The Haiti Epidemic Advisory System (HEAS)

Lymphatic Filariasis in Haiti: Translocation Implications For The United States

In a previous post, we identified possible increases in filarial transmission in Haiti.  At this time, it remains unclear if this is acute lymphatic filariasis, a true outbreak, or reporting bias following sensitization to look for the disease.  Sources have indicated Haiti may be entering the beginning of a cycle of increased transmission activity- this requires confirmation.  Particularly in regards to humanitarian workers and indigenous Haitians returning or visiting the United States, the potential to see introductions of acute lymphatic filariasis should be entertained. 

Melrose and Leggat (2005, see attached) highlight cases where substantial infection and clinical manifestation of acute lymphatic filariasis was observed in military forces serving in endemic areas.  Estimates of exposure in the Pacific Theater during World War II, for instance, suggested over 38,000 naval personnel were exposed, with over 10,000 diagnosed with the disease.  During a deployment in Samoa, 70% of exposed personnel were infected.  In one impressive example, more than 2,500 servicemen were medically evacuated to Klamath Falls, Oregon for treatment over a 17-month period.  Melrose and Leggat provide examples of infection rates among deployed personnel where, in one case, rates were high enough to withdraw entire units from service in the middle of the war to the US for treatment.  

While elephantiasis is a common visual presentation of lymphatic filariasis that comes to mind to most clinicians who have had little experience with the disease, diagnostic criteria for acute lymphatic filariasis is more broad and non-specific (from Melrose and Leggat per their reference to Hodge et al, the manuscript of which is not found in PubMED):

Involvement of the genitalia

• episodes frequently follow periods of manual labor
  
• testicular pain is the most common complaint the pain may radiate up the spermatic cord or may appear first in the lower abdomen and radiate down the spermatic cord to the testicle
  
• the spermatic cord and epididymis are thickened and indurated but the vas deferens is not involved
  
• acute hydrocele is often present
  
• pain in the medial aspect of the thigh on the affected side may be present
  
• scrotal edema is often present
  

Involvement of the superficial lymphatic vessels of the extremities

• pain in the extremity radiating distally is the usual complaint
  
• pain may be noticed first in the proximal lymph nodes
  
• the affected extremity becomes edematous
  
• a definite "red streak" characteristic of acute lymphangitis appears proximally. The affected lymphatic vessel is palpable, and usually only a single vessel is involved
  
• there is increased local heat
  
• the lesion extends centrifugally
  
• resolution begins proximally and extends centrifugally the lymph nodes draining the affected area are enlarged and tender
• severe constitutional manifestations are usually absent
  

A recent review by Manguin et al (see attached) have highlighted the potential for co-transmission of malaria and filariasis leveraging the same mosquito vectors, a compelling thought when contemplating the evolving phenomenon of dengue re-emergence in Florida and the Gulf states and potential for re-emergence of malaria and/or lymphatic filariasis.  Malaria cases have been identified in the US, with occasional instances of autochthonous transmission (as with dengue).  Evidence suggests transmission of filarial worms by mosquitoes is an inefficient process compared to malaria, and infection requires prolonged exposure to filariasis-endemic areas.  It is more likely to observe translocation of single cases from Haiti versus autochthonous transmission in the US.  However given the difficulty in sensitizing clinical suspicion to adjust differential diagnoses, current access to treatment provided by CDC (see prior post), and lack of preventive prophylaxis in the US, we propose identification of translocation and autochthonous transmission may be challenging, especially in local unsensitized communities within the US. 

Download Lymphatic Filariasis- Disease Outbreaks in Military Deployments From World War II

Download Co-transmission of malaria and filariasis

Forecasters Predict Above-Average 2010 Atlantic Hurricane Season

Forecasters at Colorado State University predict above-average tropical storm activity for the 2010 Atlantic hurricane season based on recent, unexpected warming of Atlantic tropical sea surface temperatures and a likely dissipating El Niño event.

On April 13th, hurricane forecaster William Gray released a statement noting concern regarding the expected above-average tropical storm activity over the Atlantic this coming season, noting a 58% chance of a major hurricane tracking into the Caribbean.  Haiti, located in the Caribbean, is highly vulnerable to tropical storm activity due to extensive deforestation and soil erosion, in addition to post-earthquake conditions for internally displaced persons.   More information pertaining to hurricane risk factors for Haiti can be found by accessing our February 16th Anticipatory Assessment.

Malaria Transmission in Haiti

Cholorquine-sensitive P. falciparum is the predominant strain (>90%) in Haiti and is transmitted primarily by the Anopheles albimanus mosquito. 

A. albimanus breeds in hoof and footprints, ditches, and rice fields.  More than 80% of rice fields in Artibonite Valley sampled contained larvae in a survey by Caillouet, Keating, and Eisele in November 2006 (see attached):

The minor rainy season begins in March, peaking in May.  The major rainy season begins in August, peaking in October.  Transmission of malaria is seasonal in Haiti, peaking typically in November:

 

Transmission has been reported to be localized to certain villages and intense.  A seroprevalance of 3.1% was recently determined by Eisele et al in November to December 2006:

Malaria is a current concern in Haiti due to the devastation post-earthquake, displaced and densely crowded population, and the coming minor rainy season.  Peer-reviewed literature has suggested outbreaks of malaria in post-disaster conditions is relatively common.  Falciparum malaria infection, if chemoprophylaxis is not taken before and during visits to Haiti by responders, is a serious risk for adverse clinical outcomes.  However the strain of falciparum in Haiti is chloroquine-sensitive and thus easily treated.  Although some outbreaks of malaria may occur among Haitians in the next 90 days, it is unlikely to become a crisis unto itself; the latter half of the year holds greater risk of transmission to larger numbers of indigenous people.

Download Mosquitoes in Artibonite Valley, Haiti

Download Prevalence of Malaria in Artibonite Valley

Potential Increased Security Risks In Haiti Associated With Upcoming Rainy Season

For this anticipatory assessment the Atlantic hurricane and Haiti rainy seasons for the sixteen-year period from 1993 to 2008 were analyzed. ORE (Organization for the Rehabilitation of the Environment) has been keeping precipitation records for Camp Perrin, Cayes Haiti since 1993. Measurements indicate the months of April and May received the highest levels of precipitation for the calendar year five independent times, or 31.3% of the time during the sixteen-year period. Of the remaining eleven rainy seasons where the months of April and May did not record the highest precipitation levels, a hurricane or tropical storm made landfall or traveled to within 200 kilometers of Haiti five times resulting in excessive precipitation accumulation during June 1st – November 30th, the typical months of the Atlantic hurricane season. Of note, the average high precipitation count in Camp Perrin during the sixteen-year period was 590.8 mm. During the six independent rainy and hurricane seasons where April and May did not record the highest precipitation levels and no tropical storm impacted Haiti, the highest monthly records all fell below the average for the sixteen-year period.

Analysis of open media reporting of floods in Haiti during the past sixty years revealed only a handful of flood-related events that reached the attention of international media. Of those events reported, limited details pertaining to environmental conditions and geographic locations were provided; however, in the past decade there has been documentation of flooding in Mapou, the Coquillo Nazon district of Port-au-Prince and Camp Perrin resulting in residents fleeing and local infrastructure damage.

The significance of this information to survivors of the January 10th earthquake and responders on the ground is a one-in-three chance of experiencing year-high precipitation accumulations during the upcoming months of April and May and an equal chance of mass precipitation falling as the result of a tropical storm or hurricane later in the year.  Points of concern include the low-lying areas of Haiti that could collect floodwaters. Port-au-Prince alone holds multiple floodplains, which can be observed through unclassified satellite imagery made available by the United States Army Geospatial Center. Of note, we have not seen such analysis of other areas of Haiti such as Gonaives and the southern peninsula. Extensive deforestation has left the country susceptible to greater risk of flooding and landslides. If floodplains or similar low-lying areas are providing temporary camps for survivors or transportation routes for individuals on foot, the risk of waterborne infectious disease activity increases. Moreover, security for earthquake survivor populations and responders on the ground is a concern under flood conditions based on previously recorded violent social unrest triggered by flooding.

In late September 2004 after Hurricane Jeanne brought excessive precipitation to Haiti, the city of Gonaives was flooded leaving thousands of residents without access to food, clean water or shelter. International relief agencies were able to deliver supplies to the affected population within a week of the storm’s impact; however, the quantity of initial supplies provided was insufficient. The scarcity of goods resulted in rioting among residents awaiting aid. United Nations troops were on-site in Gonaives to facilitate distribution of supplies. There is documentation of warning shots being fired by U.N. troops at distribution sites and the hiring of local gang members by U.N. officials to provide additional security during the distribution of aid to residents. The historical record of social unrest in Haiti compounded by the current multi-infrastructure collapse due to the recent earthquake leaves the local population highly vulnerable to additional rioting triggers. High levels of precipitation and potential flooding expected to occur in nearing weeks may prompt the need to employ further security and relief aid resources.

Sources:

http://www.oreworld.org/rainfall.htm

http://csc-s-maps-q.csc.noaa.gov/hurricanes/

http://weather.unisys.com/hurricane/atlantic/ http://svs.gsfc.nasa.gov/vis/a000000/a002600/a002640/

http://lcweb2.loc.gov/cgi-bin/query/r?frd/cstdy:@field(DOCID+ht0064)

http://www.scidev.net/en/editorials/haitis-lessons-for-managing-the-global-environmen.html

http://www.timesonline.co.uk/tol/news/world/article1076128.ece?print=yes&randnum=1151003209000

http://www.wunderground.com/hurricane/at200410.asp

http://www.nytimes.com/2004/05/30/weekinreview/page-two-may-23-29-in-the-caribbean-a-fatal-flood.html?pagewanted=1

http://www.planetark.com/dailynewsstory.cfm/newsid/30737/story.htm

http://www.guardian.co.uk/world/feedarticle/8521500

http://www.agc.army.mil/haiti/pdf/PortAuPrince_InundationAnalysis.pdf

Gonaives- Hurricane Jeanne (2004): Investigation of Arboviral and Waterborne Disease Transmission

(See attached) Beatty et al investigated the transmission of mosquito- and waterborne illness in Gonaive, Haiti following Hurricane Jeanne.  The time period of study was November 15 to December 22, 2004, which is within range of dengue seasonal transmission. 

Hurricane Jeanne was gathering strength when it hit Hispaniola Island, only becoming a Category 3 hurricane after passing over the island to later become the deadliest hurricane of 2004.  In reviewing the records of what happened, it is not clear whether it was a hurricane or a tropical storm when it released  330mm of precipitation in the Haitian mountains that later flooded the Artibonite Valley and Gonaives.    Mudslides were documented.  A total of 80,000 were estimated to be affected out of 250,000 inhabitants in Gonaives, which is the third most population-dense area in Haiti.  By October, it was believed 3,000 people died in Gonaives alone.  Most of the houses were huts that were under ten feet of water at street level, displacing 200,000.  One hospital was reported to be knee-deep in mud.  The most vulnerable and poorest neighborhood in Gonaives, Raboteau, reportedly flooded even in light rain and was badly affected.  Livestock and crops were washed away, a disaster the region is still trying to recover from to this day.  Other towns were affected such as Port-de-Paix, Gros Morne and Chansolme.  It is noteworthy that floods the prior May had already killed 3,000 in Mapou and other areas of southwest Haiti.

The most common clinical diagnoses they encountered were acute respiratory illness (30%), malaria (29%), pneumonia (18%), and typhoid (11%).  They diagnosed 2 cases of dengue and found high pediatric seroprevalence, indicating prior community transmission.  A 13 year old child and an infant were found to have dual infections with acute malaria and West Nile virus.  This became the first documentation of active West Nile virus in Haiti.  Curiously, no location-specific infectious disease baseline was reported or discussed in detail to enable contextualization of the results.

The authors concluded, "An outbreak of mosquitoborne disease was not detected during the period of surveillance... Our data are consistent with previously published reports, which indicate that the incidence of arboviral infections rarely increases after water-related disasters... However, malaria outbreaks are common in such settings." 

This highlights an important point of clarification in our assessments that will be updated in the next SitRep.  When we report anticipated increase in risk of transmission for dengue, it is an expression of an endemic seasonal increase, not necessarily due to any additional risk caused by the earthquake.  We will probe this issue in greater depth in future posts.

Download Hurricane Jeanne

2010 Atlantic Hurricane Season Will Increase Infectious Disease Risk in Haiti

The onset of the 2010 Atlantic hurricane season on June 1^st increases the risk of infectious disease activity among earthquake-affected populations in Haiti.  A hurricane making landfall in Haiti will bring with it mass precipitation resulting in flooding and stagnant water collection.  Saturation of grounds supporting temporary shelters for earthquake survivors will intensify already acute sanitation problems leading to increased exposure of persons to human and animal waste products.  This presents an ideal trigger for increased diarrheal disease occurrence, in addition to water-borne illnesses including leptospirosis and typhoid.  Similarly, flooding of grounds previously under drought conditions may increase the risk of anthrax infection due to the unearthing of Bacillus anthracis spores if present.  In 2009, Sud Department located within 80 kilometers of the January 12^th earthquake epicenter and subsequent aftershocks was reported to have been experiencing drought conditions.  Furthermore, stagnant water collection may serve as mosquito-breeding habitats when temperatures are fully optimized in July, increasing the risk of dengue transmission.  Increases in mosquito counts will also increase the risk of malaria transmission.  Should there be a subsequent dramatic increase in transmission of a disruptor such as diarrheal disease or dengue, logistics to mount a coordinated response could be compromised if roadways are washed out.

In the past twenty Atlantic hurricane seasons, 145 hurricanes of varied intensity have been recorded over the Atlantic Ocean, of which 9.3% (n=12) have made direct landfall in Haiti.  This data excludes tropical depressions and tropical storms that did not escalate to hurricane status, but may have still brought mass precipitation to Haiti.  Given the current El Niño event underway, the elevated temperatures of the Pacific Ocean should serve to curb the frequency and intensity of storms forming during the 2010 Atlantic hurricane season.  During the two previous El Niño years, 2006 and 1997, Atlantic hurricane activity was recorded as less than average.  As reported by the United States National Hurricane Center in 1997, tropical storm development is hindered during an El Niño event due to strong westerly winds that travel across the tropical Northern Atlantic Ocean and prevent formation of thunderstorms, which serve as precursors for hurricanes.  Despite these conditions, some hurricane forecasters are predicting an “above-average” Atlantic hurricane season in 2010.

Prior to the 2010 hurricane season, steady rains are expected to begin in March, which will serve to promote infectious disease risk factors such as diminishing sanitation and increasing mosquito populations.  These conditions will be further exacerbated if a hurricane makes landfall in Haiti.  As the next hurricane season approaches, we will be monitoring oceanic-atmospheric activity in the Atlantic region for potential indicators of tropical storm development.

  

Social Anxiety, Outcry, and Unrest in Haiti: The Biosurveillance Perspective

Ground personnel in Haiti have worried about the potential for social outcry, rioting, and the dissolution of social order as a result of the earthquake.  As has been robustly documented for years by the disaster sociology community, the tendency of communities is not to descend into to total anarchy, asocial behavior, or complete dissolution of society.  The response of the Haitian public has followed those observations- there has been lower overall violence compared to before the earthquake, isolated incidents of social tension around food distribution points, and generally an overall tone of stunned graciousness in the people.  The vast majority of reporting indicates attempts to maintain familial ties, support neighbors, and individual actions motivated to protect neighbors, friends, and family.

There is a social dynamic to be aware of here as well.  The centuries-long political and social tensions between the people of Haiti and European colonialists.  Those tensions still exist to this day in a spiritually independent Haitian society.  That said, we are seeing some of those tensions dissipate as the Haitians welcome US military support during the crisis... a window of hope for aid to Haiti.

Here is where the nexus of well-intended support and indigenous social expectation becomes a serious threat to the credibility of response operations in Haiti... and the future of Haiti's socio-economic relations with the rest of the world.  If social expectations are perceived by the Haitian public to fail, the risk of social outcry and potentially conflict will increase.

From the perspective of operational biosurveillance, there are several specific triggers for social outcry that we have observed over years of experience.  Under the right conditions, these triggers are leading indicators for social outcry and civil unrest:

(Pre-Condition) A community that has developed an expectation for access to a certain level of public health or medical care.  This has become the case in Haiti, where the UN Health Cluster, the US military, and major NGOs now are the publicly visible "point" for healthcare in Haiti.  It is strongly debatable whether the Haitian public views the Ministry of Health as the lead authority.  The focus of social outcry to-date has been directed at these organizations, not the Ministry of Health.

Perceived threat of infectious disease.  Documented as indigenous social anxiety in SMS traffic and multiple reports from ground.  This began with public concern over unburied corpses, then with the appearance of flies and vermin, followed by public awareness of disease (particularly pediatric) in their camps.

Death of children due to infectious disease.  Documented and clearly a source of concern to the Haitian people.

Death of newborns.  Documented, but unclear what role infectious disease played.  Clearly a source of concern to the Haitian people.

Death of pregnant mothers.  Not clearly documented to our knowledge but likely.  Unknown in regards to infectious disease etiology.

Perceived increasing threat over time.  Documented as indigenous social anxiety in SMS traffic and multiple reports from ground.  Public perception has been the rainy season brings infectious disease threat specifically to Haitian children.  Sporadic rainfall and public perception of increasing diarrheal and respiratory disease in children observed to-date has reinforced this anxiety.  Note this anxiety has been amplified by concern for access to adequate shelter, which is already appearing as a focus for social outcry. 

Perceived failure to control or mitigate the threat environment.  There is evidence to suggest public confidence is wavering, as evidenced by comments made by the Haitian public to the media.  The ground vaccination programs have been delayed, and it is questionable whether the children of Haiti will be reached in time before the rains.  Diarrheal and non-pandemic H1N1 respiratory disease will likely pose serious challenges once sustained rains come, with mosquito-borne disease to follow several months thereafter.  It is unlikely these challenges that are associated with challenging and complex public health response will be effectively implemented before the onset of the rains given the current state of public health capacity in Haiti.  Lower or uncertain probability / high consequence events such as pandemic H1N1 outbreaks in the IDP camps or an outbreak of anthrax in Artibonite would likely proceed grossly unchecked given the current state of public health and agricultural surveillance and response capacity.

Presence of non-routine / unexpected biological threat.  We consider the level of risk for an outbreak of pandemic influenza in IDP camps during the rainy season to be moderate within the next 30 days.  This is a conservative assessment based on the following observations and concerns:

• there is minimal capacity to handle critical level pediatric or obstetric care, particularly ventilatory care; there are multiple reports indicating a rise in pediatric respiratory illness;
• an American physician contracted an apparent influenza-like illness on the ground severe enough for him to cancel his flight home;
• outbreaks of pH1N1 have been documented already several times in other countries among the impoverished and in evacuation shelters following hurricanes- the key is crowded conditions and lack of access to care;
• it is unknown to what extend the Haitian population has experienced a wave(s) of pH1N1 transmission- herd immunity unknown and no formal epidemiological, serological survey, or vaccination campaign has been conducted to our knowledge;
• indigenous peoples and those with access to poor healthcare, malnutrition, etc have been shown to be associated with bad community and clinical outcomes; 
• outbreaks of seasonal influenza in tropical countries are often associated with the rainy season, presumably due to higher person to person contact rates- crowding in IDP camps is an obvious concern; and
• responders from all over the world are flying in to assist from areas of the world where pH1N1 circulating is still documented

The uncertainty and concerns surrounding anthrax have been reported here.  Either of these events, if they appear, will further challenge the social expectation dynamic in Haiti.  We have seen in a multitude of other countries and scenarios that unexpected, non-routine infectious disease events, especially if children and pregnant mothers are killed, are strong triggers for anxiety behaviors, evacuations and population shifts, and social outcry occasionally leading to incendiary dissent.

Summary.  The public health and medical response effort in Haiti faces serious challenges in the next 30 days.  Failure to meet these challenges and associated social expectation may contribute to the incidence of social outcry and civil unrest.  Such events may erode the public trust gained so far in the Haiti response effort.

Anthrax in Haiti

Current migration pathways for the refugees of the Port-au-Prince devastation are shown below:

Credit: USSOUTHCOM, JTF-Haiti

USSOUTHCOM and JTF-Haiti estimate emigration volumes from Port-au-Prince as follows:

• Artibonite: 62,573
• Nord-Oest: 31,250
• Grand Anse: 30,000
• Nippes: 30,000
• Oest: 22,800
• Sud: 22,425
• Centre: 14,680
• Nord: 13,500
• Nord-Est: 7,700

When considering areas of Haiti that have reported anthrax epizootics and human cases, population displacement to the Artibonite Valley represents a point of concern.  Peck and Fitzgerald reported in 2007 a dramatic increase in incidence of human cutaneous anthrax cases, from 1 case reported in 1992 to 25 cases reported in 2002. [1]  This represented a 17-fold increase in incidence.  Fortunately, the volume of cases was low:  

Below is a map of the involved area.  

Of note, there is currently a population displacement vector heading to this area of Haiti numbering over 60,000 individuals.

The most common site of infection was the face (61% of cases), with massive edema as the most common clinical finding.  Other findings included fever, local lymphadenopathy, and abdominal pain.  Pain localized to the lesion was infrequent (14% of cases).  Below is a picture of a patient who survived infection following IV penicillin:

Anthrax appeared seasonally in the Artibonite Valley, peaking in May and June.  This provides us time to engage veterinary and epidemiological support to further investigate the risk to the growing population of IDPs in Artibonite Valley.  The call to action here is an updating of the Peck and Fitzgerald assessment, as it covered anthrax reporting up to 2002.

Other areas of Haiti have reported anthrax, particularly in the southern peninsula, however dramatic increases in incidence have not been reported in recent years.  Both cattle and goats have been associated with epizootics in Haiti.  The intersection of risk is demand by IDPs for protein sources in the ensuing weeks.  The last outbreak involving human cases of anthrax in Haiti was documented on August 14, 2009, an unspecified number of human anthrax cases including one fatal case in Calumette, Bell-Anse commune. The source of the infection was suspected to be contaminated meat.  Calumette is an area east of Port-au-Prince that reported “moderate” perceived shaking.  This is an area in proximity to a population displacement vector to Jimani, DR:

We would remind the readers that anthrax is considered hyperendemic in Haiti.  Reports of this disease would not be unexpected.  The purpose of this post is to highlight potential issues for ground responders to enable rapid recognition of the disease and effective treatment.

References

1.  Peck RN, Fitzgerald DW.  Cutaneous anthrax in the Artibonite Valley of Haiti: 1992-2002.  Am J Trop Med Hyg. 2007 Nov;77(5):806-11.

2.  GIDEON.  http://www.cyinfo.com