Natural disasters can result in an increase in vector-borne diseases such as malaria. In developing countries the effects of natural disasters on communicable diseases can be severe because resources and infrastructure are often limited [Reference Watson, Gayer and Connolly1]. On 13 March 2015, Tropical Cyclone (TC) Pam struck the developing island nation of Vanuatu in the South Pacific (Fig. 1). Over half of the country's population (estimated 250 000) was affected, but Shefa and Tafea provinces suffered the most extensive damage because they lay in the direct path of TC Pam. The ensuing environmental condition was conducive to vector breeding and the risk of vector-borne disease outbreaks was thought to be serious [2].

Fig. 1. Locations of 12 rural communities in Tafea Province, Vanuatu for post-Tropical Cyclone Pam malaria surveillance in June–July 2015. Inset shows the location of Vanuatu and Tafea Province (black square).

Tafea is the southernmost province of Vanuatu, and consists of five inhabited islands (Tanna, Aniwa, Futuna, Erromango, Aneityum) with an estimated population of 34 378 in 2014 (Fig. 1) [3]. Historically malaria was endemic on all islands except Futuna, and was very rare on Aniwa [Reference Kaneko4]. On Aneityum, an integrated malaria elimination project consisting of a 9-week mass drug administration that targeted the entire island population, distribution of insecticide treated nets (ITNs), introduction of larvivorous fish to known vector-breeding sites, and malaria surveillance by a community microscopist was initiated in 1991 [Reference Kaneko5, Reference Kaneko6]. Malaria was eliminated on Aneityum by 1996, and malaria freedom has since been sustained with the exception of an outbreak of Plasmodium vivax in 2002 (Table 1) [Reference Kaneko7]. The Aneityum Project demonstrates the feasibility of malaria elimination in Vanuatu and informs the current national elimination strategy specifically tailored to the isolated community setting typical of most of Vanuatu. Substantial reduction in annual parasite incidence (API) in Tafea Province over the past few years enabled the province to target zero local transmission by the end of 2014 and sustainable elimination by the end of 2016 [3]. Cognisant of the sharp increase in malaria incidence on Tanna Island after TC Uma in 1987 [Reference Kaneko4], this study was initiated to examine if the progress towards elimination in Tafea Province had been reversed by malaria outbreak after TC Pam.

Table 1. Major milestones of malaria elimination on Aneityum Island, Vanuatu since 1991

The study was approved by the Ethical Research Committee of Karolinska Institutet in Sweden (Protocol no. 2012/4:6) and the Ministry of Health in Vanuatu. Study participants were recruited using the convenience sampling method. The purpose, procedure, risks, and benefits of the survey were explained in both Bislama (lingua franca of Vanuatu) and indigenous ‘kastom’ languages by local field assistants. The consent procedure was witnessed by a neutral third party who recorded the name of each participant as he/she enrolled in the survey.

Malariometric surveys were conducted in 12 rural communities on the islands of Erromango (estimated population 2300), Tanna (29 000), and Aneityum (1200) in Tafea Province of Vanuatu in June and July 2015, about 3 months after TC Pam struck the nation (Fig. 1). Since the population of Aneityum is relatively small and concentrated in only three communities (Anelcouhat, Port Patrick, Umej), the entire population was included. In contrast, the populations of Erromango and Tanna in particular are larger and dispersed across many communities. On these two islands, selected communities with high parasite prevalence in recent years (‘historic hotspots’) were targeted to maximize the likelihood of detecting Plasmodium infections.

Peripheral blood was obtained by finger prick for examination of Plasmodium infection by microscopy and polymerase chain reaction (PCR). Giemsa-stained blood smears were examined in the field under a light microscope by M.K. Blood (70 μl) was spotted on Whatman 31ET chromatography filter paper (Whatman, UK), allowed to dry, and stored at ambient temperature. DNA was extracted from quartered dried blood spots (17·5 μl) using the QIAamp DNA Mini kit (Qiagen, Japan) according to the manufacturer's instructions. A nested PCR protocol that amplifies a fragment of the cox3 gene in the Plasmodium mitochondrial genome was used to detect Plasmodium infections [Reference Isozumi8].

Axillary temperature was determined using a digital thermometer (Terumo, Japan), and fever was defined as having axillary temperature >37·5 °C. On Tanna and Aneityum, spleen size of children aged ⩽12 years was assessed by A.K. according to Hackett's method [Reference Hackett9]. Survey participants were asked if they had slept under ITNs the previous night. χ ² tests were used to compare the prevalence of fever, spleen rates in children, and frequencies of ITN use in the islands.

Neither microscopy nor PCR detected any Plasmodium infection from 3009 survey participants across three islands, indicating no immediate outbreak in Tafea Province after TC Pam (Table 2). Between 2008 and 2014, the API in Tafea Province decreased from 23·3/1000 population to 0·087/1000 population as a result of interventions specifically aimed to achieve zero local transmission, including mass ITN distributions and enhanced case management, surveillance and response [3]. These control measures have effectively reduced the local parasite reservoir that can serve as the source of malaria outbreak after natural disasters.

Table 2. Post-Tropical Cyclone Pam malariometric surveys in Tafea Province, Vanuatu, June–July 2015

ITN, Insecticide-treated net.

Overall, only 6·3% (190/3007) of participants had fever (Table 2). Prevalence of fever did not differ between islands (P = 0·051), although in small communities such as Cook's Bay on Erromango and Umej on Aneityum, fever was more common at 16·9% (13/77) and 12·1% (18/149), respectively (Table 2). The cause of these fever cases was not further investigated as participants did not present with complaints or other health complications.

Enlarged spleen was detected in 4·6% (41/897) of children aged ⩽12 years on Tanna and Aneityum (Table 2). Spleen rate was higher on Tanna (5·3%) than Aneityum (3·6%), but the difference was not statistically significant (P = 0·245). We previously reported the association between parasite rate and spleen rate in Vanuatu and the decrease in spleen rates on Aneityum following malaria elimination in 1996 [Reference Chaves10]. The low spleen rates observed in this study were consistent with the temporal trend on Aneityum and provided support that malaria incidence had not increased in Tafea Province after TC Pam.

About 72·8% (2175/2986) of participants responded that they had slept under ITNs the night before the survey (Table 2). ITN use was significantly higher (P < 0·001) on Aneityum (85·9%) than Tanna (67·1%) and Erromango (66·8%). Compared to our most recent surveys conducted before TC Pam, ITN usage was consistent on both Aneityum (85·9% in June 2015 vs. 85·1% in January 2015) and Tanna (67·1% in June and July 2015 vs. 62·3% in February 2014) before and after TC Pam. Higher ITN use on Aneityum was motivated by the community's perception of the threat of malaria resurgence, and its long-term commitment to sustain malaria freedom after elimination [Reference Watanabe11].

Natural disasters can create new and favourable habitats for Anopheles vectors and increase the risk of malaria outbreak, even in remote island settings with low endemicity. In the Andaman and Nicobar Islands, seawater intrusion after the tsunami in December 2004 increased the number of brackish breeding sites preferred by the local vector Anopheles sundaicus. In Andaman district, slide positive rate for malaria increased markedly from <0·4% during 2004 to 1·6% 3 months after the tsunami, suggesting heightened transmission and risk of outbreak [Reference Krishnamoorthy12]. Previously in Vanuatu, the spike in malaria incidence on Tanna in 1987 was attributed to an outbreak after serious damage caused by TC Uma [Reference Kaneko4]. In contrast, no malaria outbreak was detected on the three main islands in Tafea Province 3 months after TC Pam in 2015. Progress made in recent years towards malaria elimination in the province likely minimized the potential for malaria outbreak stemming from indigenous parasite reservoirs. Similarly, no malaria outbreak was reported in the most affected area of the Visayas in the Philippines after Typhoon Haiyan in November 2003 [Reference Savage13], as malaria had been largely eliminated prior to the typhoon [14]. Both instances demonstrate that malaria elimination is robust and can provide the additional benefit in the form of protection from malaria epidemics after catastrophic natural disasters.

Preventing the reintroduction of malaria is a major challenge even in island nations with mechanisms in place to screen arriving passengers at ports of entry, such as Mauritius [Reference Tatarsky15] and Sri Lanka [Reference Dharmawardena16]. In Vanuatu, the risk of parasite importation via human movement from other provinces to Tafea Province still exists, especially for P. vivax [Reference Chan17]. In Tafea Province, active case detection in visitors is limited to only Aneityum as part of the community-directed surveillance system established in 1993 (Table 1) [Reference Kaneko6]. High community compliance in ITN use, timely case diagnosis and treatment, and maintenance of the existing surveillance and response system will be critical to the goal of achieving and sustaining malaria elimination in the future.