Patterns of Disease after Natural Disasters

Historically, infectious disease epidemics have caused high morbidity and mortality. Disasters have the potential for social disruption and death. Epidemics are compounded when infrastructure breaks down. But can a natural disaster lead to an epidemic of an infectious disease?

How often do natural disasters cause outbreaks? Experts are conflicted about the extent of infectious agents involved. A causal relationship is difficult to prove. Pre-event baseline surveillance is often weak. Population-based, post-event surveillance is difficult. Origins of outbreaks usually are multifactorial.

The phases of a disaster as outlined below provide a useful timeline and prediction of results.

• Impact Phase (0–4 days)
  • Extrication/trauma.
  • Hypothermia/heat illness/dehydration
  • Early soft tissue infections. Infectious complications in tsunami survivors often are the result of crushing or impaling injuries from wood, rock, concrete, or metal. Wounds become contaminated with tsunami water, soil, or particulate matter. Even minor wounds and abrasions could lead to overwhelming infection with Staphylococci, Streptococci, and water-borne organisms: Vibrio, Aeromonas, Pseudomonas, Burkholderia sp., and fungi. A Dallas evacuation facility post-Katrina had a cluster of 30 adult and pediatric patients with MRSA. Following Hurricane Katrina, an increase in mortality was reported by CDC from Vibrio (Vibrio vulnificus and parahaemolyticus) soft tissue infections. Risk factors include wading in contaminated flood waters (60% of cases are wound infections and ) and 40% result from raw shellfish consumption. Contact is with brackish salt water, prevalent in Gulf states. Necrotizing fasciitis may occur. Overall mortality is 40%, Twenty percent in the subset with Vibrio from wounds. Another soft tissue infection caused by Aeromonas hydrophilia is associated with brackish fresh water. People can get sick from eating infected fish and may be associated with leech bites.
• Post-impact Phase (4 days – 4 weeks)
  • Waterborne and foodborne illnesses from contaminated drinking water (cholera, bacterial dysentery, cryptosporidiosis, rotavirus, norovirus, typhoid and paratyphoid, giardiasis, hepatitis A and E, or dermal/mucosal contact with flood waters (Leptospirosis). Amongst Katrina evacuees in the Houston Astrodome complex, 18% of the 24,000 persons housed in the Reliant Park building developed acute gastroenteritis from September 2–12, 2005. Up to 21% (adults) and 40% (pediatric) medical visits were due to GI illness during this time period. Secondary spread to shelter and medical staff occurred. Fifty percent of specimens were positive for norovirus; no other pathogen was identified. There has also been reported an increase in gastroenteric infections following earthquakes; (1976) Friuli, Italy a 5–6 fold increase in Salmonella secondary to food contamination, poor hygiene and overcrowding, (1980) Campania, Italy increase in gastroeneteritis, (1983) Popaya, Columbia increase in hepatitis A, and (1990) Luzon, Philipines increase in gastroenteritis. In the aftermath of the Rwanda crisis in 1994 (“The Great Lakes Disaster”) outbreaks of cholera caused at least 48,000 cases and 23,000 deaths within one month in the refugee camps in Goma, Zaire (now Republic of the Congo). Shigella also erupted. Following Hurricane Mitch, (1998) cholera outbreaks occurred in Guatemala, Nicargua, and Belize. There is a strong seasonal association with monsoons and warmer water temperatures. Leptospirosis (Weils Disease) is the most widespread zoonotic disease in the world. It is caused by drinking, swimming, or wading in fresh water contaminated by infected rodent urine. Leptospirosis outbreaks have been linked to floods and cyclones as far back as (1964) in Iowa to Mumbai, India (2005) and Georgetown, Guyana.
  • Communicable respiratory infections. One must be aware of the selected respiratory pathogens in post-disaster settings. These include viral (influenza, RSV, Adenoviruses), bacterial (Strep pneumoniae, pertussis, tuberculosis, Legionella, Mycoplasma pneumoniae) and diseases transmitted via the respiratory route (measles, varicella, nisseria meningitides). The proportion of acute respiratory infections among reported illnesses after Hurricane Katrina rose from ~12% (4 days) after the levee overflowed to 20% during the next (4 weeks). Southeast Asia tsunami survivors often aspirated soil, sand, and other particulate matter as a consequence of near-drowning. Tsunami Lung (cavitary lung disease + brain abscess) is a pneumonic process that occurs up to six weeks later. It is usually polymicrobial, including water-borne organisms (Aeromonas, Pseudomonas, Streptococcal species, Nocardia, Pseudallescheria boydi, and Burkholderia pseudomallei). Increased death rates from measles were reported in camps in Bangladesh (1978), Somalia (1980), Sudan (1985), Ethiopia (1987), Malawi (1988–1990), Mozambique (1988–89, 1991), Philippines, (1991), Darfur, Somalia (1994), Haripur Afghanistan, and Kakuma refugee camps, Kenya (2005). Case fatality rates of 2.3% up to 32% were reported. Measles and the risk of transmission in the disaster-affected population is dependent on the baseline vaccination coverage rates among the affected population, and in particular among children aged <15 years. Crowded living conditions, as is common among people displaced by natural disasters, facilitate transmission and necessitate even higher immunization coverage levels to prevent outbreaks.
  • Other diseases. A modern tetanus outbreak (106 cases from December 30 – January 26, 2004) occurred in Banda Aceh 4–30 days post-tsunami. Case fatality rate was 18.9%.
• Recovery Phase (after 4 weeks)
  • Diseases with long incubation periods
  • Vector-borne illnesses such as malaria, western / Saint Louis encephalitis, dengue, yellow fever, and West Nile: There is a complex relationship with floods. The incidence may initially drop then rise. Flood waters can wash away stagnant breeding sites initially. Later, collection of standing water from blocked drains can increase incidence. Human behavioral changes induced by disaster such as sleeping outside and not using nets play an important role.
  • Exacerbations of chronic disease. We tend to focus on acute trauma, but prevention of chronic disease exacerbation is paramount. During Hurricane Katrina 20.2% or 275,000 people in the greater New Orleans area had a disability of some type. Physical and/or mental disabilities were a barrier to timely and successful evacuation. Forty-one percent of Katrina evacuees had at least one chronic health condition, such as cardiac disease, hypertension, diabetes, or asthma. Special needs for shelters, wheelchairs, oxygen, large cots for the obese, glucometers, access to dialysis, etc., were some of the problems. Among Katrina evacuees in Colorado, 38.8% of households were lacking prescription medications. 42.7% had missed doses of their medications.
  • Iatrogenic complications (e.g., transfusion-related infections)

What variables does one need to consider for the development of an epidemic after a disaster? First there are environmental considerations. Cold climate favors airborne agents and warm waterborne. Winter season heralds influenza, summer enteroviruses. El Nino has increased rainfall in flood-prone areas. Increased rainfall support breeding and survival of disease vectors such as malaria while drought leads to malnutrition and increased risk of infections.

The next variable to consider is endemic organisms. Infectious organisms endemic to a region will be present after the disaster. Agents not endemic before the event are unlikely to be present afterwards. For example, there was a nine fold increase in coccidiodomycosis (valley fever) from January–March 1994 following the Northridge Earthquake. Two hundred and three cases were reported in Ventura County, including three deaths. Cases peaked two weeks after the earthquake. The attack rate was 30 cases / 100,000 population overall. Fifty-six percent of the cases and highest attack rate of 114 / 100,000 was in the town of Simi Valley, at the base of a mountain range that had numerous landslides associated with the earthquake. Large dust clouds generated by landslides and strong aftershocks in Santa Susana Mountains north of Simi Valley were dispersed into nearby valleys by northeast winds. Increased risk was associated with being in, and amount of time spent, in dust cloud. Following Mount St. Helens volcano eruption in 1980, there was a giardiasis outbreak in Red Lodge, Montana from high ash levels. The combination of increased run-off due to heat absorption by ash-darkened snow and heavy rainfall in a city with inadequate chlorination and filtration caused an outbreak in 780 persons (33% of town residents and 15% of rural residents).

Another important variable to consider are population characteristics. This includes baseline immunity to specific diseases, age (increased elderly or children), presence of chronic disease, malnutrition, diabetes, heart disease, etc., natural and induced immunosuppression such as HIV/AIDS, pre-existing political strife (less responsive to disaster teams), hygiene (cultural norms), and stress. Religious barriers are important. In 2003, the muslim Kano state in Nigeria banned the polio vaccine, claiming that it would adversely affect fertility and cause AIDS. This led to a polio epidemic that spread from northern Nigeria to 11 surrounding countries, some of which had been declared polio-free. Increased density the result from displaced populations and refugee camps is important. The artificial, crowded communities often created in the aftermath of a disaster serve as a substrate for the spread of communicable disease. 800,000 Rwandans fled into Zaire (1994), which led to cholera and shigella dysentery outbreaks. 74% of all children under five years old are said to have died in Somalia displaced camps. Meningococcal meningitis outbreaks occurred in refugees in Thailand (1979) and Khartoum (displaced Sudanese, 1988).

Pre-event structure and public health are also important variables to consider. This includes such things as sanitation, adequate and reliable food and water supply, disaster preparedness, disease surveillance, transportation, communication network, medical infrastructure, and availability of equipment and medications.

We have talked about earthquakes and volcanoes now let’s consider tsunami. On December 26, 2004, southeast Asia was hit with a tsunami. An estimated 225,000 died from eight countries, including tourists. Millions were displaced. There was substantial infrastructure destruction. WHO predicted massive excess mortality from infectious diseases. Mortality was restricted however to a brief period in time, e.g., in Sir Lanka, of 30,000 deaths, all occurred in the first week.

The last variable to consider includes type of trauma and other post-disaster noninfectious illness. Earthquake, hurricane, and tsunami can lead to crush (rhabdomyolysis) and other penetrating injuries. Tsunami and floods contribute to near-drowning and electrocution-type injuries. In fact 40% of all natural disasters are caused by floods. Floods cause about 50% of all deaths from natural disasters in the United States. A large percentage are vehicle-related drownings during flash floods. The morbidity and mortality from flooding will continue to increase as population pressures force people to marginal lands that are steep. Subsequent deforestation for the purposes of agriculture increase the risks of flash floods and mudslides. Increases in infectious diseases follow floods. In 1979 Dominican Republic flooding after two hurricanes led to an increase in cases of typhoid fever and hepatitis. An outbreak of >75,000 cases of Plasmodium falciparum occurred in Hati after Hurricane Flora (1963). Following floods in Ecudaor (1983) there was a seven-fold increase in malaria reported and a 4–5 fold increase after a flood disaster in Mozambique (2000). In 1988 in Khartoum, Sudan there were increase cases of hepatitis A and gastrointestinal infections. Whenever the power grid goes down following floods, consider the special case of carbon monoxide poisoning, There were 78 cases and 10 deaths of CO poisoning following hurricanes Katrina and Rita (August–October 2006). Volcano and fires contribute to poor air quality, which led to increased respiratory distress. San Diego County experienced a significant increase in emergency department visits for respiratory symptoms from October 21 – November 9, 2007, during fire storm 2007 when over 12 fires burned during the emergency involving over 15% of the county, including 400,000 acres.

We discussed earlier the predicted phases of infectious disease and the importance of endemic organisms. Let’s consider the case of a 51-year-old previously healthy woman of Scandinavian descent. She was a victim of the southeast Asian tsunami. The initial medical problems included a deep cutaneous wound (Acinetobacter baumannii, Stenotrophomonas maltophilia, Achromobacter xylosoxidans) to her legs, multiple fractured pelvis, ruptured bladder (Enterococcus faecium), and pneumonia (Pseudomonas spp.). Two weeks later she developed a thigh abscess (Nocardia africanum). Eight weeks later developed altered mental status due to brain abscess (Scedosporium apiospermum). Three months later had a persistent draining wound (Mycobacterium cheloniae) from the tibial region (osteomyelitis excluded). When you take a healthy but immunologically naïve host from the other side of the world, and massively inoculate with unusual tropical organisms on top of severe trauma, will act like an immune compromised host.

Finally, there has been a philosophical shift in disaster medicine. Previously the focus was on the individual with healthcare as usual for small disasters that do not overwhelm the system. The goal was to improve the health of one person at a time. Resources were assumed to be plentiful. The system will take care of itself. Today the focus is on the population. Disaster medicine is for large situations that overwhelm the healthcare infrastructure. The goal is improving the health of the largest possible number of persons. This may involve altered triage priorities. NIMS/SEMS system management is critical.