Effect Measure

CDC is reporting about 14,000 confirmed or probable cases of swine flu, although they have already said this may represent only a fraction of the total number of infected. I wasn’t able to find the latest number of those hospitalized (some of you probably have it but I couldn’t locate it with a quick search), but my recollection is that it is somewhere around 300. That puts the hospitalization rate at around 2% of the confirmed/probable cases (I round figures liberally because they are too uncertain to worry about precision). 2% is the same number the Chilean health authorities are using for serious complications there:

The Institute of Public Health in Santiago reported 24 new cases of the virus, officially known as A/H1N1, taking the national toll to 393, it said in a statement on its Web site yesterday. About 98 percent of infections in the South American country result in ?mild? disease. The rest are serious, and one was fatal, the institute said. (Jason Gale, Bloomberg)

The flu seems to be spreading with ease in Chile as it is in Australia, Japan, the UK and Spain. There is no doubt that this is a pandemic strain. WHO has yet to raise its alert system to the pandemic level (6), which emphasizes an important point: it doesn’t matter. No one who wasn’t in a deep coma needed to be alerted. I quoted once before the military adage that no battle plan survives first contact with the enemy, and this goes equally for pandemic flu plans, it seems. The WHO pandemic alert system, which was instituted in 2003 and had never been seriously tested until this outbreak, immediately met a pandemic it couldn’t handle, not because it was so severe but because it wasn’t severe enough. As a result it turned out the alert system was more of a problem than a help. The WHO could have merely announced that a novel strain of flu had spread world wide and should be considered a pandemic strain of lesser to moderate severity. We wouldn’t then be arguing about the obvious.

Instead of revamping the alert system, WHO should consider scrapping it in favor of an up-to-date information system. Thanks to the internet there is a global communication channel that makes any official international alert system obsolete. Use of the system doesn’t empower WHO under the International Health Regulations, so nothing is lost.

Unlike some, I’m sympathetic to the fix WHO got themselves in over the pandemic alert system. WHO didn’t see it coming and neither did anyone else I know of. We all accepted that the system made sense, was needed and was workable. So we all learned something from this.

Let’s move on.

Comments

  1. #1 rayj
    June 6, 2009

    An alert system is necessary because important actions can be triggered by its levels. What is needed is more study of methods. ‘Number of cases’ has not worked well enough. Adding ‘number of regions affected’ is not working either, because the way in which these data are to be combined for a final determination is not well defined. There are examples in other fields that one could consider, for instance the Torino Scale for asteroid impact hazard:
    http://en.wikipedia.org/wiki/Torino_Scale
    And the scale need not be just two-dimensional; we all have computers. Without a systematic method for communicating disease risk, however, we would be left with only, “I’ll know it when I see it.”

  2. #2 revere
    June 6, 2009

    rayJ: On the other hand, it is just because some national actions (both good and bad) are triggered by alerts that it has become subject to manipulation. Coupling actions to alerts is part of the problem, here. And nothing requires (or prevents) member states from taking actions on the basis of good information not tied to an alert system. The alert level is irrelevant to countries in North America now. It would make no difference. And the difference it makes to other countries is precisely the reason they don’t want it called a pandemic, even though it quite obviously is.

  3. #3 Snowy Owl
    June 6, 2009

    Health Services vrs Governemental Services

    It was written and we have a clear example this week in a First Nations Reserve about 350 miles North of Winnipeg, Manitoba, Canada.

    A remote First Nations reserve becomes a federal-provincial battleground, … meanwhile for People of St-Theresa Point,

    From Crofsblog
    http://crofsblogs.typepad.com/h5n1/2009/06/manitoba-health-plays-dumb.html

    St. Theresa Point, a northern Manitoba community from which 12 people suffering severe flu-like illnesses were flown by air ambulance this past week to Winnipeg. Two women and 10 children have been admitted to hospitals in the city.

    The two women are listed in critical condition and one of them, who was pregnant, has lost her baby.

    However, provincial health officials would not say if any of the new swine flu cases included those people.

    Then I invite you to read
    Something is wrong with Manitoba Health
    http://crofsblogs.typepad.com/h5n1/2009/06/something-is-wrong-with-manitoba-health.html

    A day later the Band decides to take the matter in their own Hand

    The school has been closed for several days, and public gatherings are forbidden.

    The band council has been instructing residents on how to avoid spreading the virus, using local radio and television to get the message out.
    http://www.cbc.ca/canada/manitoba/story/2009/06/04/mb-swine-flu-first-nations.html

    In Thoughts about St. Theresa Point
    Crof write
    Someone needs to take Ms. Weiss (Manitoba’s acting chief public health officer) aside for a quiet word about the role of good communication in fighting a serious disease outbreak. Not identifying the communities amounts to a cover-up, and Manitoba Health also needs to have a quiet word with their public-relations experts.
    I talked to Chief McDougall yesterday, and he struck me as an intelligent man handling a major crisis with more competence than our provincial and federal health authorities.
    http://crofsblogs.typepad.com/h5n1/page/3/

    Another day later and here comes stygmatisation
    St. Theresa Point: No room at the inn
    These are people from the reserve in Winnipeg 350 miles south.
    Kyle Wyatt, general manager of the St. Regis Hotel, said Health Canada contacted him and told him to transfer the St. Theresa guests to another facility. He did not have an exact number of people, but did say it was guests from five rooms.

    Health Canada did not explain to him why the guests had to be moved out.
    http://crofsblogs.typepad.com/h5n1/2009/06/st-theresa-point-no-room-at-the-inn.html

    Keeping an Eye on the situation Crof writes

    St. Theresa Point: Repercussions (updated)

    Via the Winnipeg Free Press, a remote First Nations reserve becomes a federal-provincial battleground: Feds ignored flu offers: Oswald. Excerpt:

    The province offered Ottawa help with pandemic planning on First Nations 13 times since May 4 but was turned down every time until Wednesday — the day it was revealed 12 residents of a northern Manitoba reserve have been hospitalized with flu symptoms.

    Health Minister Theresa Oswald made the revelation Thursday as she expressed frustration at the federal government’s response to the H1N1 flu outbreak when it comes to First Nations like St. Theresa Point, saying the province has been prevented from doing more because it doesn’t have jurisdiction.

    There are now 21 people from the community in hospital in Winnipeg and hundreds more in the community itself have registered with the nursing station as suffering from respiratory symptoms. So far just two cases are confirmed as H1N1 influenza.

    Oswald said on Wednesday the province was asked to help find housing for family members coming to Winnipeg with sick loved ones because they were having difficulty finding hotel rooms.
    http://crofsblogs.typepad.com/h5n1/2009/06/st-theresa-point-repercussions.html

    This is an exemple on the repercussions of an epidemic within a Minority in a xenophobic environment.

    This First Nations in Canada might not be different of ANY minority amisdt a xenephobic environment in the World.

    Of course many (if not all) Countries needs time to ajust their National Responses to a declaration of a pandemic by the WHO, but meanwhile, the least governments can do is to pro-actively provide informations like Preventive Measures, Warning Signs of an infection and how to Cope with an infected person.

    Regions, Locals and Communities must be ready to intervene with Pertinence and Aware of the risks.

    Pro-active distribution of information can be valuable.

    Taïwan Centers for Diseases Control reached out and pro-actively relay a confirmed H1N1 infected in our Region.

    Government must at least distribute swiftly informations.

  4. #4 tess
    June 6, 2009

    The WHO has been responsible all along to provide up-to-date information, without obfuscation, skirting responsibility, or bending to political pressure. The choice you present for WHO
    (“Instead of revamping the alert system, WHO should consider scrapping it in favor of an up-to-date information system.”) would be a favorable direction for a public health organization that has so far largely failed THE PUBLIC. History will tell WHOm public health protected FIRST.

  5. #5 Pete A
    June 6, 2009

    Because of the sheer numbers involved no pandemic plan was ever going to be successful.
    The fear governments seem to have that people are going to “panic” seems to be holding many regions back. I recently read in a local newspaper there was a case of H1N1 in the county but the location was not divulged probably for fear of causing “panic”. But if people were informed the risk of contracting the virus was imminent I am sure the information would spur them to better practices regarding hygene rather than inspiring irrational behaviour.
    I am currently trying to encourage my family to use alcohol handwash after visiting supermarkets, playgrounds ect.
    While there is still allot of misinformation and a common cynicism regarding H1N1,somewhat compounded by the WHO’s reluctance to declare a pandemic convincing people there is not a need for concern.
    While here in the UK there has been a blanket of leaflets and adverts people these days have a very short attention span and I think more information regarding the level of threat they face would be worth the risk of this mythological “panic”.

  6. #6 Spence
    June 6, 2009

    I don’t believe that the current alert system did not work. Even though the alert systems were built around the threat of a more severe strain, such as H5N1, it worked well in raising the attention to the current strain, which may still have a significant impact in the Northern Hemisphere when it returns in the Fall.

    Although the exact numbers are still hard to define, they have been reporting hospitalization rates of about 2.5% and a mortality rate close to that of seasonal flu of app 0.2%.

    However, we need to keep in mind that while seasonal flu kills an estimated 36,000 people a year, the great majority (~91%) are elderly (of which many are frail with multiple chronic diseases). The swine flu has been mostly sparing the elderly to date and has had the greatest impact on the young and high risk populations (predominately those with chronic disease but also also including high risk groups for those less than 2 years of age and pregnancy).

    Applying the information from above, during the typical seasonal flu, approximately 3,200 people under age 65 would die (9% of 36,000). However, with a novel strain, such as 2009-H1N1, given the likely higher levels of contagiousness (could be 35% of the population getting sick (typical to pandemic strains), rather thant the typical 5% – 10%). With this higher levels of contagiousness and younger population effected, we could see as many as 210,000 deaths in the U.S. primarily in the high risk groups and younger population. (0.2% X 300,000,000 population in U.S. X 35% infection rate). – Please check my math to see if I missed something.

    In addition, if you apply the same calculation to the reported hospitalization rates (2.5%), then we could have in excess of 2,600,000 requiring hospitalization (with less than 1,000,000 total hospital beds, most of which are currently occupied. (2.5% X 300,000,000 population in U.S. X 35% infection rate)

    These numbers could very well be high, given that we don’t have a good handle on the actual infections to calculate a good denominator (but built off what is being reported by CDC & WHO). However, even if they are off by a factor of 10, this will still have a pretty severe impact on our health systems. This is also assuming that the virus does not become more severe as it travels through the Southern hemisphere.

    If this was off by a factor of 10, even 20,000 deaths among the younger population and 260,000 people requiring hospitalization would be pretty significant to our health systems.

    All of this is also not considering the impact on many parts of the world that may see a more severe impact (e.g. third world countries), where they typically see a higher illness and mortality due to overall health status, nutrition and crowding levels.

    Based on what we are seeing to date, this may still be a very significant event for us when it reemerges as a second wave, potentially in the early Fall, even with an overall low mortality rate.

    I think the WHO phases, especially if a severity index is added, have done a good job to-date to warn us about this potential risk, even if the first wave in the Americas, was relatively minor.

  7. #7 M. Randolph Kruger
    June 6, 2009

    My only problem with the suggested new system is that it doesnt inform people except what to expect. Now they are talking about a severity level. But due to economics, its now yet another tier that will get people dead. So we have pan level six, severity level A or 1? Parsing…. As all of this would be rendered moot if it suddenly turned high path and we had our thumbs stuck up our tails. The suggestion of this stuff is ridiculous. Kind of like saying you need a requisition form for ammo on Dec. 7th, 1941 as the planes were overhead.

    The whole idea was to get people to prepare, to do something personally and this is now slipping back again to the all encompassing government to take care of the problems. They will pay for this in body counts due to illness, complacency generated by those that were and are complicity, or even on the high side…starvation.

    I think Revere you would agree that this hasnt worked the way it was planned and this particular pandemic just didnt want to play the way they wanted. Next month and towards August, they’ll be talking still about this. Australia is the key here and it aint looking too good for my buds down there

    It also shouldn’t escape anyones attention that all flu information in Indonesia has been canceled in lieu of a 500 million dollar contribution for a vax factory and a new dacha in Thailand. Not a word about H5N1 or H1N1, but the clinics there are reporting huge surges in “pneumonia” and dengue fever.

  8. #8 Racter
    June 6, 2009

    What I’d most like to know is: out of those ~14,000 confirmed or probable cases, how many have recovered and are therefore no longer contagious? In other words, it seems to me that it is not incidence that best reflects the current stage of the event, but prevalence.

  9. #9 revere
    June 6, 2009

    Racter: The number of US hospitalized I am told (h/t SusanCC) is above 700, roughly 5% of the confirmed/probables. If the number of true cases is 20 times higher, that’s roughly .25% of all cases, or 2500 per million. If 20% of the population gets infected, that’s 60 million people or roughly 150,000 hospitalizations. That’s roughly the same as the US seasonal flu average (200,000, although it varies significantly from season to season). Things start to get dicier if many more than 20% get infected, as may happen with a novel virus. It also depends on how nasty the virus is next fall. So we could be in for a rough ride. Or not.

    Regarding incidence (=risk) or prevalence (=burden) it depends on what you want to know. For health services, prevalence is what gets you, but of course that is a function of incidence and duration. The kind of survey NYC is doing will give prevalence, not incidence.

  10. #10 Lisa the GP
    June 6, 2009

    Well, at the risk of being overly simplistic, I’m still glad it’s not a descendant of H5N1 we’re dealing with.

    Revamping the ‘alert’ system…really it turned out to be more a ‘record’ system than an ‘alert’, tracking behind events. But, it was a first try at any such thing. We lucked out and got an ‘easy’ case to test it, so…

    Live and learn. The system was a ‘hypothesis’, now we have more data, find that the ‘hypothesis’ didn’t quite fit. Time for a new hypothesis. No big deal. Normal part of science. As you said, lets move on, not let politics and organizational ego get in the way of rationality.

  11. #11 rayj
    June 6, 2009

    Lisa: I agree with you. Have been thinking about some alternatives.

  12. #12 Snowy Owl
    June 6, 2009

    Reveres,

    There is more at steak than exclusively U.S. interests when we debliterate upon WHO.

    I do feel and understand how important North Americans are but we are talking about a World Wide Organization that had, has and might have a lot of influences on citizens of many countries, way outnumbering the rich people of North America.

    Scraping the whole system will bring WHO down in the eyes and consciousness of many Cultures.

    It is True that we are lucky to encounter a presumebly mild virus but the severity addentum should not be disregard a priori.

    Each country should assess their severity index and then getting it validate by WHO.

    We are in it now, it is not an hypotesis
    cf: Photocopy of Point Theresa Reserve datas
    http://crazybitchesrus.blogspot.com/2009/06/st-theresa-first-nation-if-is-isnt.html

    That’s what it look for isolated Community and in fact just in case you haven’t notice, we are really on our own exept when someone dare to care and forward relevant informations.

    Please do not flusch the baby with the water of the bath.

    Snowy

  13. #13 travc
    June 6, 2009

    Adding a second dimension of “severity” to the current scale “prevalence” seems in order. Severity takes longer to determine, so initial actions would be based/triggered by prevalence (following the principle of exercising the most caution). When severity data starts coming in, more often than not the actions would be scaled back and people will can say “Whew, not as bad as it could have been”.

    I instinctively dislike strict rules-based procedures, but in this sort of thing (the initial stages of a potential disease outbreak) it really is needed. Too many people have far too little information to act intelligently and too many interests would fight against potentially loosing some money by reacting.

    Explicitly adding a severity scale would allow adjusting the response actions at the time when it actually makes sense. The costs and inconveniences of declaring a potential or even a full-blown pandemic would be less frightening since they would only last until we could get enough info to tell if they are really needed “this time”. On the other hand, there would be a hell of a lot of pressure to determine the severity quickly and return a “not severe” result.

  14. #14 revere
    June 6, 2009

    Snowy: The alert system has damaged WHO’s credibility. The importance of WHO to the rest of the world is why I think the system should be scrapped.

    travc: A severity system for the world is not possible. Flu often has different virulence in different locales and locations. WHO opposed such a system until recently when it seemed to accept it under pressure from some powerful member states. If the virus has different severity in one place than another, what will you do? Add some more “dimensions”?

  15. #15 MadScientist
    June 7, 2009

    Personally I think that anyone who believes they can control a virus outbreak is delusional and anyone promoting a system to “save the world” when such an outbreak comes up is really just promoting their ego (and possibly hoping for more money for flapping their gums).

    Australia’s reactions have severely disappointed me – from suggestions of quarantine of anyone traveling from Melbourne to the nation’s capital, and the issue of TamiFlu to people disembarking from a cruise ship. Although TamiFlu will do zilch, the public officials declared “it’s better than nothing” – morons – they may as well have handed out TicTacs, at least it’s cheaper.

    As I’ve stated elsewhere, imagine trying to lock down New York City if it were hit by a devastating disease. Good luck.

    So, first we need to forget this nonsense that the spread of just any disease can be easily controlled; that is simply myth. For a good nasty bug agent A is likely to have infected many others before being discovered and isolated and the others are likely to have spread the disease before you even track them.

    So what would help? Well, perhaps nothing at this point in time. When “designer vaccines” are possible to create and test within one or two weeks and to mass manufacture within the same time frame then perhaps something useful can be done aside from tracking the spread of a disease.

    Whatever happened to that atypical pneumonia ‘sars’? It would be interesting to know why it fizzled out. I see some are claiming it’s due to action by the WHO, to which I can only respond with maniacal laughter.

  16. #16 M. Randolph Kruger
    June 7, 2009

    Lisa, picture this. They have haggled now since March about raising the pandemic level… pushing three months. No one did a thing to stop it because it was mild. Well shit if thats the criteria then they blew that too. Its a pandemic and now they are adding new tiers to already ugly. Cant get them to call pandemic until they have the new tier level next week.

    If this were 1918 and we were and are ascending the W on the wave then we very likely are just getting started. I think everyone can agree on that. But just to get to 5 it took 48 days I believe is the number. Damned glad it wasnt smallpox. But 48 days? In that time it was already across the planet….I think they call that a pandemic. So now we will have parsing up the kazoo. What ? Localized pandemic level six parsed to severity level A or 1 based upon samplings and NOT testing.

    Its expensive but who is going to allow the WHO to call pandemic in say Indonesia? You gotta test to have a pandemic of H5N1 or H1N1. They cant have this both ways and it will NEVER be called even if it is. Everyone is complacent about the last 3 tiers now… so we gonna get more toys and levels and job justifications?

    This bug if its 1918 all over again is from a timeline now moving into the August phase of 1918. Not a lot of cases, few deaths and it was all good. Then bang… the swine flu hit the fan. Rocketing numbers of cases and deaths.

    So we pay these people for WHAT? Better to just call it epidemic and let every country make the call on their own. Then we will have no testing, no reporting, no epidemics or pandemics. They didnt stop travel to Mexico or call it an epidemic because of the economic effects…So we now have to couch everything around world economic recovery… That works real good on dead people.

    Move on? Seems we cant get past square one.

  17. #17 anon
    June 7, 2009

    just because something sometimes doesn’t work perfectly is no reason yet
    to “scrap” it.
    Improve it.
    The more phases, the better. Why not phase 5.3 or such, when we passed 5 and get closer to 6 ?
    Why not add a second scale for severety ?
    When countries rely on WHO-phases and plan their responses on it,
    that’s their decision and not WHO’s.
    If the WHO-phases were badly defined, but countries still base their plans on it,
    then why didn’t they complain earlier, when the definitions came out ?

    Just because some statement may “damage” one’s credibility later is not a reason
    to switch back to a “noone knows” -philosophy.
    Do the best what you can, discuss it and try to improve.
    Better a bad WHO-system with phases 1-6, than none at all … when all sorts
    of even less competent people create their own phases, non-uniform
    and worse in average.

  18. #18 Pete A
    June 7, 2009

    I totally agree with Mr Kruger – well said that man !

  19. #19 risk_comms_guy
    June 8, 2009

    Had the W.H.O., instead of announcing that its pandemic alert had gone to level 5, taken Revere’s advice and “announced [that a] novel strain of flu had spread world wide and should be considered a pandemic strain of lesser to moderate severity” then I think awareness of the H1N1 outbreak would be far less than is now the case.

    From a communications standpoint, the WHO alert system does engage media in covering an announcement regarding an imminent pandemic. (You may recall that CNN ran Margaret Chan’s news live, uncut, for just under an hour, which was very unprecedented for a health announcement, I’d say.) This media coverage builds awareness among people in general, and forces governments to at least react (if not act) in response to the announcement.

    I therefore conclude that the alert system is a good thing, at least from a communications standpoint. I would much rather see the alert system refined (by including severity, for example) than see it disappear.

  20. #20 revere
    June 8, 2009

    risk: I think this is an open question worth discussing. I am willing to change my mind about this. I have the following comments, however.

    It needs to be clearer what the purpose of the system is. WHO serves its member nations through governments. If the idea is visibility for a particular issue, that’s one thing. If it is to induce action outside the IHR, it is another. Both may be legitimate, but they are not the same.

    The objective is also related to the consequences of confusion over the communication for the rather fragile authority and credibility of WHO. When WHO shows its technical judgments subject to pressure from member governments (which is normal and natural given the structure of the international system), it loses credibility as a scientific authority, a credibility which may be crucial in a real emergency.

    The problem of adding severity is not easily solved. It is not a single dimension, will be subject to the same manipulation as the other scale and may not be appropriate or feasible. It sounds good, but it is hard to see how it is workable. The current system sounded good, too, but a little thinking about adding severity (whatever that means) shows it is highly problematic. It also doesn’t solve widespread but not severe versus severe but not widespread (we now have examples of both). This is not a no-brainer IMO.

  21. #21 Phillip Huggan
    June 8, 2009

    The WHO Pandemic Phase system post 2006 Avian Flu scare seems to have been designed to account for incremental mutations logorithmically increasing human death-toll or potential future human death toll. Perhaps its better to get two numbers and use the incremental number as a lower bounds and some sort of unknown variable or recombination hypthesis (suggesting sharply ratcheting up flu severity possibility) as an upper bound. Ideally as more is known the spread would decrease and eventually left with one number. But maybe a simple number works for PR purposes, it is obvious the WHO Phase numbers aren’t relevant to high level administrators and researchers.

  22. #22 Risk_comms_guy
    June 8, 2009

    Revere: I agree with several of your main points:

    1. The matter of the WHO pandemic alert is an open question in my mind as well;

    2. I agree that WHO’s authority is fragile at best, and fiddling about with its alert scale erodes confidence in its technical ability and its commitment to evidence-based decision-making;

    3. That adding a severity dimension would not be a trivial or an easy thing to do;

    Beyond that, you state that there is a difference between an alert system intended to serve as a signal to governments in terms of preparatory action, versus one that is intended to focus public attention on a serious outbreak situation.

    I think a by-product of using an alert level announcement to promote public awareness of an outbreak is that it makes it more difficult for governments to decline to respond to this increased public awareness with appropriate steps. The alert level announcement thus can serve both purposes.

    Finally, I agree that the use and abuse of a pandemic alert scale is NOT a no-brainer.

    - Risk

  23. #23 Jody Lanard M.D.
    June 8, 2009

    Very fruitful discussion. I am only adding this comment to underscore the singular wisdom of risk_comms_guy, who wrote:

    I think a by-product of using an alert level announcement to promote public awareness of an outbreak is that it makes it more difficult for governments to decline to respond to this increased public awareness with appropriate steps. The alert level announcement thus can serve both purposes.

    Like the WHO Outbreak Communication Guidelines, the alert level system is ammunition, as well as information, to help publics put pressure on governments, and to help officials and agencies put pressure on each other.