When you choose to publish with PLOS, your research makes an impact. Make your work accessible to all, without restrictions, and accelerate scientific discovery with options like preprints and published peer review that make your work more Open.

PLOS BLOGS The Official PLOS Blog

Guest Blog by Oliver Morgan: The Impact of Concurrent Humanitarian Disasters

The open access publisher Biomed Central has just launched a new journal called Conflict and Health. We asked Oliver Morgan, Honorary Research Fellow at the London School of Hygiene and Tropical Medicine, to comment on a new study in the launch issue that explored the overlap between natural disasters, wars, and epidemics.



Competing interests: The author declares that he has no competing interests.

Humanitarian disasters occur frequently. In 2005, almost 96,500 people were killed and over 151,684,000 were affected by natural disasters, epidemics, civil conflict or war [1]. Given the number and scale of these events, it is surprising that there remain many questions about how they impact public health. Existing evidence suggests that the risk of epidemics following natural disasters is low [2-4]. During civil conflicts or wars, infectious disease risks tend to be higher [5-7]. However, multiple humanitarian disasters may occur simultaneously. For example, the populations of Aceh province in Indonesia and the north east coast of Sri Lanka were both affected by civil conflict when the South Asian Tsunami hit in December 2005.

A New Study on Concurrent Disasters

In a study published by a new open access journal Conflict and Health, Spiegel et al conducted an ecological analysis to investigate the frequency of overlap and occurrence of natural disasters, complex emergencies (civil conflicts and wars) and epidemics between 1995 and 2004 [8]. They identified the largest 30 events in each category as measured by the greatest number of deaths. These events were identified from an emergency events database called EM-DAT (natural disasters), the Uppsala Conflict Database (complex emergencies) and the WHO outbreaks archive (epidemics). For additional and supporting information they used the CE-DAT database on complex emergencies, the LexisNexis news service database, the Central Intelligence Agency World Factbook, the GlobalSecurity.org website and peer reviewed literature.

For each event in the three categories, concurrent disasters were defined as an event from one of the other two categories that occurred in the same geographical location (e.g. province). Natural disasters and epidemics were considered concurrent if they followed within six months of each other. Disasters that affected multiple countries were considered as single events. Terrorist attacks, transport or industrial accidents and chronic diseases such as HIV/AIDS were excluded.

During the study period they identified 3,197 natural disasters, 1,374 epidemics and 363 complex emergencies. The death toll for the 30 largest natural disasters ranged from 1,500 to 2.5 million deaths. The duration of these disasters ranged from 1 day to 7 years. Twenty seven percent of the 30 largest natural disasters occurred at the same time as a complex emergency, 23% were followed by an epidemic within 6-months of the disaster and 13% occurred at the same time as a complex emergency and was followed by an epidemic within 6 months. For the 30 largest complex emergencies, deaths ranged from 1,000 to 3 million, with duration of 1 to 41 years.

Looking specifically at complex emergencies, 87% were accompanied by a natural disaster, 63% were followed by an epidemic within 6 months, and in 60% of cases there was both a concurrent natural disaster and an epidemic within 6 months. The largest epidemics caused between 550 and 4,500 deaths and lasted between 31to 397 days. During an epidemic, a natural disaster occurred 30% of the time, a complex emergency 47% of the time and both events 10% of the time.

The authors report that these findings support the previous observation that natural disasters do not result in large-scale epidemics. Moreover, they assert that there is a hitherto unrecognised concurrence between natural disasters and complex emergencies. They conclude that epidemics are more likely to occur during complex emergencies than during natural disasters.

Strengths and Limitations of the Study

The authors used existing data sources to describe the frequency of multiple disaster types. A clear limitation of this approach is that these secondary data sources were unable to show whether the disasters occurred concurrently in the same populations. It is therefore uncertain whether these disasters did indeed overlap.

Moreover, the authors did not adjust for the different lengths of disaster duration, even though they acknowledge this as a major weakness in their study. This lack of adjustment makes their findings difficult to interpret. For example, they report that for 87% of complex emergencies, a natural disaster occurs concurrently. However, given the long period and circumstances of civil conflicts (such as in Darfur, Sudan), it is not difficult to imagine that in the course of time, both natural disasters and epidemics will occur. The authors also report that nearly a quarter of natural disasters were followed by an epidemic, a proportion that is much higher than the proportion found in other studies [2-4] and that conflicts with experiential evidence [9]. The high proportion in their study may be due to their assumption that epidemics and disasters occurring within six months of each other are linked.

What are the Implications for Practice and Future Research?

The take-home message is that epidemics occur more frequently during complex emergencies than natural disasters. This is already well established (if not always well recognised) [9]. The authors suggest that the overlap of different disaster types has implications for humanitarian organisations and professionals, which need to “bridge the gap” between complex emergencies and natural disasters. Any such gap may be overstated and health agencies, such as the International Committee of the Red Cross and Médicins Sans Frontières, have worked extensively in both settings.

Instead, one may hypothesise that the public health implication of concurrent multiple disasters is that they would have a synergistic effect; new disasters may compound existing public health problems and limit the ability for humanitarian and government agencies to respond. This synergy may be especially pronounced among populations with poor nutritional status affected by complex emergencies. A better understanding of this synergy would create a stronger case for preventive interventions among fragile populations and install a greater humanitarian urgency among donors and politicians.


1. EM-DAT. The OFDA/CRED International Disaster Database – www.em-dat.net. Accessed 24th January 2007: Université Catholique de Louvain – Brussels – Belgium.
2. Ahern M, Kovats RS, Wilkinson P, Few R, Matthies F. Global health impacts of floods: epidemiologic evidence. Epidemiol Rev. 2005;27:36-46.
3. Floret N, Viel JF, Mauny F, Hoen B, Piarroux R. Negligible risk for epidemics after geophysical disasters. Emerg Infect Dis. Apr 2006;12(4):543-548.
4. Watson J, Gayer M, Connolly M. Epidemics after natural disasters. Emerg Infect Dis [serial on the Internet]. 2007;Available from http://www.cdc.gov/ncidod/EID/13/1/1.htm.
5. Banatvala N, Zwi AB. Conflict and health: Public health and humanitarian interventions: developing the evidence base. Vol 321; 2000:101-105.
6. Salama P, Spiegel P, Talley L, Waldman R. Lessons learned from complex emergencies over past decade. The Lancet.364(9447):1801-1813.
7. Noji E. The public health concequences of disasters. New York: Oxford University Press Inc, USA; 1997.
8. Spiegel P, Le P, Ververs M-T, Salama P. Occurrence and overlap of natural disasters, complex emergencies and epidemics during the past decade (1995-2004).
Conflict and Health 2007, 1:2 (1 March 2007)
9. de Ville de Goyet C. Stop propagating disaster myths. Lancet. 2006;356:762-764.

Leave a Reply

Your email address will not be published. Required fields are marked *

Add your ORCID here. (e.g. 0000-0002-7299-680X)

Back to top