Testing. 1 2 3 Testing. Have I been infected with swine flu?

Lots of things make us sick. Right now, many of us are concerned about getting infected with pandemic flu. How do we know if this has happened? There are four basic approaches.

1. Rapid Tests.

These tests probe viral proteins for influenza signals.


  • Fast. You can have an answer within 30 minutes.
  • Easy to use. Very little training is necessary.
  • Cheap. About $20.
  • Can tell you if you have Flu A or Flu B.


  • Can’t tell you which subtype of influenza A you have, ie, whether you have seasonal flu or pandemic flu.
  • Fails to detect pandemic flu infection 50% of the time.

Given that the vast majority of Flu A cases are currently due to pandemic flu, someone with a positive result with a Flu A rapid test is almost certainly infected with pandemic flu. However, even if you have a negative result, you may be infected with pandemic flu. A negative result on a Flu A test is not a good reason to deny a patient antiviral treatment.

2. Real-time PCR (RT-PCR).

This assay uses nucleotide probes to examine the genetic material of the virus.


  • Specific. This method can be used to determine subtype, ie, whether you have been infected with seasonal or pandemic flu.
  • Sensitive. This method can detect very small amounts of virus and has been specifically designed to identify pandemic flu infections.


  • Takes 24-72 hours to produce a result in most cases.
  • Requires a high level of skill to set up the assay.
  • Can be expensive if done manually.
  • Can give false negatives if the virus develops mutations in the same regions the probes are directed towards.

RT-PCR is the most practical way to determine whether someone actually has pandemic flu. The first three disadvantages can be greatly mitigated if automated, high throughput protocols are put into place. Although this requires a substantial initial expense and considerable expertise to set up, once in place, large numbers of samples can be processed quickly and cheaply with little need for human involvement.

3. Seroprevalence studies.

This methods involves taking blood samples from individuals and determining whether they have made antibodies to the flu.


  • Can be performed after a person has been infected.
  • Can be done on large numbers of people relatively cheaply.


  • Usually not clinically useful.

The primary utility of this approach is to determine how many people were infected in a particular outbreak. This type of study is essential when determining case fatality rate.

4. Unbiased sequencing.

With this approach, a sample is taken from a patient and everything within that sample is sequenced.


  • Highest level of sensitivity. No infections will be missed.
  • Can detect novel mutations. Mutations that might cause other methods to fail will not affect this technology.
  • Can detect novel viruses. Even if an entirely new virus starts to circulate, this method will identify it.


  • Very expensive, although the cost is dropping.
  • Requires a very high level of skill to interpret the results. Relatively few laboratories have this capability.

Given the advantages and disadvantages of the different approaches, what is the most sensible strategy to utilise these methods?

Given the high failure rate for Rapid Tests, there is little justification for their use. Decisions on how to treat a patient should initially be guided by clinical judgment rather than this test.

High throughput RT-PCR should be used to test as many patients as possible. This is necessary to determine how many people are actually sick with pandemic flu as opposed to the many other infectious diseases which can cause similar symptoms. This information is critical to determine if the case fatality rate due to pandemic flu is starting to increase or if vaccines are failing.

Seroprevalence studies should be done (in fact, should already have been done) to calculate case fatality rates in different locations. Telephone surveys of influenza-like illness are not acceptable substitutes.

Unbiased sequencing should be reserved for cases of severe flu like illness but which are negative with RT-PCR assays. This would alert us to new strains or viruses.

The technology and expertise exists to put a sensible, 21st century diagnostic strategy in place that would guide correct treatment and inform policy-making. Unfortunately, the CDC seems to lack the necessary knowledge base and/or will to implement an up-to-date detection system.


3 thoughts on “Testing. 1 2 3 Testing. Have I been infected with swine flu?

  1. You write: “Seroprevalence studies should be done (in fact, should already have been done) to calculate case fatality rates in different locations.”

    I heard that labs not yet been able to develop tests of acute and convalescent antibodies that differentiate the pandemic H1N1 virus from seasonal strains. Being lab-ignorant, I don’t know if this is surprising or not; and don’t even know if it is true for sure.

    Can you find out?

  2. It’s hard to know what the CDC has or has not been able to do because they don’t tell us.

    Here is one reference you may find helpful:

    In vitro and in vivo characterization of new swine-origin H1N1 influenza viruses

    “The assessment of human sera from different age groups suggests that infection with human H1N1 viruses antigenically closely related to viruses circulating in 1918 confers neutralizing antibody activity to CA04.”

  3. Pinoy Kalusugan, there is also the purportedly-leaked,
    “CDC Director’s Swine Flu Update Brief” from July 17,
    p 29 US CFRs http://cryptome.org/h1n1/cdc-071709.pdf

    and, the WHO Epidemiological Report from May 22 Mexican CFRs by age decades,
    p 2 http://www.who.int/wer/2009/wer8421.pdf

    As soon as they found out it was at least as bad as, “Spanish flu” they really haven’t been counting cases in such a way that the public gets told what’s going on.
    “Most” cases have what seems like seasonal Influenza (except Canada found 13% of lab-confirmed cases did Not present with Fever)
    and, “seasonal” flu deaths are over 90% Over age 65; with H1N1, it is the opposite.
    Under age 49 were 60% of US deaths in August, before school opened again to vector more panflu home to the more at-risk age groups.
    5% or 6% of H1N1 cases require hospitalization, according to that May 22 WHO Report,
    which also stated:
    “Severe” cases of nH1N1 lungs’ and outcomes are similar to those of H5N1.

    Labs had at least 4 years of warning to get ready to do real surveillance.
    Science is willing, but Political will is weak…

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s