When I tell people I am an epidemiologist, most of them think it means I'm a skin doctor. I'm not (although the skin disease specialty is much more lucrative). Instead I study patterns of disease in populations and use what I see to try to figure out why the observed pattern rather than another. Since I'm a cancer epidemiologist I usually do large, highly systematic studies that often take years to execute and analyze, but some epidemiologists do much more immediate "shoe leather" epidemiology, investigating disease outbreaks. They are like disease detectives and we can often learn a surprising amount about basic science that way. An investigation just published in Emerging Infectious Diseases of a swine flu outbreak involving a tour group in China is an excellent example. Using information from this small outbreak (11 cases) the outbreak team was able to shed light on an important question of factors that increase the risk of person to person spread of the pandemic strain of swine flu H1N1.
Here's the set-up:
The index case-patient was a 40-year-old female US citizen who had traveled from the United States to Jiuzhaigou, a popular tourist spot in southwestern China; she stopped to change planes in Hong Kong and Chengdu. She noticed her first symptom, chills, on June 2, immediately before arriving in Chengdu, â23 hours after departure from the United States. After learning that she had traveled on 3 flights and had toured with a group, we obtained the manifests of all flights that she had traveled on and telephoned and asked all passengers of the 3 flights and all members of the tour group whether they had had any symptoms from May 27 through June 12, 2009. We also obtained detailed information on the activities of the tour group during the 3-day tour. Health authorities placed members of the tour group under medical observation and isolated those who had clinical signs or symptoms or positive throat swab culture results. (Han et al., Lack of airborne transmission during outbreak of pandemic (H1N1) 2009 among tour group members, China, June 2009. Emerg Infect Dis. 2009 Oct; [Epub ahead of print])
After completely blowing the SARS epidemic, at the outset of the swine flu panic China turned 180 degrees on swine flu and adopted a Draconian effort to contain the uncontainable, including fever screening and health questionnaires at international ports of entry, isolation of infected travelers, and involuntary quarantine of close contacts of infected persons. There's little evidence any of this does any good to contain influenza, but in this case the forced quarantine of the tour group provided an opportunity to do an outbreak investigation.
The actual travel itineraries and mixing with non-tour groups introduced some additional complications. As I reconstruct it from the paper, this is what happened. The 40 year old woman and her family left the US (NY) at midnight June 2 and flew all night, landing in Hong Kong at 2 pm local time. They then transferred to a Boeing 757 (presumably one class seating, two and two) at about 7:30 pm for the short flight to Chengdu (arrived 10 pm). About an hour before landing on June 2 she started to have chills, but the next morning, with her family members, joined the tour group and they all took another plane to Jiuzhaigou (altitude 2930 meters). By that time she was febrile and coughing but continued on the tour in Jiuzhaigou, joined by 7 additional members. For three days this group of 24 original from the flight plus 7 more not on the flight who joined them in Jiuzhaigou traveled in a tour bus with 70% recirculated air and 30% outside air, stayed at the same hotel, ate together, talked to each other and helped each other take pictures. At one point the index case and perhaps others shared chewing gum sticks from a pack with other members of the tour. After the 3 day tour, the 24 who flew to Jiuzhaigou together (but not the 7 who joined the tour there) flew back to Chengdu on another Boeing 757 along with 87 unrelated passengers. It was when they returned to Chengdu that the diagnosis of the index case was made and the whole tour group of 31 people (24 plus 7) was placed under "involuntary medical observation." Non-tour passengers on the two flights (to and from Jiuzhaigou from Chengdu) were contacted but none reported an influenza-like illness.
So what happened? There were three case definitions. A suspected case was someone who had one or more of 5 symptoms: fever (>38°C), cough, sore throat, chills, or headache—in a passenger of flight CZ6659 (June 3, Chengdu to Jiuzhaigou) or flight CZ6660 (June 5, Jiuzhaigou to Chengdu) or in a member of the tour group. A confirmed case was a suspected case with evidence of the virus in a swab using PCR. A secondary case was a confirmed case (other than the index, or initial, case) with symptom onset at least 24 hours after that of the index case. Of the 30 tour group members excluding the index case, the attack rate was 30% (9 secondary cases). In addition, there was a secondary case among the plane passengers on the way back to Chengdu. So why these 10 people and not the other 20? We're back to epidemiology again.
We have three plane trips and 3 days of bus/hotel tour at Jiuzhaigou. All planes (Boeing 757s) had the usual high efficiency particle filters for half the cabin air and the other half from outside. Recirculation occurred every 3 minutes, but air movement is not along the length of the plane but transversally, from side to side. We've discussed the airplane contagion risk a couple of times, here and here. The members also spent about 7 hours together in a closed tour bus. In addition, they ate together, talked with each other and did other things tour groups do socially (include share inanimate objects like chewing gum and cameras). But the pattern revealed by the epidemiologic investigation seemed to imply that the primary mode of transmission was droplet (large virus-laden particles that fall out of the air within a couple of meters). The index case was coughing and talking to others which generate droplet-sized particles. But how do we know that even smaller particles that would remain suspended in the air for hours or even days were not responsible? The could also be generated in the same way. It is a matter of some contention and importance because it bears on how risky sitting in an emergency room or traveling on a plane might be, even if you aren't near someone who is sick.
In this case, with one exception, no one but those in the tour group developed secondary cases, and the one exception was one that tended to prove the rule. None of the 91 passengers on the Hong Kong to Chengdu flight (during the last hour of which the index case started developing chills and fever) reported any influenza-like illness (ILI). Since it is thought that people may be infectious for as long as 24 hours prior to showing symptoms, this is relevant information. From Chengdu onward, 27 members of the tour group was together but 7 more were only in Jiuzhaigou with the tour and not on the planes from and to Chengdu. The authors reported that the secondary attack rate wasn't different in these two groups, suggesting at the least, that some transmission occurred during the bus tour/hotel Jiuzhaigou ground phase. The one exception to the rule that only tour members became ill was a passenger on the return trip from Jiuzhaigou to Chengdu (one of the 87 non-tour passengers). But it turns out that this non-tour individual was sitting two rows (in seat 9A) from the index case (sitting in seat 7A) and also a symptomatic secondary case (sitting in seat 7B). Note that the paper seems to have inadvertently reversed the city names for the return flight.
The index case coughed continually from shortly after onset through to diagnosis. The attack rate was greater for women than men and those under 40 compared to those over 40. This may well be a reflection of social relations:
When we evaluated the contact patterns of the tour group with the index case-patient, we found that for the 16 tourists who had talked with the index case-patient from close range (<2 m) for >2 minutes, the attack rate was 56%, whereas none of the 14 tourists who did not talk with her became ill. Members of the tour group who had talked with the index case-patient for >10 minutes were almost 5Ã as likely to become ill than those who had talked with her for 2-9 minutes. The 14 passengers who had not talked with the index case-patient did report other interactions with her, such as dining at the same table, sitting within 2 rows on the same flight or bus ride, and receiving chewing gum from her. Moreover, 3 of these 14 uninfected passengers had sat within 2 seats of the index case-patient during the bus rides but had never talked with her from close range. [internal references omitted]
In this small investigation, neither accepting chewing gum from a person (an inanimate object) nor eating at the same table nor traveling on a closed bus for 7 hours were related to getting infected. The one thing that stands out is that the people who got sick from the index case were the ones that spent more than a minimal amount of time talking to her at close range. The case unrelated to the tour was within two rows in an airplane from two symptomatic cases. All this is most consistent with droplet transmission.
There are a couple of other findings in this investigation that bear mention. One is the spectrum of symptoms among lab confirmed cases:
During this outbreak, we identified a total of 11 confirmed cases of influenza A pandemic (H1N1) 2009 infection (Figure 1). Average patient age was 36 (range 18-59) years; 2 patients were men and 9 were women. Signs and symptoms were cough (73%), fever (64%), sore throat (64%), headache (27%), chills (27%), runny nose (18%), and myalgia (18%). All 11 case-patients fully recovered; 3 (including the index case-patient) recovered on June 13, 5 on June 15, 1 on June 17, and 2 on June 18. The mean duration of illness was 11 (range 9-14) days. [internal cites omitted]}
We tend to think of a typical case of flu as having respiratory symptoms (cough, sore throat), fever, headache and muscle aches and pains. Note that 27% hand no cough, 36% had no fever, 73% didn't have headache, and 82% didn't have muscle aches. All of them had one or more of these and some others, but no one symptom was a sure sign of flu.
A second thing of note was the complete failure of the rigorous containment procedures the Chinese authorities were using:
During this outbreak, the index case-patient was febrile while traveling on 3 flights. A secondary case-patient was also febrile while traveling on the return flight (Jiuzhaigou to Chengdu). Neither patient’s illness was detected by thermal scanning at the airports. Another secondary case-patient had had a headache during the return flight. All 3 symptomatic case-patients filled out health questionnaires but did not truthfully inform health authorities of their symptoms. The other 9 case-patients began having symptoms after returning home; hence, they were also not detected by airport screening. These data suggest that thermal scanning and health questionnaires at the airports were not effective for detecting pandemic (H1N1) 2009 infections.
One thing epidemiologists are trained to think about is the possibility of systematic error, which we call bias. In this case, could it be that the high proportion of the sick who reported talking to the index case was something epidemiologist call recall bias, the tendency to remember things better if it has a special salience. Might it be that if you got the flu you were much more likely to remember talking to the index case? With a small number of cases, this could make a difference. This was addressed by the authors with this observation:
The main limitation of our investigation was the possibility of recall bias; i.e., those who became ill might have more accurately recalled their contact history than those who did not. However, the index case-patient had a highly distinctive hairstyle, which made her easy to remember. Also, interviews about the tourists’ exposure to the index case-patient were conducted within 1 week of the completion of their tour. These 2 factors should have helped minimize any potential recall bias.
This is a small study with very special circumstances. But it shows two things. One is that astute observation and collecting the right data can provide potentially useful information. The second is the information this study provided: that there is a strong suggestion that influenza was passed from the index case to 10 secondary cases by generating virus-laden droplets while talking to her fellow tourists at normal conversational distances.
- Log in to post comments
Which I always give credit where credit is due. The Administration was backtracking like hell when Joe Biden pretty accurately described the situation a few months ago about the transmission of flu on aircraft. Had we closed the borders we would have slowed it, stopped the flights, slowed it, done just about everything recommended in the IHR's and we would have slowed it.
I wonder how many people would have had a better chance if the rule book had been followed?
May I round up the "36% had no fever" to 4 people in 10? The "feverless flu" concept is one that just doesn't seem to be sinking in with front-line health care providers here in the southeast.
Three weeks ago, I moved my 17 year old son into a college dormitory with 85 other 16-19 year olds. Many of them are sick now, but very few have been running fevers, so they're not considered or counted as sick.
I really think this is having broad ramifications for the speed with which this virus is currently spreading on our college campuses. The 4 in ten kids who are sick with the flu, but, due to lack of a fever are being reassured that they do not have the flu, are walking around and spreading this at rates much faster than I think would have to be the case.
The message that should be getting out on our college campuses right now is that it is common for mild cases of the swine flu to present with no fever. We are currently experiencing an outbreak of the flu, so, if a student is experiencing a sore throat, headache, fatigue, body aches--but no fever--he or she is probably contagious at the onset of such symptoms.
My son is one such case. He has been told by a doctor twice in one week that his illness is not flu-related because he would have had a fever along with his other symptoms. As his mom, I am grateful for such mild symptoms. But I can't help but worry about the pregnant guidance counselor that he saw last week, his asthmatic roommate, or the heavy-set cashier at the cafeteria.
We've made some mistakes here in the southeast over past few weeks--I hope other parts of the country can learn from those mistakes.
so, using a language with many "p"s and "t"s,
may put you at risk
make a mouth-mask for sick speaking, the nose can be free
mandatory when you give talks or teach.
Sick teachers giving lessons - is that how it
exploded at St.Frances ?
Weird question #1: If influenza is typically regarded as most contagious during the febrile stage -- but there isn't a febrile stage -- would we then expect the patient to be less likely to transmit the virus?
Weird question #2 (expanding on anon's post): Would we expect to see lower rates of transmission at schools for the deaf?
melbren: you have the infamous problem that it is hard to tell the difference between flu and other upper repository conditions without testing.
In fact, I've got about half the flu symptoms right now myself. I think it is a cold though as the symptoms appeared gradually and there was no sudden onset. To be honest, I really hope I do have the swine flu since this is mild.
This review was just released today by the Institute of Medicine:
Respiratory Protection for Healthcare Workers in the Workplace Against Novel H1N1 Influenza A: A Letter Report
http://www.iom.edu/Object.File/Master/72/970/H1N1%20report%20brief%20FI…
The committee focused solely on the scientific and empirical evidence regarding the efficacy of various types of personal respiratory protective equipment (e.g., medical masks and respirators). Studies on influ¬enza transmission show that airborne (inhalation) transmission is one of the potential routes of transmis¬sion. N95 respirators are documented to filter out 95 to 99 percent of relevant particles and have maximum effectiveness when properly fitted to the face of users. Research results on the filtration and fit of medical masks show wide variation in penetration of aerosol particles and inadequate fit suggesting that the use of medical masks is unlikely to be effective against airborne transmission.
Just a technicality FYI-- unless the 757 aircraft mentioned was a specially configured charter aircraft, a 757 is a two cabin configuration with 2+ 2 in first class and 3+3 in coach, or rarely, one class with 3+3.
I wonder if the draconian containment policies might have affected the willingness of those on the flights (but not part of the tour group and thus were not already detained) to "fess up" to flu symptoms when contacted by authorities.
George T--Iâm in the same boat that youâre in--itâs sort of weird to be hoping for the flu, right? I suppose nature will keep us both guessing as to whether or not we have âit.â
I think my sonâs chances of having a flu are very likely, however, because he is a teenager living in a college dorm in Georgia that has at least one âofficial probableâ H1N1 case, and one âofficial probableâ seasonal flu.
I also think that our population of college kids, particularly the dorm-dwellers, should be regarded exceptionally. They are at such high risk for rapid transmission and infection. In a way--they remind me of the dry chaparral that kindled and eventually ignited those wildfires in Californiaâ¦.
Well, my situation is my girl just started school for the first time a week ago Monday. She was sick and missed school on Friday. She was bummed out about missing school even. Then she infected the rest of us. There is little doubt it's from school.
I have no way to find out if there is any H1N1 going around. All I know is that I live in an area that was closing schools a few miles away in the spring so I would assume it is still lingering in the community.
I'm going on the probably not theory and going to get the shot(s) when available. I once had a 3 week "death flu" (being dead would have been an improvement over living) and never want to go through that again.
Tymbuktu: The paper had no details on seating but I went and looked up seating plans. The two class versions had the class division (along with a barrier or curtain) between rows 7 and 9, which seemed that there would be a barrier between the cases that would have been mentioned had it existed. OTOH there would was mentioned single-class versions, and we assumed that was what was used in the South China flights. The seating charts we found also were two and two. So we might be wrong about our guess, but it was based on looking to see how this plane was configured. The paper did not contain that information.
Mark:
"...suggesting that the use of medical masks is unlikely to be effective against airborne transmission."
Is that your statement or theirs?
The statement is from the IOM report.
Anybody --
I'm having difficulty tracking down hard, "scientific-data" type information relative to the question:
Is there a relationship between the severity of a case of flu and the level of the initial exposure to the virus?
The following is the text of a brief diary I posted at the Flu Wiki:
Is there a relationship between the severity of a case of flu and the initial level of exposure to the virus?
For example, say that adult identical twins A and B attend a party at the same time. Twin A wears no mask, but twin B wears a mask that filters out 90% of airborne flu virus particles. As a result, B inhales ten times fewer infective particles.
Assuming that both develop a case of flu, will B's case be expected to be less severe? Will B be more likely to survive? Is it possible that B will not even develop a symptomatically noticeable case even though A does?
link: http://newfluwiki2.com/diary/3724/case-severity-and-the-level-of-initia…
There's a fair amount of discussion and speculation, but so far, no "hard" information has been cited.
Can anybody help?
Mark:
Thanks.
Based on my current understanding, I would respectfully have to say that the validity of their conclusion is not clear to me.
In the "Experimental Influenza Infection" section of Tellier's review ( http://www.cdc.gov/ncidod/EID/vol12no11/06-0426.htm ) of aerosol transmission, he refers to a paper by Alford et al ( http://www.ncbi.nlm.nih.gov/pubmed/5918954?dopt=Abstract ) that reports a 50% human infectious dose (in terms of a 50% tissue culture infective dose). So at least for whatever flu virus Alford et al used, there is a dose-response curve.
That being the case, then even if a mask blocks out only, say, half of the viral aerosol droplets, it would be expected to at least somewhat lower the probability of transmission.
For the record Tymbie, it doesnt matter about the configuration as the studies were done when the SARS epidemic started. HEPA filters were installed into the air systems and the delivery rate from the bleed air is just about the same all across the cabin. The biggest problem is sitting next to the air returns. If the air is contaminated and its running across your face and you inhale it doesnt matter if you were sitting in the john or in first class, you are going to get the bug inside and thats the biggest thing..
Flew to Atlanta two days ago on a day trip and back in my full face plate mask, wore it on the main concourse heading to the gate too. Had to remove it at the security check point. Some of the TSA people had paper masks on. Note to self : I if I have to fly I will buy plastic covered filters so I dont have to remove the mask. Likely no vector putting it on the deck but if they touch it with their gooked up gloved hands then they transfer that shit to your mask. Out came the disinfecting baby wipes, over to the corner away from people and a full wipe down was completed. Got only a nod from the pilot and one of the flight attendants in first class... Funny when she asked me if I wanted a drink. I said sure... I think she thought I was kidding when she rolled up with my seatmates drink and I raised my hand about that drink. Up it comes and drinking tube attachment goes on the mask, the mouthpiece is rotated into position and young Randolph enjoyed his vodka and OJ very nicely through the tube arrangement. Its big enough to drink soup though.
As for that infection thing Mark and Sal... Back when H5 was the big thing the Chinese built some sort of poof cannon that shot virus particles in some sort of media into cages with ferrets. If I recall they were trying to see how many estimated particles it took to be infected. They finally gave up as they move to the lowest setting that they could create in a cage and have it shot through in some sort of media...They ran out of ferrets. Its not helpful in relation to the less exposure or you got sick anyway argument but its about the only thing that I can remember.
Is there a simple diagnostic test for "antibodies to H1N1" like there is an HIV antibody test? (ELISA)
It might be useful to know if a person has antibodies to H1N1 (has been exposed to it).
o.jeff
From Dr. Racaniello:
"Some years ago an ELISA kit was developed to detect antibodies against H1N1 swine influenza (not the current strain). You can read about it here: http://www.jvdi.org/cgi/reprint/16/3/197. I believe the product is HerdCheck Swine Influenza Virus H1N1 Antibody Test Kit, IDEXX Laboratories, Inc., Westbrook, ME. I am not aware of a similar kit to detect antibodies to the 2009 H1N1 strain."
O. Jeff.... Quick test but I am hearing they are not accurate in 30-40% of the cases and its only a typing test by A, B seasonal....
Reason its off? Antigen change? Dont know. RT-PCR test will give you the noise but not in time to save you very likely.
Ssal - good point, but the IOM found that surgical masks likely provide an order of magnitude less protection than NIOSH approved respirators, not only 50% less. This is because of the less effective filtering material and lack of a tight seal against the wearer's face. The ultimate policy issues is - how much risk should healthcare workers be expected to assume to do their jobs. Most infection control folks assume that a higher level of risk is okay compared to most occupational health folks, but it is not their decision. Here in the US, the acceptable level of risk to workers has been previously decided by Congress and by several Supreme Court decisions.
On an ethical note:
If the Chinese government exceeds its reasonable authority and detains ill people, is it possible to get informed consent from such people to do an epidemiological study?
epistemology: It's not clear you need that in the US, either, for a public health emergency situation.
O. Jeff
http://www.cdc.gov/h1n1flu/guidance/rapid_testing.htm
Question: Yesterday we attended an out-of-town family birthday party. Upon attending, my sister-in-law told me my niece had "a small cold," and that she had caught it too. No biggie. Then their cousins arrived, the ones who had passed it to them. Both had noisy coughs a day and a half after developing a fever.
How likely is this to be H1N1? In my experience as a parent, colds rarely cause a noisy cough and if it does, it usually develops several days after onset. These kids had fever followed by terrible coughs within 48 hours.
The silver lining is that all had mild symptoms -- out of the three kids that were "ill," all were up, around, and playing happily. The worrisome news is that because it was an out-of-town party, I couldn't think of a reason to cut our visit short, so I kept my kids outside playing as much as possible to avoid those close quarters.
Is there anything we can take 24 hours later (I'm thinking Emergen-C? that would help if we were exposed?)
Curious: Not possible to say, but there is likelihood of flu. Common colds don't cause fever. But there are a lot of viruses out there, including respiratory syncytial virus and metapneumovirus and adenovirus and . . . and . . . besides influenza. So best to keep an eye on everyone and if they start looking sick, bring 'em in to the pediatrician or whoever gives medical care to them. My son in law just had a cough and fever, felt better for a day and then it came back with fever. That's a warning sign for secondary pneumonia and after a day or two, at my urging, he went in and yep, he had pneumonia. They gave him some zithroycin and within a day or two was much better (although after a week he still has a cough). Kids can go sour much faster, though, so be vigilant but don't get crazy over it. Most cases just resolve.
Thanks, Revere. I'm keeping a close eye on everyone. I had a pneumonia shot this spring, as I'm asthmatic, but my kids are so young (6 and 4), I'm hoping that their pneumovax shot will still provide some protection.
What struck me was the speed with which everyone got sick. My SIL babysat for the sick kids Thursday at lunchtime and by Friday afternoon, she and my niece were ill. This doesn't seem to follow the three day incubation period.
I'm also surprised by your statement that colds don't cause fever. I never knew this. So when my kids get a fever with respiratory symptoms, that means they're fighting a different bug like one of those you mentioned? I always thought it was a cold! Good to know, though I'm sure in most cases the treatment is the same.
Knocking wood -- and glad to hear your son-in-law is recovering.
Curious: Any "common cold" (many possible viruses) can cause fever but they usually don't, and this is one of the main ways to explain to people that flu is not "Just a bad cold." Muscle aches and pains, characteristic of flu, are also very uncommon with colds. But unfortunately there are a lot of things that aren't "head colds" that also aren't flu.
Regarding the incubation period, evidence is piling up that swine flu has an incubation period of under two days, so what you describe is not inconsistent. But keep an eye on things but also take a deep breath before doing anyting and try to say calm and clear eyed. When your kids are involved, that's pretty difficult, as I know personally, but that's my (pretty worthless) advice.
You know parents too well! But yesterday (and the fact I didn't hightail it out of there when I realized the other kids were sick) is proof positive I've gained some perspective.
Everyone is symptom free right now, which means I'm (mostly) anxiety free. If that changes, we'll deal with it then. Lots of wood knocking going on, and it's a good reminder to keep up what we've been doing -- lots of fruits and vegetables every day, plenty of sunshine and sleep, and keeping hands clean.
Thanks. Oh, and you were missed during your vacation -- I come every day to read your posts!
Dear Revere,
Some of us are trying to get the infection control community to accept the concept of aerosol transmission vs. droplet/ airborne transmission since there is no bright line separating droplet and airborne transmission. Respiratory aerosols containing particles of many sizes are expelled when people breath, talk, cough, etc. Many of these particles are respirable. Close contact with an infected person can result in droplet spray exposure to the mucous membranes of the eyes, nose and mouth. However, close contact also maximizes the opportunity for inhalation of respiratory aerosols. Thus, the association between droplet spray exposure and infection is confounded by inhalation exposure. In the situation described above, it is impossible to say whether the close contacts of the case became infected via inhalation of virus or via droplet spray, although in animal models inhalation of influenza virus causes infection much more efficiently than instillation of nasal drops. The absence of evidence of long-range transmission does not mean that transmission via respiratory aerosols/inhalation did not occur.
"There's little evidence any of this does any good to contain influenza"
I guess it was based on SARS, which appeared to be communicable before symptomatic. This probably still does some good. If someone is communicable a few days after symptoms they won't spread it further. Very unlikely to stop a spread (the pre-screen R0 would encompass a very narrow range where these types of distancing measures would work), but slowing it down looks good. It seems like learning on the fly happens in 1/2 a week fits and starts so time is nice.
My beef wouldn't be with China here in 2009. It would be with Russia. I'm still waiting for a follow up to their GMO potato research.
"The absence of evidence of long-range transmission does not mean that transmission via respiratory aerosols/inhalation did not occur."
This is relevant for face masks, designing face masks, bed spacing and whether or not to open windows. If you find 95% of transmission is little droplets or only 5% is, that's probably good enough for procurement purposes (for the part of USA healthcare budget that doesn't go to bloated insurance industry), or something like that.