually working and not out ill with influenza.

Once again cohort studies (the same kind of potentially biased research that led to the conclusion that flu vaccine cuts mortality by 50 percent) are behind these claims. Tamiflu costs $10 a pill. It is possible that people who take it are more likely to be insured and affluent, or at least middle-class, than those who do not, and a large body of evidence shows that the well-off nearly always fare better than the poor when stricken with an infectious disease, including flu. In both 2003 and 2009, reviews of randomized placebo-controlled studies found that the study populations simply weren’t large enough to answer the question: Does Tamiflu prevent pneumonia?

What I want Tamiflu to do, and it does, is decrease the chance that my patient will die of severe flu. Decrease the chance. That’s the best one can for with an antiviral, and if there is not an effective vaccine, the best I will be able to offer my patients.

But prevent pneumonia? Huh? Is this a reference to prophylactic studies? And does the author mean influenza or bacteria pneumonia? This kind of obscure sentence drives. Me. Nuts. It casts doubt with obscurity.

Again with be binary, does it prevent pneumonia? What one learns early in medicine is that the effects of our interventions, are incremental and additive. No single intervention is a magic bullet to treat or prevent disease. Flu treatment and prevention will be most efficacious when we do many interventions.

As late as this August, the company’s own Web site contained the following statement, which was written under the direction of the FDA: “Tamiflu has not been proven to have a positive impact on the potential consequences (such as hospitalizations, mortality, or economic impact) of seasonal, avian, or pandemic influenza.” An FDA spokesperson said recently that the agency is unaware of any data submitted by Roche that would support the claims in the company’s September 2006 news release about the drug’s reducing flu deaths.

Why, then, has the federal government stockpiled millions of doses of antivirals, at a cost of several billion dollars?

Because there is some effectiveness of the medication against the disease and should there be a Tamiflu sensitive flu pandemic and we do not have an effective vaccine, we may be able to save a few people who might have otherwise died. Again, welcome to the messy world of the practice of medicine.

And why are physicians being encouraged to hand out prescriptions to large numbers of people, without sound evidence that the drugs will help?

They are not. Here in the Northwest, most patients are not being treated with Tamiflu now that we know who may and who may not die of the illness. In medicine if you wait for perfect studies to give irrefutable answers, good luck. If that is the standard, then we will do nothing. But I disagree with the author. I think the preponderance of data is good enough to make decisions: treat the really ill and those at high risk of death and vaccinate as many as you can to prevent the disease.

The short answer may be that public-health officials feel they must offer something, and these drugs are the only possible remedies at hand. “I have to agree with the critics the antiviral question is not cut-and-dried,” says Fauci. “But [these drugs are] the best we have.” The CDC’s Nancy Cox also acknowledges that the science is not as sound as she might like, but the government still recommends their use. And as with vaccines, she considers additional randomized placebo-controlled trials of the antiviral drugs to be “unethical” and thus out of the question.

This is the curious state of debate about the government’s two main weapons in the fight against pandemic flu.

It is not curious. It is the real world, with limited time and resources. We have to fight the war, as that evil bastard said, with the Army we have. Sometimes even the irreducibly corrupt will make the right decision.

At first, government officials declare that both vaccines and drugs are effective. When faced with contrary evidence, the adherents acknowledge that the science is not as crisp as they might wish. Then, in response to calls for placebo-controlled trials, which would provide clear results one way or the other, the proponents say such studies would deprive patients of vaccines and drugs that have already been deemed effective. “We can’t just let people die,” says Cox.

We can’t let people die. Maybe it is me, but the tone of the article is that “unethical” and “not letting people die” are used in a manner to suggest these are not valid reasons for the current policy. They are very valid reasons to making decisions.

Students of U.S. medical history will find this circular logic familiar: it is a long-recurring theme in American medicine, and one that has, on occasion, had deadly consequences. In 1925, Sinclair Lewis caricatured a medical culture that allowed belief–and profits–to distort science in his Pulitzer Prize-winning book, Arrowsmith. Based on the lives of the real-life microbiologists Paul de Kruif and Jacques Loeb, Lewis tells the story of Martin Arrowsmith, a physician who invents a new vaccine during a deadly outbreak of bubonic plague. But his efforts to test the vaccine’s efficacy are frustrated by an angry community that desperately wants to believe the vaccine works, and a profit-hungry institute that rushes the vaccine into use prematurely–forever preempting the proper studies that are needed.

So a work of fiction is now considered part of medical history? Great. In that case I think we should use Marcus Welby MD as example of the history of the profession. He was a nice, caring, MD who did what ever he could to help his patients. I love the profession being reduced to an evil caricature. Well, as long as we are using a work of fiction as a characterization of how medicine works, might as well go to the Onion to see how journalism works. If me and mine are going to be reduced to an evil cliched stereotype, lets do the same with journalists.

I have to admit this kind of characterizaion  irritates me. I just came back from the national infectious disease meetings, and, while there are docs who are just as greedy and ignoble as any journalist, what drives people in public health and the CDC is to do good. I know no one believes that. It sure isn’t the money.

The annals of medicine are littered with treatments and tests that became medical doctrine on the slimmest of evidence, and were then declared sacrosanct and beyond scientific investigation. In the 1980s and ’90s, for example, cancer specialists were convinced that high-dose chemotherapy followed by a bone-marrow transplant was the best hope for women with advanced breast cancer, and many refused to enroll their patients in randomized clinical trials that were designed to test transplants against the standard–and far less toxic–therapy. The trials, they said, were unethical, because they knew transplants worked.

False analogy. With the treatment of breast cancer with bone marrow transplant, there was not 50 years of convergent data to suggest bone marrow transplants worked for breast cancer. Nothing. Again, we have a rich data set to show that flu vaccine decreases flu morbidity and mortality.

When the studies were concluded, in 1999 and 2000, it turned out that bone-marrow transplants were killing patients. Another recent example involves drugs related to the analgesic lidocaine. In the 1970s, doctors noticed that the drugs seemed to make the heart beat rhythmically, and they began prescribing them to patients suffering from irregular heartbeats, assuming that restoring a proper rhythm would reduce the patient’s risk of dying. Prominent cardiologists for years opposed clinical trials of the drugs, saying it would be medical malpractice to withhold them from patients in a control group. The drugs were widely used for two decades, until a government-sponsored study showed in 1989 that patients who were prescribed the medicine were three and a half times as likely to die as those given a placebo.

Again false analogy. I remember the day when we gave everyone lidocaine and other antiarrhythmics, back when I was an intern. Antiarrhythmics are toxic drugs. We knew it in the day, we thought they were less toxic than the diseases they treated Antivirals and the vaccines have a long track record of safety and significant toxicities are rare.

Demonstrating the efficacy (or lack thereof) of vaccine and antivirals during flu season would not be hard to do, given the proper resources.

Remember I told you that was coming. Now its not hard to do with the proper resources. Getting to Mars would not be hard to do with the proper resources. A quarter of our population is uninsured and as a result has double the mortality rate when compared to the insured. Wait. That’s not based on a double blind study either. So it probably isn’t true. The best thing to do is take away everyone’s insurance, since it is the well-to-do who are healthy and can afford health care and it is not access to health care that makes the difference.

The best solution would be to take all the money used to fund the NCCAM, as they have never demonstrated efficacy of alternative therapies, and spend the money on influenza research. That would be a win-win.

Take a group of people who are at risk of getting the flu, and randomly assign half to get vaccine and the other half a dummy shot. Then count the people in each group who come down with flu, suffer serious illness, or die. (A similarly designed trial would suffice for the antivirals.) It might sound coldhearted, but it is the only way to know for certain whether, and for whom, current remedies actually work.

I love that cold hearted. Evidently the authors haven’t watched someone die of influenza or its complications, a clinical problem I am getting reacquainted with thanks to H1N1. I think there is enough data with flu vaccine efficacy to have a high expectation that the vaccine will prevent flu and, as a result, its complications. I don’t think we need to kill a cohort of the elderly to prove it, but I also doubt such a trial is feasible in that it would be difficult on enroll people given our current knowledge.

It would also be useful to know whether vaccinating healthy people–who can mount an immune response on their own–protects the more vulnerable people around them. For example, immunizing nursing-home staff and healthy children is thought to reduce the spread of flu to the elderly and the immune-compromised. Pinning down the effectiveness of this strategy would be a bit more complex, but not impossible.

We have four studies that show just this effect: vaccinating health care workers decreases mortality in their patients, although the effect is  nuanced and variable. Again. As Dr. Jeffersons review said,

“Staff vaccination had a significant effect on influenza-like illness (vaccine effectiveness [VE] 86%, 95% CI 40-97%) only when patients were vaccinated too. If patients were not vaccinated, staff immunization had no effect. Vaccinating health-care workers did not appear efficacious against influenza (RR 0.87, 95% CI 0.46-1.63). There was no significant effect of vaccination on lower respiratory tract infections: (RR 0.70, 95% CI 0.41-1.20). Deaths from pneumonia were significantly reduced (VE 39%, 95% CI 2-62%), as were deaths from all causes (VE 40%, 95% CI 27-50%) (16631547).”

In the absence of such evidence, we are left with two possibilities. One is that flu vaccine is in fact highly beneficial, or at least helpful.

Again with the binary. The effect of the flu vaccine and other interventions is a continuum. It depends on who gets it, their ability to respond with antibody and whether they get secondary complications. The data is clear that it is helpful. It is if you read ALL the literature.

Solid evidence to that effect would encourage more citizens–and particularly more health professionals–to get their shots and prevent the flu’s spread. As it stands, more than 50 percent of health-care workers say they do not intend to get vaccinated for swine flu and don’t routinely get their shots for seasonal flu, in part because many of them doubt the vaccines’ efficacy.

I spend each flu season talking to my colleagues about the vaccine and the reasons they do not get the vaccine rarely have anything to do with the data, see for a smart ass rant on the topic.

The other possibility, of course, is that we’re relying heavily on vaccines and antivirals that simply don’t work, or don’t work as well as we believe. And as a result, we may be neglecting other, proven measures that could minimize the death rate during pandemics.

What is the difference between ‘do not work as well as we believe’ and ‘at least helpful’?  Again, I keep harping on the issue that the response to the flu vaccine is nuanced, not black and white. And what are these other issues to minimize the death rates?

“Vaccines give us a false sense of security,” says Sumit Majumdar. “When you have a strategy that [everybody thinks] reduces death by 50 percent, it’s pretty hard to invest resources to come up with better remedies.” For instance, health departments in every state are responsible for submitting plans to the CDC for educating the public, in the event of a serious pandemic, about hand-washing and “social distancing” (voluntary quarantines, school closings, and even enforcement of mandatory quarantines to keep infected people in their homes). Putting these plans into action will require considerable coordination among government officials, the media, and health-care workers–and widespread buy-in from the public. Yet little discussion has appeared in the press to help people understand the measures they can take to best protect themselves during a flu outbreak–other than vaccination and antivirals.

Social distancing is suggested. And based on what randomized controlled trial? The data to support social distancing are limited and in a recent Annals article was not optimistic:

“Sixty (8%) contacts in the 259 households had RT-PCR– confirmed influenza virus infection in the 7 days after intervention. Hand hygiene with or without facemasks seemed to reduce influenza transmission, but the differences compared with the control group were not significant. In 154 households in which interventions were implemented within 36 hours of symptom onset in the index patient, transmission of RT-PCR–confirmed infection seemed reduced, an effect attributable to fewer infections among participants using facemasks plus hand hygiene (adjusted odds ratio, 0.33 [95% CI, 0.13 to 0.87]). Adherence to interventions varied.”

All of us in health care know the difficulty in getting patients to adhere with recommended interventions. Not mentioned is Dr. Jefferson’s review of this issue where he notices that the quality of the data to support other means to prevent influenza is not very good:

”The quality of the studies was poor for all four randomized controlled trials and most cluster randomized controlled trials; the observational studies were of mixed quality” and “The quality of the methods varied in these studies but may reflect the difficult and real life circumstances in which they were carried out (19773323).”

What a surprise. More less-than-perfect studies. You know, if we just had the resources…

“Launched early enough and continued long enough, social distancing can blunt the impact of a pandemic,” says Howard Markel, a pediatrician and historian of medicine at the University of Michigan. Washing hands diligently, avoiding public places during an outbreak, and having a supply of canned goods and water on hand are sound defenses, he says.

Shut down society. It would work. But if 30% are infected over three months, as worst case scenarios show, who will be around to care for the ill?

Such steps could be highly effective in helping to slow the spread of the virus. In Mexico, for instance, where the first swine flu cases were identified in March, the government launched an aggressive program to get people to wash their hands and exhorted those who were sick to stay home and effectively quarantine themselves. In the United Kingdom, the national health department is promoting a “buddy” program, encouraging citizens to find a friend or neighbor willing to deliver food and medicine so people who fall ill can stay home.

I also have a career devoted to trying to get dedicated health care workers to consistently wash their hands. The reasons for the prior failures to increase compliance are multitudinous and the reason for our current excellent success are complex, but expecting the public to wash their hands reliably is probably not realistic. As the Annals article said:

“Only half of the index patients in the face mask plus hand hygiene group re- ported regular use of a surgical mask during follow-up. Face mask adherence among household contacts was lower. Adherence to the hand hygiene intervention seemed low compared with rates recommended in health care settings but was similar to rates in previous community studies. In addition, effects in our study may tend toward a lower bound on the effects that might be observed in a pandemic with heightened public awareness.”

Everything seems so simple: do a quick study on Tamiflu and vaccine efficacy, have people wash their hands. Why didn’t the CDC think of that? Bumbling, incompetent bureaucrats one and all. And while we are at it, lets have everyone lose weight and exercise regularly.

And once you have influenza and are ill enough to be in the hospital, heading for a ventilator and the perpetual night, what are you doing to do instead of Tamiflu? Social distancing? Got me. I am open for ideas.

In the U.S., by contrast, our reliance on vaccination may have the opposite effect: breeding feelings of invulnerability, and leading some people to ignore simple measures like better-than-normal hygiene, staying away from those who are sick, and staying home when they feel ill. Likewise, our encouragement of early treatment with antiviral drugs will likely lead many people to show up at the hospital at first sniffle. “There’s no worse place to go than the hospital during flu season,” says Majumdar. Those who don’t have the flu are more likely to catch it there, and those who do will spread it around, he says. “But we don’t tell people this.”

And relying on hand hygiene and social distancing may have the opposite effect: it will lead people to a sense that they can’t get the flu and so not get a protective vaccine and then acquire the disease and die. Again, its this all or nothing approach. As if we are not doing and recommending the simple measures. My hospitals are full of signs and pamphlets: wash your hands, wear a mask, stay home if ill. And the patients and visitors routinely ignore them. We want children to not be in the family birth center since they can be ill and not inform people of their symptoms. One visitor smuggled her child into the unit in a paper sack. Good luck with simple measures.

There is a rule in health care: no good deed goes unpunished. No matter what we do there will be unintended adverse consequences. Sometimes, and in medicine it is often, it is not what is the best answer, but the least bad answer.

All of which leaves open the question of what people should do when faced with a decision about whether to get themselves and their families vaccinated. There is little immediate danger from getting a seasonal flu shot, aside from a sore arm and mild flu-like symptoms. The safety of the swine flu vaccine remains to be seen. In the absence of better evidence, vaccines and antivirals must be viewed as only partial and uncertain defenses against the flu. And they may be mere talismans. By being afraid to do the proper studies now, we may be condemning ourselves to using treatments based on illusion and faith rather than sound science.

Always end with vaccine safety fear mongering. Trying to be an ethical researcher who is concerned about the health and well being of the people for whom we are responsible  is being afraid. I have noticed one thing in 25 years of practicing medicine: death lasts along time.

So what are the options?

You could read the entire literature on flu vaccine. I have, over the last 25 years, read most of it. It is compelling. Or rely on two researchers.

We could stop vaccinations. I think that would be wrong. We can do more studies and make better vaccines. Always a good idea. Eventually, if we can get the resources, perhaps definitive studies will be done as to the effect on mortality in elderly.

We could vaccinate everyone west of the Missisippi and no one to the east (I live in the west and have no loved ones in the east) and look at death rates.

What I think is needed is universal vaccination. We need to get everyone vaccinated to protect those that can’t get vaccinated or can’t respond to the vaccine. Of course, that’s impossible. It is a yearly shot and we lack the resources to accomplish this. What we will end up with is a refinement of the current approach. More emphasis on vaccinating the caregivers and family of the vulnerable, more emphasis on hygiene. Medical knowledge is always imperfect and always changing, one hopes for the better

And a year from now we will have some interesting data on morbidity and mortality from the current pandemic. It will be another brick in the wall of information, and it may be a good fit, or it may be an outlier. But we will not have any randomized placebo controlled trial since the mortality of this strain is greatest in children and pregnant women, who wants to volunteer their children or pregnant spouse to participate? Dr. Jefferson? The authors of the paper want to offer up their children? I think based on the data of prior influenza vaccines, and the fact that this is just another flu, I would be unwilling to risk sacrificing my children to appear cold hearted and unafraid.

Maybe those that read this will have a better understanding that medical knowledge and its application is imperfect and incremental.

I grew up knowing that the universe is big. “Space is big. You just won’t believe how vastly, hugely, mind-bogglingly big it is. I mean, you may think it’s a long way down the road to the chemist’s, but that’s just peanuts to space.”

Yet there was a time, just a century ago, where the universe was thought to consist of our galaxy. That was it. And erroneous conclusions followed from that narrow perspective. Even Einstein modified his equations with a cosmologic constant to account for the small universe. Bigger universe, bigger data sets, lead to different conclusions.

Vaccine efficacy is a huge universe. The Atlantic article, “Does the Vaccine Matter?“, is an island universe. True as far is it goes, but limited. I’m sticking with the whole universe.

Posted in: Science and the Media, Vaccines

Leave a Comment (60) ↓