Media Home Page | Contact Us
CDC Telebriefing Transcript
Increase in Influenza-Related Deaths in the United States
January 7, 2003
CDC MODERATOR: Thank you, Alan, welcome everyone, and thanks for joining us. Today, we are going to talk about an article in this week's Journal of American Medical Association. The article discusses an increase in influenza deaths in the United States. We'll start the call with a short narrative and then take questions.
Joining us online are Dr. Keiji Fukuda. His name is spelled K-e-i-j-i F-u-k-u-d-a, and Dr. Larry Anderson. His name is spelled L-a-r-r-y A-n-d-e-r-s-o-n.
Dr. Fukuda is a CDC expert on influenza. Dr. Anderson, an expert on respiratory syncytial virus. Both are authors of the article.
DR. FUKUDA: Hi. Thanks, Dave. Let me just try to put the study in context in a couple of minutes, and then go over what I think are some of the main points, and then we'll throw it open for questions.
I think that as a scientific study, one thing that you should take home is that for the first time, what we've been able to do is combine national viral surveillance data, and national mortality data to estimate deaths from influenza and RSV, or respiratory syncytial virus.
And the reason why we have done this, and why it's kind of an event, is that it allows us to come up with more precise information and also to come up with separate death estimates for different individual viruses.
Now when you look at the overall results, they indicate that, overall, influenza causes more deaths than RSV in all age groups, except very young children in which RSV causes more deaths. But the other large point is that both pathogens are major causes of death in the United States.
Now when we look at influenza, the deaths from influenza have increased from an average of 20,000 deaths per year to an average of about 36,000 deaths per year.
Now I want you to understand that these are average numbers. They're average numbers of deaths per year. When you look at individual seasons, the death totals from influenza can vary a lot, and they can range from really very few measurable deaths to, in bad years, up to 50,000 to 70,000 deaths, and so there's a wide range of deaths. The 36,000 is the average number but the range can go up as high as 50,000 to 70,000 deaths in bad years.
Now when you look at the impact of influenza, or you look at the epidemiology of influenza, this variability is one of the characteristics that we see for this disease.
Now the study also indicates that RSV is associated with about 11,000 deaths per year in the United States, on average, and in the study, we see that the deaths from RSV vary less than the deaths form influenza do over individual seasons.
Now, in terms of the public health importance of this study, what it really drives home is that influenza and RSV are both major causes of death in the United States. The deaths from influenza have increased in the 1990's compared with earlier decades and there are probably two major reasons for this. One is that the population in the United States is growing older and the number of people who are over 65 is increasing quite rapidly.
The second major factor is that during the 1990's, the most common influenza virus that circulated was the influenza AH3N2 virus. This is one of three main influenza viruses that circulate worldwide. Now these findings have important implications because, again, the population that's 65 and older, is really the fastest-growing segment of the U.S. population.
Now I think that some of the bottom-line points that we hope that you take home is that influenza and other respiratory viruses such as RSV really have been under-appreciated health problems, and a lot of people tend to think of them as causes of things like colds, but, really, I think these figures show you that they're really major causes of serious illnesses and death in the United States.
Now there are certain groups of people who are at increased risk of serious illness and death from influenza, and there are major steps that we can take to protect these people.
One major step is that we need to increase the use of influenza vaccine in those groups of people who are at risk of developing serious complications from influenza, and these are largely the elderly, and then people with a variety of chronic medical conditions, and we also need to increase vaccination coverage rates in people who take care of those vulnerable people, and these are especially health care workers and household members.
There have been some recent steps taken to facilitate the increased vaccine coverage that we need.
As an example, HHS recently approved the use of standing orders in facilities that take care of patients, Medicare patients, and this removes one important obstacle for that group of people getting vaccine more easily.
In addition, the Advisory Committee on Immunization Practices and CDC have recently broadened the influenza vaccine recommendations to include all people 50 years and older, and then recently, ACIP and CDC recently began encouraging the use of flu vaccine in children who are six to twenty-three months of age, and vaccination of that group of people who take care of those young children.
There are also some other important steps that I think need to be taken. One of the most important is that physicians and patients need to continue influenza vaccination beyond November.
Traditionally in the U.S., vaccine activities for influenza really tail down in November and there's no reason for this, and what we're trying to point out is that these vaccination activities really ought to continue as long as vaccine is available, and as long as people who are at high risk for serious complications remain unvaccinated.
I think the final point is that it's clear that we need more research. We need better research or more research in influenza vaccines that are more protective for elderly people.
We need to understand better why immune systems in elderly people are able to respond less well to flu vaccine than younger immune systems and why they don't fight off influenza infections as well as younger immune systems.
And then finally we need to have approved RSV vaccines that are going to be protective both in young children and in elderly people.
So I think that I'll stop right there and why don't we just throw it open for questions.
CDC MODERATOR: Alan, we're ready to take questions now.
AT&T MODERATOR: Thank you. Ladies and gentlemen, if you do have questions press the one on your touchtone phone. You'll hear a tone indicating you've been placed in queue, and may remove yourself from queue by pressing the pound key. If you're using a speaker-phone, we ask that you please pick up your headset before pressing any numbers.
Our first question will come from the line of Seth Bornstein with Knight-Ridder News Services. Go ahead.
QUESTION: Yes. Dr. Fukuda, thank you for doing this. I have a couple of questions and follow-ups on just the time periods we're talking about. You're saying through the '90s, the average has been 36,000 compared to previous decades. Can you give the specific time period you're talking about with the average of 36,000 deaths per year compared to the specific time periods where the average was 20,000.
And I'm wondering, second, is this a difference in--an actual increase in the deaths, or are we seeing a difference in the way it has been, the deaths have been attributed, the cause of deaths have been attributed?
DR. FUKUDA: Sure. Good question. If you look at the previous estimates of deaths from influenza in the United States, the most recent ones were covering the decades from the 1970's into the 1980's, and so that's where we have the estimates of an average of 20,000 deaths per year.
This study, and the 36,000 figure provides an update on those, and is covering, providing information from the 1990's.
QUESTION: So you're talking 1990 to 1990 inclusive, to--I mean, what was the end date of your study, the beginning date of your study?
DR. FUKUDA: The 36,000 figure is from the 1990-91 season up through the 1998-99 season.
QUESTION: And to look at that, you're saying that there are more deaths, this is not a difference in attribution for cause of deaths you say, and at the same time, this is when we saw more vaccine. So how do you say we're vaccinating more people and yet the deaths are going up?
DR. FUKUDA: Okay; good questions again. The increase from the 20,000 to the 36,000 is a real increase. You know, there are differences in methods that are used in this study and in previous studies, but we believe that this represents a true increase, and the basic reason for that, which is also much of the reason for--to answer your second question--is that the population is aging.
We know from a lot of different studies that influenza vaccine works pretty well and it works pretty well in preventing a variety of outcomes in such as hospitalizations and deaths and illnesses. But at the same time that we see increasing vaccination rates in the United States, we're also seeing this really very tremendous increase of older people in the United States, and as you grow older your immune system works less well.
You know, if you look at the aging of the population and you look at the demographics, the group of people who are 85 and above is, again, one of the fastest-growing segments in that elderly population, and this is a group of people in whom the vaccines are less protective, less immunogenic, and they're the people who are especially vulnerable to influenza and other respiratory viruses.
So we have two things going on at the same time. We have a large group of people who are growing more vulnerable and we also have better vaccination rates going on. So these are going on simultaneously but the sort of balance of all of this is that we have had an increase in deaths in the United States from influenza.
QUESTION: Thank you.
AT&T MODERATOR: And next we'll go to the line of Anita Manning with USA Today. Go ahead, please.
QUESTION: Hi; thanks very much. I have a couple of questions about RSV--but I don't know what the previous estimate was, and was it mainly thought to be a problem of young children and now we're learning that it's older people? And then I have a second follow-up question and I'll just give them both to you now.
Does the flu vaccine offer any protection against any other viruses such as RSV, or other respiratory viruses? And I'll put this on mute and listen. Thank you.
DR. ANDERSON: Well, RSV, traditionally, has been considered a disease of primarily the infant and young child but, actually, fairly early on, there was evidence of disease in elderly patients, or adults as well, and in the '90s, there have been a number of studies, many of those done at the University of Rochester, demonstrating, in prospective studies, that RSV is a substantial cause of disease in elderly populations.
Previous estimates--we have not done this kind of study previously with RSV, so we don't really have earlier estimates, and since we did not have the surveillance data in the way that we do now, we really can't go back and look at prior years. So we really cannot do the kind of thing that was done with influenza and compare the numbers or rates of deaths that we estimate are attributable to RSV from previous years to this time period in the study.
I mean, there have been other estimates that we, and other groups have done, looking at a possible death attributable to RSV, but done in a different way, not the kind of analysis that was done in this study.
In terms of does influenza vaccine provide protection from respiratory syncytial virus and other viruses, the specific answer is no, and then I'll defer to Keiji, if he wants to comment further.
But I think one of the things that may occur with influenza vaccines is kind of an augmentation of the fact that other diseases may have on individual patient, after infection with influenza, and that's a topic we really don't have good data on, but it's conceivable that influenza vaccine, probably not for RSV but for some other viral infections, may decrease the severity. But that's a guess. And Keiji may or may not want to comment on that topic.
DR. FUKUDA: Yeah, just to amplify on what Larry said. You know, influenza vaccines are specific for influenza viruses, and as far as we know it provides direct protection only against influenza viruses, although as Larry points out, there may be other benefits that we don't fully understand.
But another thing is that influenza viruses change continually each year, which is why the vaccine has to be given every year to cover these new viruses as they come up.
CDC MODERATOR: Thank you. Next question, please.
AT&T MODERATOR: That will come from the line of Joanne Silberner [ph] with National Public Radio. Go ahead.
QUESTION: Can you give us an update on how this year's vaccination is going. Are there shortages? How many people have gotten vaccinated? Do you have any data on that?
DR. FUKUDA: Joanne, we don't have data on how many people have received influenza vaccine for this year but we do have some other data. You know, in the past couple of years, on average, there has been about, oh, I don't know, 75 million doses, 70- to 75 million doses of influenza vaccine that have been distributed in the United States.
This year, there has been much more vaccine produced and distributed than any other year before, approximately 94 million doses of influenza vaccine. So we have a lot more vaccine which has been produced and distributed this year compared with earlier years.
What we don't know is how many people have actually received those vaccines right now.
QUESTION: Okay, and a follow-up is can you divide out how much of the increase in deaths is due to the aging of the population versus how much is due to the more virulent virus?
DR. FUKUDA: That's a good question. I don't think--we certainly haven't done that in this study and I don't think that right now I could separate out, you know, how important is one factor versus the other.
QUESTION: I mean, wouldn't that be pretty simple cause you know the ages of the people?
DR. FUKUDA: I'm not sure it would be so simple. It's something that we could try, partly because, you know, in different years we also have mixes of viruses. You know, we don't have--we sometimes have years in which we may basically isolate only one virus type or another virus type, but in other years we have different mixes of viruses out there, making it more difficult or complicated.
CDC MODERATOR: Thank you, Joanne. Next question, please, Alan.
AT&T MODERATOR: That'll be from the line of Marilyn Marchione [ph] with Milwaukee Journal. Go ahead, please.
QUESTION: Hi; good morning; thank you. I wonder if you could remind us, please, what is the vaccination rate among those 65 and older, and then I have a second question that relates to your data.
When you have these three categories, I'm wondering if each of these categories of death is sort of a subset of the other. In other words, when you have underlying pneumonia and influenza deaths, is that a subset of underlying respiratory and circulatory deaths? Because what I'm trying to do is establish how many RSV deaths are occurring in the elderly versus young.
DR. FUKUDA: Okay. The first one is easier to answer and I'll try to make the second answer clear. The vaccination rates in the elderly, in the last few years, have ranged somewhere between, oh, 65 to 67 percent. So that's in people who are 65 years and older.
Now in terms of your second question, the underlying pneumonia and influenza death is the narrowest, or smallest group of deaths in this paper, and it is a subset of the respiratory and circulatory deaths, which is a broader group of deaths, and then that is a subset of "all cause death," which is literally what it means, deaths from all causes. So they are subcategories of each other.
The reason why this paper has three categories--it may seem like an unnecessary complication--but the reason why it has it is that traditionally, in the past, for many decades, when people have done these analyses, they have used the pneumonia and influenza death category and they have also used the "all cause death" categories, and they've used them for different reasons, which I won't go into here.
That's the reason why we continue to do that in this paper, but we also developed this third category, the respiratory and circulatory death category because, for a variety of reasons, we feel it provides the most reasonable all-around ball park figure for deaths from these viruses.
QUESTION: And if I could ask a follow-up, then. I don't mean to dispute the data that the deaths are larger than have been previously anticipated or previously estimated. But before that 20- to 40,000 death estimate we always heard was flu and its related complications--pneumonia. Is what's new in here the circulatory death portion?
DR. FUKUDA: No, all of the deaths--or I should not say all, but most of the deaths from influenza and viruses like RSV are from complications of these viruses, typically. There are some people who develop a primary infection and then go on to have fulminant disease directly from the virus and to die from the virus.
But more typically what happens is that someone gets infected with one of these viruses and then it precipitates complications. So it leads to, for example, bacterial pneumonia, or if someone has a preexisting heart condition, it'll lead to worsening of that heart condition, and it's these complications that result in death.
So these viruses sort of set off a train of events which leads to the person's death. The increase in the 20,000 to the 36,000 estimate is not really a reflection of this new category.
It's a real reflection that we're just seeing more deaths from these viruses.
QUESTION: Thanks very much.
CDC MODERATOR: Thank you, Marilyn. Next question, please, Alan.
AT&T MODERATOR: That will be from the line of Brian Bectel [ph] with Infectious Disease News. Go ahead, please.
QUESTION: Hi; thank you. Dr. Fukuda, and Dr. Anderson, if you could chime in as well, I was wondering if you could just break down the percentage of deaths in age groups, specifically, in young and old.
DR. FUKUDA: Yeah, if you, just as a rough breakdown, if you look at the influenza deaths, if you look at that 36,000, we estimate that 90 percent of those take place in people who are 65 and above, and then the remaining 10 percent or so take place in younger age groups.
If you look at the RSV deaths, I believe that we estimate about 78 percent of those deaths take place in people 65 and above, and the remaining take place in younger age groups.
QUESTION: And how much of this rise that you're reporting is better surveillance and how much is a real rise in deaths?
DR. FUKUDA: Again, I think that most of that increase, we believe, is a true increase in deaths. If anything, we've been fairly conservative in how we have come up with the 36,000 deaths. The previous estimated deaths were based on the so-called all cause death, and the estimate of 36,000 is a narrower category but it's still a much higher number.
So it's a conservative estimate, and, again, we think that it reflects a real increase and not sort of a methodologic change.
QUESTION: And a real quick follow-up. Do you know how this year is shaping up in terms of deaths? Or is it too early to tell?
DR. FUKUDA: It's too early to tell. These analyses are based on both viral data but also all of the deaths which are reported in the United States and typically it takes a year and a half to two years for all of the deaths in the United States to be made available for analysis.
QUESTION: Thank you so much.
CDC MODERATOR: Thank you, Brian. Next question, please, Alan.
AT&T MODERATOR: Yes, sir. That's come from the line of Ann Cairns [ph] with The Wall Street Journal. Go ahead.
QUESTION: Hi; good morning. I have a couple of questions, if I amy. I just wanted to clarify. One of the reasons you gave for the increase in the flu deaths in the '90s was that the Influenza A strain was the most prevalent. I just want to clarify that that is an especially nasty form of the flu. Is that why?
DR. FUKUDA: Well, if you--you know, there are three main types of influenza viruses which have been circulating since 1977. Two of these are Influenza A viruses, Influenza AH3N2 is one of them, Influenza AH1N1 is the second one, and then influenza B viruses. So these are the three main viruses which have been circulating, and, you know, each year they change a little bit and so there are different strains. But just keep in mind, there are three main types.
DR. FUKUDA: And then if you look at those three main types and you look at the kinds of impact that they have had, the influenza AH3N2 viruses have typically been associated with higher levels of death, higher rates of death and serious illnesses than the other two influenza viruses, and that's also the one which has been the most common in this decade.
QUESTION: And is there any understanding as to why they've been the most common in this decade?
DR. FUKUDA: No. You know, the percentages and the types of viruses that circulate are just something that we monitor but we don't really know why one predominates over another in any given year.
QUESTION: Okay; great. And just one other question. What is the latest you would recommend that people in high-risk categories get the vaccine, if they should be getting it past November, but how late--
DR. FUKUDA: Well, you know, I think that the basic concept is that, you know, it's best, it's certainly best to get vaccinated before the influenza season has really kicked in, which is why we typically recommend that October and November is the best time for people to get vaccinated, especially high-risk people to get vaccinated.
But, you know, if you look at this year, for example, influenza activity is still pretty low in most parts of the country and there are people who have high-risk conditions who have not been vaccinated. So if there's vaccine in your area, and especially if you're a person with a high-risk condition, or, again, one of those people who take care of them, a health care worker, a family member, you know, there's every reason to get vaccinated and no reason not to get vaccinated.
It's a little bit hard to say that, you know, there's a definite end date by which you shouldn't get vaccinated but that's the basic concept.
QUESTION: Okay; thanks very much.
CDC MODERATOR: Thank you, Ann. Next question, please, Alan.
AT&T MODERATOR: Yes. We'll go to the Associated Press. Lindsey Tanner, go ahead
MS. TANNER: What strain are people being vaccinated against of flu this season?
Secondly, do you have overall vaccination rates for recent years for all ages?
And thirdly, when was the flu vaccine first recommended for people age 65, and then for those over 50?
DR. FUKUDA: Let me see if I can get all these right, but the influenza vaccine, which is an inactivated flu vaccine, contains-- always contains 3 viruses, or has contained 3 viruses interested past several years, so it's a trivalent vaccine, and it contains viruses similar to the influenza-H3N2, H1N1 and B viruses, which were in circulation. And so this year's vaccine is no exception, and the viruses which are contained in the vaccine this year are similar to the viruses which are circulating worldwide right now, and so the coverage is pretty good.
In terms of your second question or I think your third question, the recommendations to vaccinate elderly people have been in effect since the U.S. first started making flu vaccine recommendations, which was in the early 1960s. So since the early 1960s the elderly have been recognized as one of the groups which are specifically vulnerable to influenza, and they have been specifically targeted for vaccination.
The group 50 years and above, that recommendation was first made, I believe, in the year 2000, so that's a more recent recommendation.
Lindsey, you asked a second question which I forget. Can you repeat it again?
QUESTION: Yes. I'm wondering if you have overall flu vaccination rate estimates for all ages combined. Earlier you gave them for older people.
DR. FUKUDA: Well, in people--again, in people who are older than 65 and older, the rates have been about 65 to 67 percent in the last few years. In people who are younger than 65, if you look at those groups that are younger than 65 and have a high-risk condition, you know, for example, someone who has diabetes or heart disease, or children with asthma, the rates are much lower, ad on average the vaccination rates in those groups have been about 30 percent or so in the last several years. I think we have additional data on those, but I would have to go back and look at them to give you figures for other groups, but those are the ballpark figures.
CDC MODERATOR: Allen, next question, please.
AT&T OPERATOR: Yes, sir. That will be from John Lowerman with Bloomberg News. Your line is open.
QUESTION: You just answered my question. Thank you.
CDC MODERATOR: Thank you, John.
AT&T OPERATOR: And we'll now go to Steven Smith with the Boston Globe. Go ahead, please.
QUESTION: Hi, good morning. A couple of questions. I'm wondering, once you adjusted for the increase in the 65 and older cohort that you referenced, whether there was a true increase in the mortality rate?
I'm also wondering whether you look ed at morbidity issues?
And thirdly, if you could give a snapshot of what you're seeing this year vis-a-vis influenza? I know that for the third consecutive year in the Northeast it's shaping up as an especially mild season. The Massachusetts Department of Public Health reports only 2 isolates so far, but I'm wondering what you're seeing nationally.
DR. FUKUDA: When we look at the rate in the study, we see that the rates of death have not increased in the elderly, and so basically again, underscoring the point that it's the increase in elderly people that's going on, so it's not that these viruses are becoming more virulent than in the past.
In terms of morbidity, that's a really good question, and that's one of the things that we want to look at. We haven't--you know, we concentrated on death in this analysis, but we will be looking at morbidity to see whether those rates have also increased, such as hospitalizations.
In terms of this year's activity, again, I think it appears to be fairly early in the season. Again, for those of you that follow influenza, I think that you know that when the season begins, when it peaks, can really vary a lot between years. The most common time frame for influenza to peak is in February, but there are years in which we have seen a peak in March, and there are also years in which we've seen a peak in December. And so right now activity levels in general have been pretty low in the country, but when I look at the curves, it appears to me that we can clearly expect activity to pick up later on, you know, perhaps in a few weeks, perhaps in several weeks.
QUESTION: Do you attribute this relatively low level that you're tracking this year and the low levels the past 2 years to the strains in circulation, to the efficacy of the vaccine that's been developed? Have you been able to parse that out?
DR. FUKUDA: I would like to think that it was due to the influenza vaccine activities, but I don't think that's true. I don't think our vaccine coverage rates are high enough either in the United States or anywhere else in the world really to affect the epidemiology or the spread of these viruses. It's just one of the characteristics. You know, flu has been studied for decades or even over 100 years, and when you look at influenza over this long time span, it's just one of the aspects of influenza that in some years you have lower levels of activity and probably lower numbers of infections, and in other years you have higher numbers of infections and more illnesses and deaths.
QUESTION: One final question. When you look at vaccination levels, is there a rate at which herd immunity would be achieved if you, say, vaccinated 80 percent of the population? Does anyone know if herd immunity can be reached and what sort of level would have to be attained?
DR. FUKUDA: I think that 80 percent is a figure which is often mentioned, but there are some things that you ought to understand. For example, herd immunity within an institution like a nursing home would be possibly a different phenomenon and a just different threshold than herd immunity in a community or in a country. So I think that for institutions, you know, people will frequently say that 80 percent is the figure at which herd immunity may develop, and that might well be true. I think that for larger populations, certainly for a big population like the United States or larger communities and cities, I think that that's less clear what sort of levels we really need to attain the achieve herd immunity.
CDC MODERATOR: Thank you, Steven.
Next question, please, Allen.
AT&T OPERATOR: Yes, sir. That will come from the line of Gina Culotta from New York Times. Go ahead, please.
QUESTION: Hi. I have two questions. One of those, I was looking at the JAMA paper and I guess Table 2, where you have Estimated Annual Influenza Associated Deaths From 1976-77 to the '98-99 seasons, is the table that gives the deaths. But I look at that table and I cannot figure out where the 20,000 and the newer figure of 36,000 are in that table. Where do I look at, underlying pneumonia, underlying respiratory and circulatory, or all causes or what?
DR. FUKUDA: Gina, I think--let me explain Table 2 and Table 3. I think that the figures that we're talking about come from Table 3. And let me explain the difference between Table 2 and Table 3. Table 2 reflects the analysis of sort of an intermediate model that we used in this study, and this is a model which looks only at influenza, but it--because it's only looking at influenza, we're able to look at a longer time span. So we're looking at data from 1976 through 1999. And so that's one model.
What we then did was go on and develop a fuller model, somewhat more complex, but a model which was also able to take RSV into account, and that's--those results are available in Table 3. And because we had national viral surveillance for RSV from 1990 onwards, the analysis is from 1990 through 1999, and again, this is sort of the more full model. And so the 36,000 figure comes from that second row of numbers under the underlying respiratory and circulatory deaths, and if you look at the bottom of the fifth column, you'll see that 36,155. Do you see that?
QUESTION: I have another question then. Because then if you would compare the 20,000 that was from previous years, that wa with a different model, so how could you be sure that when you're changing models--and you said before that it was not a matter of changing models, but how would you know when you actually changed the way you did it and you got--the 20,000 I guess was the--I see a 25,000 mean for these other years, but I don't know--
DR. FUKUDA: No. The 20,000 does not come out of this paper. The 20,000 comes out of studies that were done actually also be CDC previously.
QUESTION: Well, is it possible that maybe it was high all along and you just use a different model? That's what I was wondering, like you know, when you change the way you do the calculations in the '90s, how do you know what you're comparing?
DR. FUKUDA: Okay. If we were to look at the 20,000 figure, that would be 20,000 all-cause deaths. Those were the numbers that were previously used. If we look at the all-cause deaths in the newer model, the figure comes out to about 51,000. So that's--
QUESTION: So now I'm really--could you explain then how you make a comparison? Because it gets more and more confusing to me. Because first we were comparing 20,000 and 36,000, but if you are comparing all-cause deaths, then it's 20,000 versus 51,000, but then you've changed models.
DR. FUKUDA: Okay. This is going to get a little bit complicated, but let me try to explain it. In the past, the models that were used to come up with estimates of flu-related deaths used all-cause death estimates, and that's because that approach had been used for decades previously, and those models were somewhat simpler than the models that we used in this study.
In this study we changed the model, and the reason why we did that, one of the reasons that we did that was that we now have a lot of virus surveillance data available, and so we developed a model which allows us to use that virus data, and we did this--and again, we did this because it allows us to come up with more precise estimates and it allows us to come up with estimates of death related to specific viruses. And there are a number of other methodologic issues why we moved, but I think that's probably the most important one to know.
And then again in the past, there have generally been two categories of deaths which have been analyzed, the so-called pneumonia and influenza death, and then the all-cause death. Now, in this analysis we created a third category of death, these respiratory and circulatory deaths, and the reasons we did that was that for certain reasons, we were not satisfied that the all-cause death approach was the best way to estimate death from these viruses. And you know, for example, all-cause deaths includes deaths from accidents, deaths from fires and so on, and those kinds of deaths can increase in the wintertime completely independent of influenza circulation or RSV circulation.
So there are a number of reasons for why we wanted to move to a different category of deaths, which we did, and it's--that means that the respiratory and circulatory death category is a more conservative way of estimating deaths from these viruses.
QUESTION: I still have the same question though. It sounds like--I don't want to be a total pest about this, but it's still confusing to me how you can compare numbers using two different models, and say that the number of deaths has increased in the '90s, when in the model you used previously was so different.
DR. FUKUDA: Gina, again, if we were to use the older models, which we have done in unpublished analyses, the numbers are much higher also. But when we use this current model--you know, we use this current model because it gives us more precise and more specific estimates, but those--when we use the same model, we also come up with much higher estimates, but we do want to move on to this new model because we think it's a better way to do it.
QUESTION: One last really easy question this time. Is there any reason why people who are not in risk groups, should they have vaccines too if there's enough available?
DR. FUKUDA: Yes, yes. I think that what we have--what we try to do is that--you know, we're especially concerned about that group of people that are going to develop serious complications, but when you look at the recommendations for flu vaccine that are given out each year, you know, we certainly mention that--we certainly recommend that anyone who wants to avoid developing influenza should get vaccinated against the flu, and then some of the priority groups that we recommend vaccine for, such as health care workers and other household members, are not people who necessarily have high-risk conditions, and so, yes, definitely flu vaccine is also recommended for other people who do not have high-risk conditions.
DR. FUKUDA: Thank you.
CDC MODERATOR: Thank you, Gina.
Allen, is there any more questions?
AT&T OPERATOR: Yes, sir. We have a question being queued from Ted Vigowski with Atlanta Public Broadcasting. Go ahead, please.
QUESTION: Dr. Anderson, you might be the best one to address this very quick question. Of all the years I've covered RSV, I'm always curious as to why we haven't been able to nail down a vaccine for this very [inaudible] virus. Is it a constellation of viruses like flu that change so much that make it hard to do that? Or if you could tell me anecdotally what you think the problem is about coming with a vaccine for RSV. Thank you.
DR. ANDERSON: If I knew the answer, I would be in good shape. But I think there are a couple underlying themes just from the epidemiology of RSV that suggest that we may have trouble with developing a vaccine, and one is that you get repeated infections throughout life, and that with these repeated infections, as illustrated in the data presented in this paper, you can have substantial serious disease with complications. So that tells us that even natural infection provides limited protective immunity. It provides some, but limited.
And it's not just as in the case of influenza, that there's a shift in change of the antigenic structure of the virus such that the virus has kind of work to evade previously-induced immunity. And we're not really sure exactly why we have this problem, but I suspect it has to do with how RSV causes disease, and we suspect, but do not know, that part of that is based on the virus's ability to kind of induce inflammatory response, a host response that actually contributes to the disease process itself. Certainly a lot of groups have been working on developing an RSV vaccine in terms of trying to understand what causes disease and what might be a protective immune response that we could then try to induce with a vaccine, and trials with both live virus vaccines, primarily in young children, and then subunit or non-live virus vaccines in the elderly.
So there's been a lot of work, and unfortunately, it hasn't been successful. And if we knew why, we would hopefully be able to then work to make a safer vaccine, but it's a very good question.
CDC MODERATOR: Thank you. Do we have another question, Allen?
AT&T OPERATOR: Yes, sir, from the line of Maggie Fox with Reuters. Go ahead.
QUESTION: Good morning. I joined a little late, so I'm sorry if this repeats an earlier question. I'm wondering if you have recently added to your recommendations that young children, I believe, under 2 get the flu vaccine, and I'm wondering if you can speak to the percentage of fatalities or serious morbidity from this group, and if that number has changed?
DR. FUKUDA: Sure. Thanks, Maggie. For the first time what the Advisory Committee on Immunization Practices and CDC have said is that it encourages, which is kind of softer language than "recommends." It encourages influenza vaccine for children 6 months to 23 months of age, and then the people who take care of them, so basically health care workers and household members who take care of them. And that recommendation was made because there are now a number of studies which indicate that this group of children is at risk for hospitalizations and serious complications from influenza than older children who are healthy. And so simply by being that young, there's an increased risk of getting hospitalized from influenza-related complications.
When you look at this group it's clear that the rates of hospitalizations are elevated. The rates of death are minimally elevated in that group compared with healthy children, so really the rates, the increases in rates of deaths are not anywhere what you see compared to the increases in the rates of deaths in the elderly people. So that recommendation is really based more on concern that these children are being hospitalized in developing those complications, but it's not driven so much by an increase in deaths in that group.
QUESTION: Can I ask another question, please?
CDC MODERATOR: Go ahead, Maggie.
QUESTION: What we used to say in the past was that influenza caused anywhere between 20,000 and 40,000 deaths a year. We used to refer to 40,000 in a bad year. Is the bad year now revised upward or are we just more in bad years lately?
DR. FUKUDA: Yeah. In the past that 20,000 to 40,000 was always wrong. It was really--it should have been really zero to 40,000, with the average being 20,000. So what we're saying now is that the average is closer to 36,000 and that the upper range is closer to 50 to 70,000 in severe years, but the lower range--again there will be years in which activity will be very mild and the number of measurable deaths would be much closer to zero, so there's the big range and then there's the average.
QUESTION: Thank you.
CDC MODERATOR: Thank you, Maggie. Next question, please, Allen.
AT&T OPERATOR: Yes, sir. We have a follow-up question from the line of Seth Lowenstein with Knight-Ridder News Service. Go ahead.
QUESTION: Again, thanks again for doing this. Looking at the Center for Health Statistics data, it always lists pneumonia and influenza deaths at around 65,000 a year, in that ballpark, especially the last--at the last--2000. Does this first with what you see in the National Center for Health Statistics?
And then another follow-up question. Just, is it fair to say the last 3 years have been dramatically smaller than the 36,000?
DR. FUKUDA: Let me take the NCHS question first, the National Center for Health Statistics.
If you look at people who die with pneumonia and influenza, or one of these other categories, in the United States, there are a lot of people who die from these diseases throughout the year, and most of those deaths are not related to influenza, and so what we're saying is that there is a percentage of those people who die, with those diagnoses, who are related to influenza, and so yes, our figures--actually, our analyses are based upon the data provided by the National Center for Health Statistics but what we do is go ahead and estimate that fraction or that percentage of deaths which can be attributed to influenza or RSV and that's something that the NCHS statistics do not do.
QUESTION: Oh, good. Thank you.
DR. FUKUDA: And then you had one other question. What was that?
QUESTION: Yes. Is it fair to say that the last three seasons, or the last two seasons and then this partial--how much below this average of 36,000--I mean, we know that they've been relatively mild years.
Is there any ball park estimate you can give us for the last two seasons, and maybe this season? Obviously, it's too early for this one.
DR. FUKUDA: Yes. You know, we monitor, we also monitor deaths from something called the 122 cities mortality reporting system. These are deaths which are reported by 122 cities vital registrar's offices, and they capture about one-third of the deaths which occur in the United States.
But this is sort of a rough way for us to get a feel for what kind of impact influenza is having on mortality in the country.
So if we go back to the--let me see, I think the 2000-2001 season, I believe that in that system there were no measurable increases in deaths in that year, whereas last year, if we look at the results in that system, deaths were increased for a few weeks in the United States but, again, I don't have exact figures, and the data that we're putting out now is based on the entire death data set, the National Centers for Health Statistics data set, and again, there's always a couple of years time lag before we can analyze those data.
So I can't give you precise figures for the last couple of years.
QUESTION: Well, I know not precise, but if we're looking at--it's fair to say, then, from what you see, the 2000-2001 season may have been close to the zero range--
DR. FUKUDA: Yes--
QUESTION: --that we were talking about. And the 2001-2002, obviously not close to zero but is it fair to say much lower than the 36,000 range?
DR. FUKUDA: You know, I would say that there were deaths, and again, I really don't know where the numbers are going to fall out for that year, and so I don't know whether you can say it's going to be much below 36,000. It may turn out to be close to 36,000. It may be an average year. It may be lower than an average year. I really can't tell you right now.
CDC MODERATOR: Thank you, Seth. Alan, we have time for one more question.
AT&T MODERATOR: Yes, sir, and that'll come from, a follow-up question from Anita Manning with USA Today.
QUESTION: Hi; thanks. And this will be a nice one to go out, because I'm going to ask, I wanted to ask where we are in development of better flu vaccines, and what the public health message is to people how are worried about RSV, how they can protect themselves.
DR. FUKUDA: Anita, in terms of influenza vaccine, I think that there is a lot of work going on in a couple of different areas. One is, you know, which is most immediately relevant, is just how to improve vaccination coverage rates.
You know, again, I think there's a lot of thinking going on about how to remove obstacle and how to improve vaccination coverage rates in groups of people, and there are some of those things which I mentioned at the beginning of the call, that are being done and that should be done.
And then in terms of influenza vaccines, I think that, you know, that there has been a lot of work done on developing different kinds of flu vaccines. Recently, there has been movement to get approval for live attenuated influenza vaccines, and certainly there's research going on to improve inactivated vaccines.
I think, and I can't tell you more than that, where we stand, but there are, you know, certainly groups working on improving vaccines, and the third sort of component there is to understand why the immune systems in older people don't respond as well, and, again, there is a lot of work going on in that area, but it's a very complex problem, trying to figure out just why the immune system doesn't respond as well as younger immune systems do.
And then for the RSV, I'll turn that over to Larry.
DR. ANDERSON: Well, without a vaccine or an anti-viral, we don't have a specific way to prevent either the infection or the disease associated with the infection.
There are, however, some things that, in general, decrease the risk of acquiring a variety of infections, and that's some of the infection control procedures, and that's actually particularly important in hospitals in preventing nosocomial transmission, and there of course the risk is often higher because you have patients that are ill with a number of diseases, and that's actually an important feature of controlling RSV in at-risk populations.
And to some extent you can probably decrease the risk of acquiring any viral infections or many of the viral infections by handwashing, not sharing food and water, the kind of classic things that decrease the chance of acquiring infection from someone else. But we, unfortunately, without a vaccine, do not have a good prevention strategy for RSV, particularly in adults. In young children there is the RSV immunoglobulin, the infant--in young child, but that of course is not licensed or applicable to the adult population.
CDC MODERATOR: Thank you, Anita, and thanks to everybody for joining us. Alan, thanks for your help.
AT&T MODERATOR: Thank you. Ladies and gentlemen, that concludes your conference call for today.
Thank you for your participation and for using AT&T's executive teleconference service. You may now disconnect.
This page last updated January 7, 2003
Department of Health and Human Services