Back to the West Nile Virus listserv archive.

Date: Fri, 25 Oct 2002
Posted by: Michael Gochfeld (gochfeld@eohsi.rutgers.edu)

Lois Levitan wrote:

I would like to make a few points following up on Rich Lampman's comments re: magnitude of perceived and real risks relating to West Nile Virus. My comments are on the numbers, and what's behind them, not on the substantive issues of the discussion about risk, which is an important discussion to have, but for my 2 cents will have to wait for another time:

Rich mentions the oft-repeated statistic that 1 human case of WNV in 150 is said to lead to serious complications. He notes that these numbers are drawn from a 1999 serosurvey in Queens, NY.

1. Both Rich and I have made the same assumption re: the source of this statistic. If CDC or other epidemiologists know otherwise or could elaborate, I would appreciate if they would post this information to this listserv. Since the statistic is being repeated so widely and used for different purposes, it would be helpful to have a clear and precise explanation of how it was derived.

The queens serosurvey data that were made available in mmwr do not allow a clear interpretation since denominator data are not available. It would have been easy for c.d.c. to provide more useful data, but that's what we have available at this time.

There are also data published on the romanian epidemic of 1996 which are also not complete. The authors inferred about 94,000 infections leading to about 300-400 identified cases with about 10% fatality (this is from memory). C.D.C. played a role in that investigation and i was told by one of the authors not to rely dogmatically on the estimates, that they might be too high or too low (but that's true of any estimates).

So as i said in my previous message, it wouldn't be difficult to design a study of seroprevalence in selected areas (including some out west where WNV is still spotty). It might reveal that the virus has already arrived in places where it hasn't been identified.

2. Rich compares the <1% serious illness (ie, "1 in 150") with the 6% fatalities in Illinois this year. However, his comparison is based on an error of interpretation: The "1 serious case in 150" refers to 1 case showing serous clinical symptoms out of 150 infections. Most people infected with WNV do not know they are infected (unless they have been part of a research serosurvey). What the "1 in 150" implies, therefore, is that for every person who is hospitalized we should multiply by 150 to get a handle on how many people have been infected.

This would obviously be useful to know.  there is at least one other stratum, maybe more

Bitten by infected mosquito  ----->no infection established

Infected person

Mild illness  (not identified as west nile), maybe flu-like

Mild CNS infection----headache

CNS-infection diagnosed and hospitalized------->recovered----->no sequelae

Fatality                                                                                 Permanent sequelae

There is no assurance that the percentages for each transmission would be the same from place to place or year to year.

Lois's following point is correct and needs to be conveyed clearly. Michael Gochfeld

3. The comparison that Rich is making is in the fatality rate among people who are known to be WNV-positive. This has been about 12% of hospitalized/known cases over the past 3 years prior to 2002. Thusfar this year it is far lower than previously, about 5-6% of reported cases.

However, the numbers are not as crystal clear as one might suspect. The "numerator" -- or number of people who have died -- is clouded by assessments of the cause of death, when a WNV-infected person is already ill from other diseases. And the "denominator" -- or number of WNV cases -- is clouded by unclear standards in what is counted as a "WNV case." In the first year, 1999, only serious central nervous system infections made it into the news and were counted; in the second and third years a few sub-clinical cases were recorded as WNV-positives. This year it appears that many more sub-clinical positives are being recorded. But how many? It would be useful to have a tally that distinguishes between cases that led to encephalitis and others, so that the media and others can quote reliable statistics, if for no other reason.

4. I wonder how precise the "1 in 150" figure is, if based only on the Queens serosurvey of 1999, a study involving fewer than 700 people in the small geographic area that was the initial WNV hotspot. How much confidence do epidemiologists have in the "1 in 150" figure (now and in 1999)? Can anyone shed some light on this? Or provide information/interpretation from subsequent serosurveys? Or information about serosurveys that are planned or in process?

5. Another statistic that I believe is also derived from the 1999 Queens serosurvey is that about 20% of WNV-infected people have mild flu-like symptoms. This is derived, I believe, from responses from the 19 people in the Queens serosurvey who tested WNV-positive (of the approximately 700 participants) None of the 19 had serious illness nor were among the 60 or so people in the NYC metropolitan area who showed clinical symptoms that year. Ie, It would not have been known these people had been infected with WNV if the blood tests had not been done for research purposes.

Six of the 19 WNV-infected people said they had not been feeling well in the past few weeks (ie, about 30% of a small sample) while about 10% of the non-infected population said they also had not been feeling well. I believe that the 20% figure was derived by subtracting the 10% background noise from the 30%. This is a very small sample. Have other studies shown similar results?

I look forward to feedback on the questions raised here. Thanks in advance.

-- 
-----------------------------------------------------------------------------------------------------------
WESTNILEVIRUS-L is an email discussion group for communication
and discussion about West Nile Virus, particularly regarding policy,
risk reduction and public education issues. It is moderated by
Dr. Lois Levitan at Cornell University's Center for the Environment.
To subscribe (or unsubscribe), send an email request to envrisk@cornell.edu.
Subscribers are encouraged to post to the group by sending an email to:
WESTNILEVIRUS-L@cornell.edu.  Please send only unformatted text,
without attachments.   Archives  are posted at: 
http://www.cfe.cornell.edu/erap/WNV/WNV-LArchiveIndex.cfm
------------------------------------------------------------------------------------------------------------