Announcement

Collapse
No announcement yet.

How to evaluate reports of "mystery" diseases

Collapse
This is a sticky topic.
X
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • How to evaluate reports of "mystery" diseases

    With the large concern throughout Flublogia about the number of reports of mystery diseases scattered around the world, I figured FT could use this thread on what information people should look for in those reports in order to try to understand what is really going on, as well as whether a report is concerning.

    Some of the more important questions to ask yourself when reading a mystery disease report are:

    1. What diseases are explicitly listed in the report as suspect? Do they make sense, given the location and symptoms reported?

    If a disease is listed as suspect in a report, and is consistent, there is a very good chance that disease is involved. Such reports should be taken as though that etiology is likely responsible. If the listed etiology is not consistent because it does not match the location (i.e. malaria in winter in Northern US), or the symptoms (flu suspected with diarrheal symptoms) it is worth keeping an eye on and possibly ignoring the suspect diagnosis.

    2. How credible is the source?

    There are many sources that have repeatedly posted mangled, incorrect, or downright false reports in the past. Such sources have less credibility in future reports.

    3. Is there any reason, other than an actual outbreak, that we might be seeing this report?

    In areas with little health infrastructure, screaming "mystery disease" is one of the easiest ways to draw the attentions of national health authorities. If the article is mentioning poor hospitals or dirty water or such things, it may be a method of trying to improved those issues. Obviously, if those issues would cause such an outbreak (dirty water in a diarrheal outbreak), there could be such an outbreak there as well. There are also many political reasons to generate false outbreak reports, even more so if the source is NOT located in the same country as the reported outbreak.

    4. How many people are ill, and how many have died? What is the apparent CFR? In what area?

    The etiologies suspect in a respiratory outbreak with a 0.5% fatality rate are very different than the ones suspect in an outbreak with a 50% fatality rate.

    5. Is there any evidence for person-to-person spread? Are any HCW ill?

    Obviously, as HCW will almost never have the same exposures as the patient, if an HCW is ill, the illness is almost certainly spreading person to person. Statements such as "Person X died and her father died two weeks later" are also strongly suspect of perosn-to-person transmission. On the other hand, expressions like "Almost every family has someone affected" tend to indicate exposure to a common, but perhaps very large source.

    6. Is there any pattern to exactly who is ill?

    An illness in all adult males tends to suspect an occupational exposure. An illness in all children might suspect an exposure at school. Bird flu as it currently exists tends to infect young adults (who raise infected poultry). A novel virus or a pandemic virus would infect just about everyone.

    7. How long has this been going on? Is there any pattern to when the cases occurred?

    This tends to lead to the epidemiology as well. If the rate of new cases seems to be growing exponentially, that is suspect of person-to-person spread as well. If everyone took ill at once, that suggests a single exposure. If the number of cases per day is roughly constant, that suggest an ongoing exposure, but could also be produced by person-to-person spread with countermeasure (i.e. isolation, contact tracing, quarantine). Also, the longer an outbreak has been going on (especially if we are talking four months or more sometimes), the less concerning its spread would be, as it would have already spread if it could.

    8. How much effort has been put into diagnosis this outbreak?

    A mystery disease in a remote village without a health post could be almost anything, and has only a very small chance of being something novel. A mystery disease in a modern hospital in a developed country has a very high chance of being novel. One of the early signs that Pandemic H1N1 was not seasonal flu was the inability of the Oaxacan health authorities to diagnose it.

    9. How much effort has been put into treating this outbreak? Do antibiotics (or antivirals or anti-parasitics, etc.) seem to work?

    The more effort that has unsuccessfully been tried to treat patients, the more concerning an outbreak is. If certain treatments seem to work, that tends to push the diagnosis in that direction.

    10. Have similar outbreaks been reported from that area in the past? How often?

    If the area rarely sees "mystery diseases", then a mystery disease in that area is concerning. If such things are reported frequently, it might indicate an underlying situation (poor food, dirty water, etc.). Keep in mind, though, that a previously undiagnosed virus does count as an underlying situation, but in that case, we would be seeing the same symptoms and epidemiology in all such outbreaks. Also keep in mind that a large history of such reports (dating back five years or more sometimes) might be a sign of case #3 - that these reports are being used to gain attention.

    11. Are there any other mystery diseases in the general area with THE SAME SYMPTOMS AND EPIDEMIOLOGY?

    That might indicate wider spread. The presence of many different "mystery diseases" in an area is not suggestive of a wider outbreak.

    12. Could this "outbreak" really be the result of simultaneous outbreaks of multiple known diseases?

    That is often one of the easiest accidental ways to get a "mystery disease" report.

    13. How many articles have we seen on this outbreak, and over what period of time?

    Many of the reports we see are a single article (perhas repeated over several days in several different sources) and nothing further. That is strongly suggestive that when the local health authorities investigated, they found that the situation was not as the article described. More concerning is a situation where the source posts daily or near-daily updates on the situation. More concerning still, is when more and more sources start posting the story.

    14. Is there any mention of ill animals?

    Many of the emerging diseases are zoonoses, that is they are transmitted from animals to people. The mention of ill animals might tend to indicate such a disease (i.e. dead birds might indicate H5N1, dead gorillas might indicate Ebola, dead livestock anthrax, etc.). Keep in mind, though, that if something doesn't seem right, we might be back in the situation of #1, where the mention of the dead animals is simply a way of stating that such an illness is suspect.

    15. Exactly what are the reported symptoms of the outbreak? Do they make sense?

    This probably should have been higher up the list. Keep in mind that there might be a difference between the actual symptoms and the reported symptoms. Any reports of incoherent symptoms (i.e. "uneasiness in the eyes", "itching in the body") are a sign of a poorly written report. Reports of symptoms that don't go together often might be a sign of something unusual, but can also be a clue to a compound outbreak.

    16. Exactly where is this outbreak?

    The etiologies that would be suspect in one part of the world, are usually ridiculous elsewhere. Also, knowing how widespread an outbreak is might tend to rule out certain exposures (we would not suspect foodborne illness if the outbreak is spread across and entire state or country).

    I will add to this over time. Anyone else who has such tips should post them in this thread.

  • #2
    Re: How to evaluate reports of "mystery" diseases

    Thank you, Alert. Very thoughtful analysis.

    Comment


    • #3
      Re: How to evaluate reports of "mystery" diseases

      Over the past few days, reports of a "mystery illness" have come out of Nepal. The thread is here

      http://www.flutrackers.com/forum/sho...d.php?p=341047

      This gives us several things to look for in such reports. In football, a yellow flag is thrown when someone does something that just isn't right, signifying a penalty. I see at least four yellow flags in the recent reports.

      1. There is the mention of fever that lasts for 10-15 minutes. That seems to be an impossible symptom. Body temperature just doesn't change that fast due to illness. Other impossible symptoms from previous outbreaks have included bodies that decompose instantly after death, lungs that reach temperatures of 140F, etc. Inclusion of impossible symptoms tends to nullify the symptomatic descpition given in the report.

      2. There is the mention of "emitting saliva". I view that as a nonsense symptom - I'm not sure what that would even mean - spitting? drooling?. Other such symptoms in previous outbreaks have included "itching in the body", "uneasiness in the eyes", etc. If you can't tell what the symptom even means, that too is a flag. Of note here is that nonsense symptoms might indicate symptoms that the witnesses cannot describe accurately, but in this case with the other flags around, that does not seem to be the case.

      3. There is severe inconsistency of the time period over which the outbreak takes place. Some articles say one week, others one month, one says two months, and one actually contains several conflicting time periods. Such flags tend to preclude any estimation of the epidemiology.

      4. There is mention of patients inevitably dying within hours, but the death toll remains consistent for several days. While some illnesses can kill within hours and are difficult to treat, such an outbreak would have an escalating death toll. This is a particularly strong flag that something other than the reported outbreak is causing the report.

      It is conceivable that there is a geographic flag here as well, as the articles do not agree on precisely where the outbreak is.

      While the story is not yet over on this outbreak, all sources have gone silent for over two days (so the reports might be over), with the most recent report suggesting that the "mystery illness" was fabricated to draw attention to underlying disease in the area. I have seen few reports with as many inconsistencies as this outbreak, so thought it would make a good example here.

      Comment


      • #4
        Re: How to evaluate reports of "mystery" diseases

        In press reports from Nepal and India, the notion of a "mysterious disease" seems to be quite common.

        May be one should read : "undiagnosed disease" in stead of: "mysterious disease" .
        “Addressing chronic disease is an issue of human rights – that must be our call to arms"
        Richard Horton, Editor-in-Chief The Lancet

        ~~~~ Twitter:@GertvanderHoek ~~~ GertvanderHoek@gmail.com ~~~

        Comment


        • #5
          Re: How to evaluate reports of "mystery" diseases

          Originally posted by Dutchy View Post
          In press reports from Nepal and India, the notion of a "mysterious disease" seems to be quite common.

          May be one should read : "undiagnosed disease" in stead of: "mysterious disease" .
          I fully agree. I often post the word "unidentified" or "undiagnosed" disease in place of "mystery" in the thread title of such outbreaks.

          Although, I have noticed that more often than not, it isn't that the disease has not been diagnosed yet, it is that it isn't a report of a single disease - either it is several illnesses, or it is an incorrect report for one reason or another. Of all the "mystery illnesses" reported out of those two countries over the past two years, I am aware of very few (less than 10&#37 where a single etiology was later determined. In remote parts of those countries, the health care system can be so nonexistent that the only way to get any health care is by screaming "mystery disease". And that maneuver is not particular to those two countries.

          Comment


          • #6
            Re: How to evaluate reports of "mystery" diseases

            Bump This.

            Here at FluTrackers, moderators and members routinely report on disease outbreak that could represent new or novel infectious agents.

            Often these outbreaks are labeled as "unknown" or "mysterious" diseases. In post #1, alert outlines some critical issues to consider when evaluating media reports on "new" or mysterious diseases.
            http://novel-infectious-diseases.blogspot.com/

            Comment

            Working...
            X