Epidemiological characteristics of malaria from control to elimination in Hubei Province, China, 2005–2016

Background Hubei Province, China, has been operating a malaria elimination programme. This study aimed at investigating the epidemiologic characteristics of malaria in Hubei Province (2005–2016) to plan resource allocation for malaria elimination. Methods Data on all malaria cases from 2005 to 2016 in all counties of Hubei Province were extracted from a web-based reporting system. The numbers of indigenous and imported cases during the disease control (2005–2010) and elimination (2011–2016) stages, as well as their spatiotemporal distribution, were compared. Results A total of 8109 malaria cases were reported from 2005 to 2016 (7270 and 839 cases during the control and elimination stages, respectively). Between 2005 and 2010, indigenous malaria cases comprised the majority of total cases (7114/7270; 97.9%), and Plasmodium vivax malaria cases accounted for most malaria cases (5572/7270; 76.6%). No indigenous malaria cases have been reported in Hubei Province since 2013. Imported malaria cases showed a gradually increasing trend from 2011 to 2016, Plasmodium falciparum was the predominant species in these cases, and the number of counties with imported cases increased from 4 in 2005 to 47 in 2016. During the control and elimination stages, the most likely spatial clusters for indigenous cases included 13 and 11 counties, respectively. However, the cluster of indigenous malaria cases has not been identified since September 2011. For imported cases, the most likely cluster and three secondary clusters during both stages were identified. Conclusions Hubei Province has made significant achievements in controlling and eliminating malaria; however, the region now faces some challenges associated with the increasing number and distribution of imported malaria cases. Priorities for malaria elimination should include better management of imported malaria cases, prevention of secondary malaria transmission, and ensuring the sustainability of malaria surveillance.

and measures need to be adjusted from control to elimination. The aim of the control stage was to reduce the morbidity and mortality of malaria among the population [5]. However, the malaria elimination stage requires the detection and response to individual malaria cases, and elimination of the source of infection to ultimately prevent local malaria transmission [6,7]. Since 2010, substantial progress in malaria elimination has been made in China [3,8]. The indigenous malaria burden in China has decreased, and locally transmitted cases decreased from 1469 cases in 2011 to 43 cases in 2015 [9].
Historically, malaria transmission has been unstable and prone to outbreaks in Hubei Province. Malaria has been prevalent in this region, and widespread outbreaks occurred in the 1950 and 1970s [10]. After decades of control efforts, the incidence of malaria reduced significantly, from 3688.9/100,000 population in 1973 to 1.2/100,000 population in 2009 [11]. As the NMEP was launched in May 2010, Hubei Province has undergone the change from the disease control (2005)(2006)(2007)(2008)(2009)(2010) to elimination stage (2011)(2012)(2013)(2014)(2015)(2016). While studies on the changes in epidemiological malaria characteristics between disease control and elimination stages have been conducted in China [3,5,12,13], few of these studies investigated the changes in malaria epidemic characteristics from the control stage (2005)(2006)(2007)(2008)(2009)(2010) to the elimination stage (2011)(2012)(2013)(2014)(2015)(2016) in Hubei Province. In this study, the epidemiological characteristics of malaria in Hubei from 2005 to 2016 were described and analysed, so as to help devise a strategy for resource allocation for malaria elimination in Hubei Province.

Study area
Hubei Province is located on the middle and lower reaches of the Yangtze River, which lies on the southcentral mainland of China. The province has a population of 57.2 million and covers 185,900 sq km. It includes 102 counties, and Wuhan City is the provincial capital ( Fig. 1).

Data collection
This retrospective study was conducted to compare the characteristics of malaria epidemiology between disease control and elimination stages. Data on malaria cases at county level in Hubei Province from 2005 to 2016 were obtained from the China Information System for Disease Control and Prevention (CISDCP). All clinically diagnosed cases and laboratory-confirmed cases were reported to the CISDCP. Clinically diagnosed cases were defined as individuals presenting with malaria-like symptoms, with a history of travel to a malaria-endemic area during the malaria transmission season, but showing no signs of malaria with a laboratory test. Laboratory-confirmed cases were defined as individuals with confirmed disease through microscopic examination, rapid diagnostic testing (RDT), or polymerase chain reaction (PCR). Both types of cases were included in this study.
In China, an indigenous malaria case is defined as a malaria infection that was acquired locally and in which the patient does not have a history of travel. An imported case is defined as an individual having a history of travel to a malaria-endemic country outside China and an onset In this study, the period from 2005 to 2010 was defined as the control stage and the period from 2011 to 2016 as the elimination stage.

Seasonal fluctuation analysis
A seasonal index for each month from 2005 to 2016 was used to identify the seasonal patterns of malaria incidence (including both indigenous and imported cases). The index was calculated as the average of the number of cases in a given month (e.g., July 2005) divided by the mean of number of cases in the corresponding month during the 12 years of analysis (e.g., July 2005-2016). There was no seasonal fluctuation when the seasonal index for a given month was close to 1.0.

Space-time clusters analysis
A space-time scan statistic was employed to test whether malaria cases were distributed randomly over space and time, and if they were not, to detect space-time clusters and evaluate their statistical significance. SaTScan software (version 9.4, Kulldorff and Information Management Services, Inc.) was used to conduct this retrospective space-time scan statistic [14,15]. The maximum spatial cluster size of the population at risk was set to 50%, and the maximum temporal cluster size was 50% within the study period. The most significant clusters as well as secondary clusters were identified. The ArcGIS software version 10.0 (ESRI Inc., Redlands, CA, USA) was used to visualize space-time clusters of indigenous and imported malaria cases.  (Table 1).

General epidemiological malaria characteristics during the control and elimination stages
Plasmodium vivax malaria cases accounted for the majority of total malaria cases during the control stage (5572/7270; 76.6%). In contrast, most malaria cases were attributed to Plasmodium falciparum infection (518/839; 61.7%) during the elimination stage. The proportion of malaria cases associated with P. falciparum increased from 0.1% in 2005 to 68.2% in 2016 (Fig. 2).
The majority of malaria cases (80.7%; 5865/7270) were clinically diagnosed cases in the control stage. During the elimination stage, clinically diagnosed cases accounted for 4.1% (34/839) of total cases. All malaria cases were laboratory-confirmed cases since 2012.

Season fluctuation of indigenous and imported malaria cases
Indigenous malaria cases showed a clear seasonal fluctuation during the study period (Fig. 3). A relatively high seasonal index of indigenous malaria was observed from May to September, during which 71.6% (5154/7203) of all indigenous cases were reported. The seasonal index reached a peak in August (2.26). For imported cases, two peaks (one in January to February and a second one in May to September) were observed. There was significant difference in the proportion of monthly cases between indigenous and imported malaria (χ 2 = 106.28, p < 0.0001).  Table 2).

Spatiotemporal cluster analysis
The spatiotemporal cluster analysis of malaria cases showed that both indigenous and imported cases were not randomly distributed throughout the study period. During the control stage, the most likely cluster of indigenous cases covered 13 counties. This cluster had 4273 cases with a relative risk (RR) of 18.58 (p < 0.001) and accounted for 58.8% (4273/7270) of the total cases during the control stage. Four different significant spatial clusters were identified for imported cases during the control stage (the most likely cluster and three secondary clusters). The most likely cluster included one county, in Wuhan City. Three secondary clusters covered five counties, which were located in Yichang City (1 county), Shiyan City (3 county), and Jingzhou City (1 county) (Fig. 4, Table 3). During the elimination stage (2011-2016), the most likely cluster of indigenous cases included 11 counties.  For imported cases during the elimination phase, the most significant cluster covered 6 counties, all in Wuhan City; three secondary clusters included three county, in Yichang, Jingmen and Enshi cities (Fig. 4, Table 3).

Discussion
The study analysed longitudinal surveillance data from 2005 to 2016 to determine changes in the epidemiological characteristics of malaria in Hubei Province during disease control and elimination stages. The study revealed a remarkable decrease in indigenous malaria cases in this region over the past 12 years; in contrast, the number of imported malaria cases increased significantly during the study period.
In the control stage, malaria cases decreased since 2008. A total of 429 malaria cases were reported in 2010, which was the lowest number of malaria cases in control stage. Malaria prevalence was effectively controlled by the high levels of coverage of interventions [5], including mass drug administration, insecticide-treated mosquito net distribution and spraying insecticide indoors [16,17]. During the elimination stage, no locally transmitted malaria was maintained from 2013 to 2016 with the implementation of NMEP. The strategy of malaria elimination emphasized detection, investigation and disposal of malaria cases and foci, and the '1-3-7' strategy was implemented to ensure effective case detection and confirmation, case classification and treatment, and focus investigation and action [18].
The decrease of malaria prevalence over the past 12 years in Hubei is driven by two factors. One factor was the implementation of the Global Fund to Fight AIDS, Tuberculosis and Malaria, from 2003 to 2012. These initiatives strengthened anti-malaria measures, such as active case detection, standard treatment of malaria patients, improvements in the direct reporting network system of malaria cases, the training of more rural physicians, health education and promotion of malaria in high-risk populations, and vector control in high-transmission areas [19][20][21]. Another factor was the cooperative malaria control between Hubei and four other provinces (Jiangsu, Shandong, Henan, Anhui) since 1974, which facilitated the application of unified measures, regular exchange of experience and information, as well as annual inspection and monitoring [22]. Imported malaria cases accounted for most cases during the elimination stage, and P. falciparum was the predominant species in these cases. The majority of imported cases in Hubei Province originated in African countries with endemic malaria where P. falciparum is the main parasite [23][24][25]. P. falciparum infection typically leads to severe disease and death. It has been reported that severe malaria was associated with medical staff not recognizing malaria at an early stage [26][27][28]. Hubei Province has not had any indigenous cases of P. falciparum malaria since 1963, and therefore a lack of awareness of the disease exists in medical staff, which compromises their ability to diagnose and treat these cases. Malaria-related deaths have been reported in Hubei Province [27]. Early discovery, diagnosis and treatment should be prioritized for imported malaria cases [29], and healthcare workers should pay more attention to diagnosis and treatment of imported cases, particularly those originating in Africa.
Imported malaria was more widespread during elimination stage than during control stage. It was identified in four counties in 2005; this increased to 47 counties in 2016. Malaria re-introduction caused by imported cases has been reported in areas that are non-endemic for malaria [30][31][32]. Imported malaria may lead to local transmission in some areas of Hubei where transmitting vectors remain [33,34], posing a great challenge to malaria elimination in the province. Enhanced surveillance systems should be sustained to maintain elimination of local transmission in Hubei Province by monitoring malaria importation and potential transmission in risk areas [35]. Moreover, cooperation between public health, health care, immigration and quarantine services, the commercial sector, and police will likely play an important role in the prevention, detection and management of imported malaria [25].
The analysis of this study showed a weak seasonal fluctuation for imported malaria cases (with peaks occurring in January to February and May to September). Chinese New Year holidays are celebrated in January/February, and workers return to China from overseas countries to   [38].
Space-time scan statistical analysis can be used to detected high-risk areas, which provide significant references to prioritize the resource assignment in malaria elimination. The clusters of indigenous malaria cases were seen in 11 counties (Zaoyang, Xiangzhou, Zengdu, Yicheng, Xiangcheng, Fancheng, Zhongxiang, Laohekou, Nanzhang, Guangshui, Jingshan) during control and elimination stages. The two most important vector species for malaria transmission, Anopheles sinensis and Anopheles anthropophagus, have existed in most parts of these cluster areas for a long time [39], and malaria epidemics have occurred frequently in these areas over past decades [11]. However, a cluster of indigenous malaria cases has not been identified since September 2011.
The most likely cluster of imported cases was located in six counties of Wuhan City (Qiaokou, Jianghan, Hanyang, Jiangan, Wuchang, Dongxihu) from 2005 to 2016. As this region has better medical care than the other cities of Hubei, many overseas labourers return to this area [23,40]. The secondary clusters were detected in the Yichang, Jingmen, Shiyan, Jingzhou, and Enshi cities. In these cities, many labour service and construction companies offer overseas job opportunities in Africa and Southeast Asia, both of which have endemic malaria [41]. Overseas labourers generally work outdoors at construction sites, they lack awareness of the risks of malaria and personal protection. Interventions strengthening their awareness of the risks of malaria and occupation-based vector control measures should be implemented to reduce the risk of malaria infection among export labourers [42,43].
There were two limitations to this study. First, underreporting and misreporting of malaria cases may have existed during the study period, this has been attributed to asymptomatic infection, capacity of malaria testing, and the availability of a healthcare facility. Second, a few imported cases missing information regarding the origin of the infection regions during the control period may have affected the observed origin of imported malaria cases.

Conclusions
The prevalence of indigenous malaria has declined dramatically over the past 12 years in Hubei Province, China. Although no indigenous malaria cases have been reported in Hubei since 2013, maintaining the status of non-locally transmitted malaria faces some challenges, such as the increase in the number of imported malaria cases and an increasingly widespread distribution of imported cases across the province. Accordingly, the management of imported malaria, prevention strategies for re-introduction of malaria, and malaria surveillance should be strengthened.