Operational Biosurveillance

Dialogue Between a Meteorologist and an Operational Biosurveillance Analyst

Dialogue Between a Meteorologist [M] and an Operational Biosurveillance Analyst [OBA]:

 

[M] A forecast is a prediction of an event before it begins.  If the event has begun (e.g. a hurricane) at a given location, updates are called "nowcasts." For example, the initial wind speed forecast for hurricane Sandy at Battery Park might have been 60 mph. If the storm has started and the forecast is raised to 70 mph, that would be a nowcast. 

 [OBA] There are many in public health who claim forecasting cannot be done.  This is simply not true for a wide variety of infectious disease conditions.  Indeed we are discovering there are a multitude of health outcomes only peripherally related to infectious disease that can be forecasted as well.  Among those who are practicing physicians, this has been long captured in the classic discourse between physician and patient.  For example, "don't smoke or you may develop lung cancer".  The meteorological notion of a "nowcast" has been done routinely now for several years at the National Infectious Disease Forecast Center as well as some of our other engagements overseas such as Haiti.  "Nowcasts" are also performed by public health, however at nowhere near the depth of coverage as we are able to provide now via direct linkages to clinical medical data (which is a major source of public health advisories).  Public health advisories are but the barest subset of what is observed on a daily basis in clinical medicine and that is relevant to the Public Interest.

[M] Climatology represents the long term averages (in meteorology, usually 30 years or more). Climatology tells us that, in NYC, January, on the whole, will be colder than July.   Persistence is the tendency of the weather of tomorrow to be like the weather of today.  A forecast is considered "skillful" if it beats both "climatology" and "persistence".  If it is sunny and 75° today and exactly the same thing occurs tomorrow, even a perfect forecast is not "skillful" because there was no change in the weather.

[OBA]  Excellent point.  However what is considered state of the art in meteorology is not so in operational biosurveillance.  Indeed, the vast majority of healthcare providers operate in reactive mode, simply addressing disease one patient at a time and deciding whether or not to treat with antibiotics.  With the exception of influenza, most physicians have no idea what their local baselines are for rhinovirus, metapneumovirus, Salmonella, etc etc etc.  In other words, a lack of comprehensive understanding of one's own local baseline.  This unfortunately extends to local antibiotic resistance patterns, where we are clearly losing several important battles with resistance in our communities.

So, current state of the art in operational biosurveillance is a few exceptions representative of "skilled" forecasts (e.g. influenza vaccine drift) and a greater majority of "unskilled" forecasts.  However what is considered "unskilled" from the strict definition of meteorology is considered utterly "skilled" by your average healthcare provider and the public because they are unaware of their baselines in the first place.

[M]  We have previously discussed the concept of adopting the Watch / Warning terminology used by meteorology in the event a forecastable major adverse health event presents itself. If I recall correctly, watches by the National Weather Service are issued at longer time periods (~ two days out) with a 30% threshold. In other words, if the forecasters believes there is a one-in-three chance of an occurrence, a Watch is issued. If, as the start of the event grows closer, a Warning is issued when the confidence reaches 70% at roughly 24 hours out. 

An issue we will have to consider as our forecasts are adopted on a wide basis is, to what extent did our forecast cause a different outcome than what would have occurred naturally? Triple-digit tornado death tolls were relatively common prior to the creation of the tornado warning program. They are exceedingly rare now (even with a large population) because of the warning system. The forecasts have changed the outcomes (fewer deaths than would exist without the warning system). 

Assume we issue an "Influenza Watch." Our watch then prompts a proactive and effective vaccination program. Our watch may have the effect of surpressing the number of cases needed to "verify" (another word would be "validate") our forecast.  In other words, if we forecast 10% of the population will contract influenza and there is a subsequent vaccination program, and the actual number is turns out to be 6% was it because we made an inaccurate forecast or was it because of the vaccination program?

[OBA]  This is a crucial point.  Anecdotally, we have observed patients flocking to our pediatric practice requesting influenza vaccination because 1) we issued advisories to them months in advance of the forecasted drift season; 2) the bulk of our patient families support influenza vaccination; and 3) the community is small enough where "everyone knows each other" and therefore share stories of children who were unvaccinated and later hospitalized with influenza.  So, the pre-event forecast (i.e. well ahead of the first confirmed influenza case in the county) began the community orientation process, followed by discussion after discussion family by family in the clinic as they asked about the forecast, followed by that single case of a hospitalized unvaccinated child to show the "proof of threat" as well as "proof of preventive action".  We then issued "nowcasts" updating the public as to what we were seeing in the hospital: that the hospitalized were almost universally unvaccinated.  We observed several of the "vaccine refusers" reconsider their position in the context of this powerful local community influence.

All of the above said, operational validation does not equate to academic validation.  And we readily acknowledge the question of validation to be a tremendous challenge when prognosticating disease. Conversely, we have often pointed out that academic studies and peer-review does not often transition into the reality of daily clinical medicine.   There is tremendous challenge here. 

Only by building years of operational experience may we have the necessary data to address this challenge.

 

 

Progress in Operational Infectious Disease Forecasting in Haiti

In collaboration with several National Academy of Science members, we have completed an assessment of forecast signatures in Haiti for cholera.  The results are a promising beginning to anticipating endemic transmission patterns in Haiti to facilitate proactive preparedness and response.

Disaster Conditions Declared in Pestel, Grand Anse, Haiti

The Haiti Epidemic Advisory System (HEAS) is scrambling now to issue an Emergency Appeal, as cholera has most certainly reared its head yet again, this time in Grand Anse with 445 cases and 59 fatalities.

Observations Regarding the Implementation of the Haiti Epidemic Advisory System

We have successfully triggered a social network to support reporting of disease events in Haiti called the Haiti Epidemic Advisory System (HEAS). We now have 150 members with representatives from nearly every major NGO, smaller NGOs, unaffiliated clinicians, and key Haitian authorities. What we have executed is a system that mirrors a basic sociological process we have seen in many other operating environments to support warning functions such as:

• The World Health Organization IHR reporting system (official, publicly visible) versus ProMED, the Global Outbreak Alert and Response Network (GOARN), and the now-defunct World Health Organization Outbreak Verification List (OVL) systems (unofficial, offline).
• The National Biosurveillance Integration Center (official) versus the BIWAC/Wildfire system (unofficial, offline) In other words, we continue to observe a very important sociological phenomenon that relates to the process of getting people to routinely report information- specifically, important public health information – that would not be reported under formal networks.

We are still in crisis mode in Haiti.  As such, the challenges for effective formal medical surveillance remain.  Others have published outlining the need for alternative reporting systems for disease events, again highlighting a basic sociological phenomenon.

Especially as it pertains to emerging crises, we have found over our 12-year career that “tip” information regarding infectious disease events is often received through trusted one-to-one communications. When this one-to-one communications are coupled with well-architected, intelligent social networks, a new, more robust type of situational awareness and collective intelligence emerges that can drive agile, proactive, rapid response in emergent situations.

This is absolutely a complimentary process to formal public health reporting, as currently facilitated by the Haitian Ministry of Health, PAHO, and CDC in Haiti.  Metrics for success when constructing a warning system are always so difficult to define. But we do have two important ones:

• People are reporting disease events in a way that supplements reporting in the formal medical surveillance system.
• Several of the major NGOs have proactively moved medical countermeasures and logistics into position in anticipation of a possible infectious disease crisis – as a result of advisories sent to them from the HEAS community effort.

We are currently in emergency mode monitoring for a potential crisis of pediatric diarrheal disease.  Unfortunately, the same cultural resistance to change in the public health community is present in our attempts to properly fund and maintain the capability: 

"[thanks for your] detailed follow-up message, which has been reviewed by the... health teams in Haiti. They appreciate the interest but, for now, find the current procedures satisfactory to their objectives."

As mentioned before, this same sociological process delayed public benefit of weather forecasting by decades.  In the case of Haiti, it is actively impeding our attempts to mitigate large scale child mortality with the advent of the rains.  Unfortunately in this case, the risk relates not only to the tragedy of children dying in Haiti after there was a major loss of life due to the earthquake, but also represents an important contributing factor to destabilizing the region.  We have seen in other cultures that unexpected fatalities involving children- particularly if perceived to be preventable- is a key trigger for social outcry.  In the context of the current situation in Haiti, where the public is already aware of a mismatch between donations and ground implementation of those monies, there is a very real possibility of civil unrest.

Warning systems are contentious constructs.  The reaction two nights ago at the UN Health Cluster to a recent posting of an HEAS advisory related to malaria is revealing.  Comments by public health professionals were indignant, saying "we already knew that".  The disconnect is the translation of awareness to proactive action.  In the case of malaria, there have been robust discussions with the MENTOR Initiative about malaria surveillance, control, and emergency preparedness.  MENTOR is an excellent effort that has been proactively engaged in dialog with the HEAS community, very much to their credit and to the benefit of the community.  However, pediatric diarrheal disease is more likely to abruptly overwhelm the already fragile medical infrastructure and therefore more likely to fall in the range of being labeled a crisis or disaster.  It is therefore more likely to contribute to social destabilization.  

Translation to action is a difficult process.  The issuance of a warning is an important step in sensitizing a community who internalize that information, cross-check the validity of the warning, and then decide whether the information is relevant to their operations or not (i.e. confirmatory behavior).  Another way to view confirmatory behavior is to consider it a part of active surveillance.  We often see the "all is clear" type reporting through this process, which enables detection of negative as well as positive signature patterns.  Proof of successful translation to action was demonstrated by the above-mentioned metrics for success:

• People are reporting disease events in a way that supplements reporting in the formal medical surveillance system.
• Several of the major NGOs have proactively moved medical countermeasures and logistics into position in anticipation of a possible infectious disease crisis – as a result of advisories sent to them from the HEAS community effort.

The more troublesome aspect of situational awareness-driven warning systems is they promote transparency.  In other words, it is a vehicle for accountability.  Receipt of a warning automatically places the recipient in a position of accountability.  If they act to mitigate a crisis before it happens, then the recipient and those they server are pleased if the crisis indeed manifests (or no social outcry is observed that challenges the credibility of the recipient organization).  If the recipient does not act, and a crisis occurs when their receipt of the information was known broadly, then the question of "why did you do nothing" appears.  For a bureaucrat, this places them directly in the line of sight, a most difficult position to be in when risk-averse.

Information Sharing In Disaster Response: Towards An International Standard

Wired.com mentioned our efforts in a recent article that quoted USSOUTHCOM: "they are poised to integrate some crowdsourcing to see what the situation is on the ground with respect to medical information. Eventually, that kind of tool could be used for biosurveillance –- monitoring and mapping the outbreak of infectious diseases."

The 2010 Haiti disaster response effort was my first time interacting with a US federal agency in such an open framework as that offered by APAN.  I was introduced via our participation in Crisis Mappers and collaborative discussions with InSTEDD and Ushahidi.  Although I appreciate USSOUTCOM's desire to remain conservative in their assessment of APAN's utility, from the perspective of operational biosurveillance it is frankly a breath of fresh air.

Two key observations that are relevant to the Wired article (and why it is so important to recognize Crisis Mappers community and APAN as setting the new world standard in crisis communications):

1. The bigger the organization, as embodied by funding lines, personnel, and complexity of task it has claimed, the less likely it is able to participate effectively in near-real time operational biosurveillance communications.  This is especially true when considering the ultimate objective of issuing warnings.
2. As the emergency response community has always emphasized, the need to understand which individuals are gatekeepers versus bridges is critical. 

Currently the best and most timely information we receive from Haiti is from individuals, the majority of whom work in smaller, more agile organizations.  We avoid gatekeepers.  One multi-billion dollar US NGO flatly declined to participate in information sharing, claiming "we already know there are outbreaks in Haiti".  Technically speaking, at the time of the comment there were no outbreaks reported in Haiti, just sporadic, expected reporting of baseline disease.  This particular NGO has a history of little to no cooperation with other NGOs, a common problem in the NGO world.  But this is not to pick on a particular organization or community of organizations- this is a human problem related to trust and other factors that dictate whether one individual(s) will chose to share information with another.

The challenge of managing crisis information flows was highlighted brilliantly by Uriel Rosenthal, Paul 't Hart, and Michael T. Charles (1989): 

• In crisis situations, formal rules and procedures give way to informal processes and improvisations.  Necessity is the mother of invention.  Under the pressure of the circumstances, official authorities may be overruled by situational leaders.  Formal authority loses out to legitimate power.  For making critical decisions, official qualifications and jurisdictions happen to be less important that the capacity for a resolute and rapid response to the crisis.  This is precisely what Crisis Mappers and APAN represent, as well as the entire warning process of the 2009 pandemic H1N1 influenza from our perspective.

• Public agencies can be viewed as information processing systems... They tend to be disposed to consume information that is easy to categorize, leaning towards so-called programmed decisions.  Here I would propose "public agencies" in this era includes both governmental and non-profit organizations

• Crisis information does not fit in with this pattern.  Crisis information is threatening, it is frightening, it is unfamiliar, it is new, it is surprising and it has 'never been heard before'.  Crisis information is unprecedented, and it is critical, yet ambiguous.  It may come in massive quantities, yet 'just that damned thing we must know, we do not' (at other stages during the same crisis, lack of new data, subsequent rumor circuits and other forms of non-information may produce another and often unthought of challenge to decision makers).  It cannot be categorized in usual ways.  Nonetheless, crises demand rapid information processing, but also very careful information processing.  Under these circumstances the usual patterns of organizational and interorganizational communication do not suffice.
  

• In crisis situations there is a considerable increase in the volume and speed of upward and downward communications... High level officials directly communicate with low-ranking bureaucrats.  Key decision makers run the risk of being overwhelmed by an overload of information from sources of which they are unfamiliar.  

• In crisis situations, on the other hand, decision makers tend to give priority to the source of information instead of its content.  They are inclined to rely on trusted, and liked sources.  In conflict crises they may become completely consumed in closed communication circuits made up solely of allies, adherents, and friends.  This point is dead-on.

• Paradoxically enough, potentially rewarding channels of communication are often closed off by strategies or tactics that are destined to isolate the opponents.  In such cases it may become impossible to acquire accurate information about the most crucial aspects of the crisis.

With the tremendous variety of international assistance being provided by so many different professional disciplines, organization types, and governments, action certainly speaks loud in terms of coming together to help others.  But the issue of open information sharing is such an important one, and one we must continue to push on.  The unfortunate fact remains there are several major organizations responding in Haiti now, flush with funding, who are not participating in situational awareness.

Participation in situational awareness is absolutely essential for the following reasons:

• rapid identification of events that may result in additional loss of life or further disruption in the response theater
• rapid identification of victims' needs not otherwise identified 
• improved response logistics, efficiency, and cost

I would propose it be considered a mandatory international standard for any responding organization or individual to share situational awareness if they intend to participate in a disaster response theater.  Donors and taxpayers should insist on this standard as well. 

References:

Rosenthal U, 't Hart P, Charles MT.  The World of Crises and Crisis Management.  In: Rosenthal U, 't Hart P, Charles MT, eds Coping With Crises: The Management of Disasters, Riots and Terrorism.  Springfield: Charles C. Thomas; 1989:3-33.

Official Versus Unofficial Reporting in Operational Biosurveillance

Current experience with the Haiti disaster response has provoked memories of the early days of the 2009 H1N1 pandemic, where unofficial reporting mechanisms preceded official reporting.  It is clear the work of the unofficial human "mash-ups" like Crisis Mappers Net, with partners like Ushahidi, InSTEDD, HealthMap, and many others are showing a new, agile mechanism for gathering and reporting crisis information.  Right now we are observing tremendous chaos and delays in reporting information from the larger NGOs and official organizations who are on the ground in Haiti- where the bigger and more beaurocratic the organization, the more difficult it is to coordinate information flow in an environment where agility is becomes most important.  This is the classic hallmark of a disaster, where the demand for information is tremendous... a demand at odds with the very nature of a cumbersome beaurocratic entity.

Below is a chart of unofficial versus official reporting in the early days of the 2009 H1N1 pandemic.  Even inside the United States there were tremendous time differences between unofficial reporting of suspected cases and official reporting by CDC, which was reflective of lack of available test kits for pH1N1, the hierarchical nature of reporting at the local, state, and national levels, and the methodical process of forensic epidemiological investigation.  When considering the interface of national level reporting to WHO, the same kinds of time delays were also observed.

The numbers in the left and right-hand columns represent the difference in days between first report of suspect cases of H1N1 and first report of laboratory confirmation.

It has been said here we are witnessing the birth of a new professional discipline in the realm of operational biosurveillance... a discipline that places a higher premium on unofficial reporting because the output of this community is timely, relevant warning of biological threats to society.

As we are observing right now in Haiti, the phenomenon of time delays between unofficial and official reporting is certainly not unique to operational biosurveillance.

Panic and Flight Behavior Reporting During Infectious Disease Events, Crises, and Disasters

The idea of people “panicking” due to disaster is a misnomer.  The literature repeatedly shows the normal response is to bond and protect loved ones- to the point of even placing individuals in greater harm. 

 

Assertions of the sociological and anthropological community are this phenomenon is universal for disasters of any etiology- including epidemics.  However, a literature review of studies supporting the assertion that social behavior in the context of epidemic-caused disasters is similar to that seen in natural disasters is limited.

 

We undertook a review of ProMED reports for the ten-year period January 1, 1998 to December 31, 2007.  Known biases in ProMED during this period included a bias towards English language reporting on human health events.  The results were interesting.  Only 1.7% of the reports in this period covering infectious disease events affecting humans or animals contained the word “panic”.  Of these “panic” reports, we noted the following:

 

1.     Concern about the potential for “panic” (61%)

2.     “Panic” reported with no behavioral evidence (27%)

3.     “Panic” reported with behavioral evidence (12%)

 

Bottom line, when “panic” was reported in association with an infectious disease event, it was a rare phenomenon.  Evidence of actual anxiety behavior associated with the word “panic” was even more rare: 0.2% of all reports during a ten-year period.

 

The concept of “panic” as people behaving like asocial animals is a myth.  All of the instances of anxiety behavior reported in association with “panic” were protective behaviors. 

 

To dive even deeper into this, we explored the number of times flight behavior was reported in ProMED during the same ten-year period.  By “flight” we were specifically looking for public-initiated evacuation or abandonment of their community or family members.

 

In total, only 0.1% of all reports contained mention of flight behavior where public-initiated abandonment of a community or family members due to an epidemic was reported. 

 

Family member abandonment was documented in one instance: Marburg hemorrhagic fever in Uige, Angola.  On the date of this particular report, May 4, 2005, there were 313 cases and 280 deaths (89% fatalities) reported the prior day. This was more than six weeks after the discovery of the first outbreak of Marburg hemorrhagic fever ever reported in Angola.  A family reportedly abandoned a one-month-old baby whose mother had died of Marburg infection and fled the village.  There was no indication the family abandoned any other family member.  Neighbors feared contact with the baby and allowed the child to starve to death.  The context of this report was one of epidemic resurgence involving substantial struggle with effective containment.  Fatalities were reported on a daily basis.

 

In summary, The majority of “panic” and flight behavior reporting involved zoonotic diseases associated with causing active serious illness in humans at the time of reporting.  Specifically, provocation of community and family abandonment involved diseases that, at the time of reporting, caused serious and lethal human disease.  All reporting of community or familial abandonment occurred in undeveloped or developing countries.  Of note, no reporting of “panic” or flight behavior was documented for Haiti in this ten-year period.

 

Despite the known biases in ProMED during the period of study and the fact all global infectious disease events during the time period studied were not reported in ProMED, we have found it to be highly analogous to our operational experience. 

 

When considering the current critical information requirements for the Haiti disaster, it is our hope this information places “panic” and flight behavior reporting in perspective- it is a very rare phenomenon, but when documented usually represents a situation worthy of time-sensitive awareness and scrutiny.

 

 

NOTE: New Website Link For Haiti Operational Biosurveillance

In order to facilitate a dedicated website and output for disaster response in Haiti, we will move all further postings on Haiti to "Haiti: Operational Biosurveillance".  Please update your bookmarks and RSS links accordingly.

Arrival of 53 Haitian orphans in Pittsburgh: All In Good Health

The arrival of 53 orphans from BRESMA orphanage in Port-au-Prince were examined by physicians in Pittsburgh today.  All were in good health.  Some children had mild respiratory illness considered common and routine for their age.  Pittsburgh is the first city in the United States to receive orphans from Haiti. 

Haiti: Prioritized Critical Reporting Requirements and Indicators For Operational Biosurveillance

Geographic Scope: Hispaniola Island, inclusive of the countries of Haiti and Dominican Republic

Emphasis: infectious disease capable of generating acute disruption within the response theater

Specific Diseases of Interest: anthrax, dengue, diarrheal disease, pandemic influenza

Event Indicators:

1.     Report of public-initiated evacuation, migration, or flight from habitation site due to perception or rumor of disease

2.     Report of public “panic” due to perception or rumor of disease

3.     Report of medical infrastructure, either indigenous or response agency-deployed assets, becoming acutely overwhelmed due to disease

4.     Report of abrupt increase in disease prevalence, especially involving an efficiently transmitting agent such as influenza

5.     Report of acute disease affecting animals potentially serving as protein sources to the indigenous population such as cattle or poultry

See prior post, Haiti: Operational Biosurveillance Priorities For Potentially Disruptive Infectious Disease Events, for background information.