Operational Biosurveillance

Perspective: Bus Driver in Guangdong China Dies of H5N1, No Exposure to Poultry

CBS News is reporting the death of a bus driver infected with H5N1 in Shenzhen, Guangdong Province, China.  He apparently had no direct contact with poultry, and 120 close contacts were asymptomatic. This was the first case of human H5N1 reported in China since June 2010.  A couple of operational perspectives:

• Be dubious of any report of "no activity" of H5N1 in Asia.  Humans report only what they are inclined to tell people based on their ability to recognize disease- note this kind of problem exists right here inside the United States as well.  China's a very big country, and no one truly knows precisely what level of H5N1 is exhibiting in animals or humans at any given moment.  Obviously a human fatality is more likely to provoke recognition (and hopefully) reporting.
• "No direct contact with poultry"- maybe... maybe not.  There remains a large swath of uncertainty regarding environmental contact through restaurants, the bus the man drove, etc.  Point being, don't jump immediately to the conclusion that human-human transmission has occurred without sucking on a large crystal of sodium chloride.  But that said, limited human-human transmission is believed to have occurred in the past- the key is sustained human-human transmission.
• Of course, the man's occupation as a bus driver is the key piece of information jacking people up- i.e. the thought of multiple undefined contacts in a city of 10 million people is enough to force any public health officer into a cold sweat.
• "120 close contacts were asymptomatic"- ok, but if one wants to negate a hypothesis of sustained human-human transmission, it would be useful to do serologies on these folks.

So run down the hypothesis of sustained human-human transmission.  Check for reports of influenza-like illness in Shenzhen.  Specifically for reports of medical clinics and hospitals reporting an increase in influenza-like illness.  Fatalities are only the tip of a much larger iceberg if we are dealing with an efficiently transmitting virus.

As a final point, Chinese officials should be thanked for reporting this important event.

Biosurveillance Targeting And Critical Information Requirements

Previously we discussed the difference between detection, recognition (which included prioritization of threat), and warning.

In this case, let's use the example of ProMED, which certainly delves more into the recognition aspect of the detection-warning cycle.  Here you have subject matter experts curating information from the firehose of input available on the internet on a daily basis.  They often provide excellent context and background information, forming collaborative partnerships with such terrific groups as GIDEON.  GIDEON is essentially a "living textbook" of information on infectious disease  worldwide.

But ProMED has had its detractors over the years, having adopted a now-archaic (by social media standards) form of community dialog.  ProMED is heavily moderated, which is a good thing when considering all the potential garbage that could hit the system.  However, it does introduce substantial time delays, which relegates the system to a forensic, after the fact operational posture.  Further, you will often see bias exhibited in displaying "what's hot and sexy" versus a particular subject matter expert's interest.  In other words, it is a mish-mash of things happening around the world without a clear sense of what is truly relevant and important.

But how does one establish relevance?  That is indeed the question, and the same one we dealt with in the DHS National Biosurveillance Integration Center (NBIC).  During that time, we observed that many agencies actually did not even have a prioritized list of issues they monitored on a daily basis.  The concept of critical information requirements from the military operational environment had not permeated the analyst watch cells we interacted with.  Of course, this created tremendous confusion regarding what to warn about on the back end.  ... And it took hours and hours of discussion for people to arrive at an integrated agreement on what pathogens take priority, particularly when having this discussion among physicians, veterinarians, public health officers, etc.

We would see CDC, for instance, obsess about an isolated outbreak of Ebola in Africa that was hundreds of clicks from the nearest airport and completely ignore reports of an unusual respiratory disease in southern China.  When dealing with limited ground verification resources, this poses quite the conundrum when our stateside first responders may encounter the infectious agent in question before warning is issued by a national authority, as what happened during the 2009 H1N1 influenza pandemic.  Certainly Ebola is far more "sexy" than influenza... but one is clearly more acutely and broadly disruptive to a society than the other.

So let's examine some examples:

• Anthrax in Argentina.  In this particular report, it is unclear why the outbreak was brought to the entire world's attention.  Does this outbreak exhibit unusual epidemiological features?  Does it exhibit any suspicious features?  Is it reflective of a concerning long term trend?  It would seem this kind of report should be tagged as a "routine event".
• Oseltamivir-Resistant Pandemic H1N1.  Now this is of casual interest, albeit not at the top of our list.  It is of course good for clinicians to know where we are in terms of drug resistance, but certainly not an unexpected phenomenon.
• NDM-1 Enterobacteriaceae In India's Water Supply.  This is a slow-moving threat of high impact, highlighting the challenge of prioritizing fast-moving, potentially acutely disruptive threats versus slow-moving insidious phenomena.  This is clearly an important warning to the world, however certainly of far greater immediate relevance to the people of New Delhi than here in the United States.  Of course, the definition of a crisis depends entirely on perspective.
• Influenza A/H3N2v In The United States.  This particular issue is at the top of our monitoring list.  But it is an interesting example because we likely would never have noticed this phenomenon just a few short years ago thanks to technology.  This would be an example of the odd dichotomy of better technology yielding greater social vulnerability to anxiety as a result of being able to percieve more threats.  The phrase "ignorance is bliss" of course is relevant here.  But that said, this is a top issue because of the opportunity to explore shortening of the time delta between detection / recognition / warning and response.  While we in the United States are singularly concerned with this issue, the people of Haiti would probably not see this as particularly concerning because they are still dealing with cholera and the myriad of other more morbid and mortal infectious disease endemic to their country.  Again the problem of perspective.

ProMED certainly has a problem trying to be all things to all people, with is a tar pit operationally.  In recent years we have seen multiple appeals for funding vis a vie the Jimmy Wales / Wikipedia Syndrome of proving a great service to the world, creating an expectation for that service, but not establishing a business plan to sustain that service before creating the social expectation in the first place.

Use of prioritized information requirements to structure the outputs is important, as it does create some degree of an edge to the reporting scheme.  It also tightens the screws on the plate of accountability, where issues detected, recognized, prioritized, and warned are more visibly displayed to the international community for their consideration.  But this kind of process requires another level of analysis by moderators and indeed, the user community.  This is ultimately important because we as a global community simply cannot react to all things all the time.

 

 

Biosurveillance: The Difference Between Detection, Signal Recognition, and Warning

For the last 15 or so years, we have seen an explosion of different electronic disease surveillance systems emerge.  Everything from laboratory diagnostics and reporting systems to syndromic surveillance to using the internet for so-called "event-based surveillance".  And there has been a virtual orgy of academic frenzy around the "bright lights and shiny object" effect of the technology.  This is representative of billions of taxpayer dollars over these years dispensed from a variety of federal mechanisms.

Yet, with all of this expenditure (much of which is ongoing), we have still so very little proof of operational value.  The "research and development ice cream cone" continues to lick itself without any apparent metric held by the funders for the operational "so what".  Arthur Reingold's challenge points from 2003 still appear to be unanswered ten years later.  Below, his points have been adjusted to fit a more broad consideration of the topic:

1. Demonstrating that [x, y, z type of] surveillance reduces morbidity and / or mortality following [an infectious disease crisis / disaster].
2. Types of [infectious disease crises / disasters] that [x, y, z type of] surveillance is likely to detect.
3. Response to apparent increases in illnesses signaled by [x, y, z type of] surveillance.
4. Plausibility that [x, y, z type of] surveillance can yield more timely identification of the population at risk in [an infectious disease crisis / disaster].
5. Likelihood that [x, y, z type of] surveillance will produce useful information about naturally occurring infectious diseases.
6. Circumstances under which [x, y, z type of] surveillance is likely to strengthen local and state health departments.

Certainly Reingold was not the only person during the heyday of syndromic surveillance to raise questions about the validity of the approach, but the gist of the above challenges were echoed several times in the literature.  But the academics lept onward undeterred, and the marketing associated with their programs grew to astounding levels of claimed accomplishments.

One such claim was that made by John Brownstein of HealthMap on Aug 3rd at an IARPA OSI Program Proposer's Day, who stated (paraphrased) that HealthMap was "the first surveillance system in the world to detect and publicly report the 2009 H1N1 influenza pandemic".  This impressive claim invokes scrutiny into the defintion of detection and how it applies to evaluating any disease surveillance system... and of course, the taxpayer funds associated with that program.

Merriam Webster defines the word detect as

1. to discover the true character of;
2. to discover or determine the existence, presence, or fact of
   

Webster's definition of recognize goes deeper than detect, however:

1. to acknowledge formally;
2. to acknowledge or take notice of in some definite way;
3. to perceive to be something or someone previously known

However it is the word's Latin origin, cognoscere that brings this point into focus.  Cognoscere means to know.  Of course the word know implies that a human being's mental processes have been engaged.  Which implies 1) there is a human being looking at the data and thinking about it. 

The definition of warning is provided by Webster as:

1. to give notice to beforehand especially of danger or evil
2. to give adminishing advice to
3. to call to one's attention

So this implies a human being, having thought about the data, has recognized and prioritized a concern... and communicated that specific concern to others.

The analytic team that warned the senior health intelligence officer in CDC's Director's Emergency Operations Center (DEOC) noted at the time that HealthMap's information was 1) not posted on their website as fast as the original Mexican source posting 2) was not highlighted in any unusual way to call singular attention to the information among dozens (hundreds?) of other colored pins on their map, and 3) where was no transfer of information to ProMED or CDC from HealthMap evident to suggest that human beings had communicated about the information during early to mid April 2009.  Therefore, the following may be concluded:

1. The first point of detection was actually the Mexican journalist who wrote the first article on the La Gloria situation.  HealthMap simply functioned as a Google News parser- a function anyone could have performed using Google Alerts.  If you happened to click that one pin on the map, you would have 'seen' the report, albeit then you would have to understand whether what was being reported was unusual...
   
2. Human beings were required to recognize unusual indicator patterns in the initial reports (n=1 to 3), not through recognition of simply volume of reporting as captured by HealthMap's Heat Index. 
3. Human to human communication clearly had to occur to push the warning through, as information provided via pull websites and push email notifications were inadequate due to daily information overload at WHO and CDC.

The point of the above is not to singularly criticize HealthMap but to make a point: collection of data is insufficient, by itself, to generate operational impact.  This was a very hard lesson learned by the syndromic surveillance community as well.  And if you are fighting an embedded, decades-old professional culture imbued with a "not invented here" bias (i.e. public health), it is unlikely the data will be exploited properly.

So let's give Brownstein a break for a minute and focus on one of his collaborators: the Canadian Global Public Health Network (GPHIN).  Brownstein's claim at the IARPA meeting was a duplication of GPHIN's claim to the world of having been the first public health system to detect SARS (see attached manuscript, Figure 1). However there was unverified rumor at the time that the web-harvested Chinese media article published on Nov 27, 2002 went unnoticed by GPHIN's staff for some time due to lack of machine translation for Chinese (unverified) and lack of a Chinese linguist.  Even if these rumors were false, and GPHIN's system immediately punched a warning out to its customers, there is no evidence anyone was paying attention.  In the multiple manuscripts and powerpoints given by GPHIN staff, not one contains copies of emails or phone records to indicate human beings had 1) recognized the problem, 2) prioritized the problem (out of the dozens of other reports for the day), and 3) communicated directly with the Canadian PHAC, the US CDC, or WHO specifically about this prioritized problem.  This is concerning when considering how much information was actually missed at the time (see attached report by Polyak et al).

The analogy is football: it's like being thrilled you threw a "Hail Mary" pass, but unawares that 1) no one caught the ball and 2) ran it to the end zone... 3) and thus you lost the game in the process.  Sound familiar?  It should.  These are very similar issues to those in the original 9/11 Commission Report.

Currently the federal government continues to chase its tail, spending more funds on biosurveillance research: the self-licking ice cream cone.  Billions of dollars and 10-15 years later, we still are not holding these systems' collective feet to the fire asking "show me don't tell me".  We continue to reach for the bright lights and shiney objects represented by the latest fad technology ("social media" being the current one) without thinking about how to promote the human process of understanding the data, accumulating knowledge, and using that knowledge to impact the potential outcome of the threat.

These are lessons currently highlighted in the recent IOM report on the DHS National Biosurveillance Integration Center (NBIC), however the problem is far bigger than just NBIC, DHS, or the people funded by DHS.  This is a biosurveillance community-wide problem that affects the credibility of the entire field. 

And if we do not fix it, the funding and public support may eventually disappear.

Download GPHIN Manuscript

Download Emergence of SARS in 2002

 

 

 

More Evidence of Human-Human Transmission of Swine-Origin Triple Reassortant Influenza A(H3N2) Virus In The United States

On December 23rd, CDC released another statement about the evolving situation with swine-origin triple reassortant influenza A(H3N2) virus in the United States.  Three more human infections, two of which associated with a daycare in West Virginia.  The latter is of greater concern than the former due to probable human-human transmission.  Time delays between submission of samples and public report by CDC remain high:

(Case A) Indiana, no evidence of human-human transmission- exposure from swine. 

• Day 0 (Oct 20th).  Patient symptomatic, hospitalized for 4 days.
• Day 2 (Oct 22nd).  Influenza-positive per in-house hospital laboratory testing.
• Day 8 (Oct 28th).  Indiana State Public Health Laboratory: "inconclusive influenza A virus".
• Day 11 (Oct 31st).  CDC confirms via genomic sequencing presence of "A(H3N2)v" (formerly known as S-OtrH3N2)
• Day 64 (Dec 23rd).  CDC publicly releases the above information.
• Summary:
  • Time delta: state public health recognition of unusual virus= 8 days
  • Time delta: clinical processing of patient to CDC awareness= 11 days
  • Time delta: CDC issuance of warning to the public= 64 and 56 days from initial clinical processing of the patient and state notification to CDC, respectively

(Cases B) West Virgina, evidence of human-human transmission.

• Day 0 (Nov 12th, approximate).  Pediatric patient recognized symptomatic by parents.
• Day 7 (Nov 19th).  Patient, who was already hospitalized for an unrelated issue, develops fever.
• Day 9 (Nov 21st).  Sample taken.  In-house hospital laboratory discovers sample is positive for influenza A.  Patient discharged home.
• ... No further information regarding when West Virginia's state laboratory or CDC processed the sample, which was A(H3N2)v-positive.
• Summary:
  • Time delta: state public health recognition of unusual virus= unknown
  • Time delta: clinical processing of patient to CDC awareness= unknown
  • Time delta: CDC issuance of warning to the public= 34 days from initial clinical processing of the patient (approx. Nov 19th) 

(Case C) West Virginia, same cluster / thread.

• Day 0 (Nov 29th).  2nd child at daycare center ill.
• Day 9 (Dec 7th).  West Virginia Office of Laboratory Services processes sample as "inconclusive"- forwards on to CDC.  This patient was part of a retrospective assessment, where clinical processing did not occur.
• .. No further information regarding when CDC processed the sample, which was A(H3N2)v-positive.
• Summary:
  • Time delta: CDC issuance of warning to the public= 34 days from state public health laboratory processing of the sample to CDC issuance of public warning

 

The key point is of course, we only know what we are able to know.  Our team leaned into the discussion in November to ensure CDC was aware we were 1) monitoring the situation and 2) are carefully documenting their performance in coupling warning to response.  And quietly behind the scenes we are engaging clinical practitioners in an effort to encourage more sampling among symptomatic patients who were vaccinated for this season.  We expect the predominance of this focus to be on pediatric patients, however we are not seeing social sensitization yet among pediatricians.  Messaging from CDC requires clarification and guidance regarding when to sample.

Currently, we are observing between large time deltas between clinical processing of these patients and CDC's issuance of public warning.  It is our opinion that if A(H3N2)v suddenly displaces current seasonal A/H3N2 or gains significant prevalence, the first point of recognition may be an abruptly overwhelmed pediatric medical infrastructure versus dramatically time-delayed reporting from CDC. 

The observations of public health performance during the 2009 influenza pandemic begin to repeat themselves...

 

Analysis of Warning Process: Swine-Origin Triple Reassortant Influenza A(H3N2) Virus In USA

We have been monitoring the situation with swine-origin triple reassortant influenza A(H3N2) virus (S-OtrH3N2) inside the United State for some time now.  Recent reports of likely human-human transmission sparked debate, discussed here in this post.

1.  First, the ECDC's risk assessment was an excellent discussion of the situation.  CDC's report may be found here.

2.  The time delays in issuing statements about the evolving situation reflect the forensic process and length of time it takes to recognize novel pathogens and execute public warning inside the United States.  In this case, a joint press releaseon Sept 2nd and 5th, respectively, from CDC and Pennsylvannia state officials announced recognition of S-OtrH3N2 pediatric cases in Indiana and Pennsylvania stemming from exposure to infected swine.  The breakdown of timelines as follows:

Case A (Indiana):

• Day 0 (July 23rd).  Parental recognition of a child with flu-like illness.  Child brought to local ED, clinical laboratory testing at the hospital positive for type A influenza.   Child found to have been vaccinated the prior fall, had chronic medical condition, discharged from the ED same day without anti-virals.
• Day 1 (July 24th).  Child returns to ED with a worsened clinical picture and admitted.
• Day 4 (July 27th).  Child does well and is discharged from the inpatient ward to home.  No further issues.
• Day 25 (Aug 17th).  Indiana State Department of Health Laboratories test the sample and discover suspect swine-origin influenza A (H3N2) virus.  CDC notified and sample forwarded to CDC for confirmation and further analysis.
• Day 27 (Aug 19th).  CDC confirms findings of the Indiana State DOH Lab. 
• Day 41 (Sep 2nd).  CDC reports publicly results of genomic sequencing: swine-origin triple reassortant influenza A(H3N2) virus that now includes segments of 2009 pandemic H1N1 virus.  Report indicates patient had contact with infected swine.
• Summary:
  
  • Time delta: state public health recognition of unusual influenza virus= 25 days
  • Time delta: clinical processing of patient to CDC awareness= 25 days
  • Time delta: CDC issuance of warning to the public= 41 and 14 days from initial clinical processing of the patient and state notification to CDC, respectively
    

Case B (Pennsylvania):

• Day 0 (Aug 20th).  Parental recognition of symptoms, take child to local ED.  Nasal swab positive for influenza A.  Child was vaccinated the prior fall.
• Day 3 (Aug 23rd).  Pennsylvania State Department of Health Laboratory idenifies suspected swine-origin influenza A (H3N2) virus.
• Day 4 (Aug 24th).  Pennsylvania state officials notify CDC of the findings.
• Day 6 (Aug 26th).  CDC confirms findings.
• Day 13 (Sep 2nd).  CDC reports publicly results of genomic sequencing: swine-origin triple reassortant influenza A(H3N2) virus that now includes segments of 2009 pandemic H1N1 virus.  Report indicates patient had contact with infected swine.
• Summary:
  • Time delta: state public health recognition of unusual influenza virus= 3 days
  • Time delta: clinical processing of patient to CDC awareness= 4 days
  • Time delta: CDC issuance of warning to the public= 13 and 7 days from initial clinical processing of the patient and state notification to CDC, respectively

Questions were asked openly of CDC, USDA, and Pennsylvania state officials on Sept 12th about how these cases were recognized.  More to the point, guidance for clinical sampling was requested publicly.  This request was never answered.  At this point, the situation had migrated to the top of biological threat monitoring list.  Lack of response was not reassuring.

On Nov 23rd, a probable human-human transmission cluster of S-OtrH3N2 cases was reported by CDC, with an evident 7-10 day lag between initial clinical processing, laboratory based recognition at the state level, and CDC issuance of a public statement.  This was a much better time delta and likely reflective of surveillance grid sensitization this particular issue. 

Overall, we see a drop from a time delta between clinical processing and CDC issuance of public warning of these events from 41 to 13 to 7 days as the situation evolved.  It was not until Nov 26th did media reports indicate a seed strain had been sent to vaccine manufacturers in case S-OtrH3N2 begins to overtake current circulating strains of influenza virus in the human population.

There are several summary observations to make at this point:

1. S-OtrH3N2 is our #1 pathogen of concern for the United States and the world at this time, from the perspective of potential social disruption.  Projected impact would be similar to that seen in a vaccine drift year of seasonal influenza (e.g. 2004 in the United States).  Thus far no fatalities have been observed, but the sample size is far too small at this time to make firm assertions of case fatality rate or morbidity profiles. 
2. CDC should be praised for proactively forwarding seed stock to vaccine manufacturers.  What will be interesting to observe at this point is the time delta between initial recognition of sustained human-human transmission and an order to ramp up and distribute S-OtrH3N2 vaccine to local healthcare providers.  In this instance CDC stands in much better position to execute more efficient warning to response operational coupling.
3. The above said, we note with great concern how long it took to recognize and effectively communicate to the public the initial situation in Indiana.  From the perspective of this author, a pediatrician practicing in a rural setting, there is no healthcare provider sensitization to this issue- i.e. there is no change at all in sampling procedures for children presenting with influenza-like illness.  It is as though this season is like any other.  Especially after the 2009 H1N1 pandemic experience and the current economic recession, we observe a general reluctance to test for influenza unless there is a compelling reason to do so.
4. Not all states have the surveillance capabilities of Pennsylvania and Iowa.  There is an incomplete understanding of how pervasive S-OtrH3N2 transmission is inside the United States at this time. 
5. Abrupt community inundation of pediatric flu-like illness, especially among children previously vaccinated for the 2010-2011 influenza season, should prompt an index of suspicion at least to obtain samples for testing by clinicians.  Submission of samples that are processed by state laboratories and moved up the chain to CDC is the only way to assess this evolving situation accurately.