Surveillance efforts after mass drug administration to validate elimination of lymphatic filariasis as a public health problem in Vanuatu

Background Vanuatu was formerly highly endemic for lymphatic filariasis (LF), caused by Wuchereria bancrofti and transmitted by Anopheles mosquitoes. After a baseline survey showing 4.8% antigen prevalence in 1998, the country conducted nationwide (in one implementation unit) annual mass drug administration (MDA) with albendazole and diethylcarbamazine citrate from 2000 to 2004 and achieved prevalence of 0.2% by 2006 in a representative nationwide cluster survey among all age groups. Methods Post MDA surveillance was conducted from 2006 to 2012. After MDA, the country was divided for surveillance into three evaluation units (EUs) formed by grouping provinces according to baseline prevalence: EU1: Torba, Sanma and Malampa; EU2: Penama; EU3: Shefa and Tafea. The study compiled all past data and information on surveys in Vanuatu from the country programme. This paper reviews the surveillance activities done after stopping MDA to validate the interruption of transmission and elimination of LF as a public health problem. Results Post-MDA surveillance consisting of at least three transmission assessment surveys (TAS) in each of the three EUs was conducted between 2006 and 2012. Sentinel and spot check surveys identified a few villages with persistent high prevalence; all antigen positive cases in these sites were treated and additional targeted MDA conducted for 3 years in 13 villages in one area of concern. All three EUs passed all TAS in 2007, 2010 and 2012 respectively, with no positives found except in EU2 (Penama province) in 2012 when 2 children tested positive for circulating filariasis antigen. Assessment of the burden of chronic filariasis morbidity found 95 cases in 2003 and 32 remaining cases in 2007, all aged over 60 years. Conclusions Vanuatu has achieved validation of elimination of LF as a public health problem. Post-validation surveillance is still recommended especially in formerly highly endemic areas.


Background
Vanuatu is a Pacific Island nation and one of the 16 Pacific Island countries and territories included in the Pacific Programme to Eliminate Lymphatic Filariasis (PacELF) started in 1999 [1,2]. PacELF is part of the Global Programme to Eliminate Lymphatic Filariasis (GPELF) which was launched in 2000 and aims to eliminate lymphatic filariasis (LF) as a public health problem by 2020 [3]. Elimination as a public health problem in this context includes interruption of transmission by mosquitoes and provision of services for those suffering from acute and chronic LF morbidity (acute attacks of lymphangitis, lymphoedema/elephantiasis and hydrocele).
Vanuatu is known to have been highly endemic for LF, based on a few surveys conducted prior to 1997, which have been reviewed in [4]. Some control efforts using mass drug administration (MDA) for up to 3 years in seven islands did not succeed in interrupting transmission [5]. At the start of the PacELF control programme in Vanuatu in 1999, the national LF antigen prevalence measured by immunochromatographic test (ICT) was estimated to be 4.8% in a baseline survey of 51 villages. Through intense efforts of annual MDA campaigns combining albendazole and diethylcarbamazine citrate (DEC) with vector control, antigen prevalence was reduced to 0.2% by 2006 as estimated in a survey of 90 household clusters. Programme implementation and MDA campaigns from 2000 until the final round of MDA in 2004 and the stop MDA survey in 2006 have been previously reported [4]. The current paper reviews the monitoring and post-MDA surveillance activities (including spot check site surveys after 2004) as evidence towards validation of elimination of LF as a public health problem according to current WHO validation process [6]. Available information on morbidity is also included in the present paper.
The monitoring and surveillance framework GPELF 2000 guidelines: Initially the Vanuatu LF programme followed the first (2000) GPELF guidelines [7]. These called for a mapping survey to assess LF endemicity in designated implementation units, testing of at least two sentinel sites and two spot check sites at baseline before the first round of MDA and during MDA, and, after achieving a microfilariae (Mf ) prevalence rate of less than 1% in these sites, performing a lot quality assurance sampling (LQAS) survey of 3000 children aged 6-10 years, born after the initiation of effective MDA rounds. The LQAS survey was designed to determine whether further MDA rounds can be stopped using a cut-off of <0.1% circulating filariasis antigen (CFA) prevalence.
PacELF 2003 guidelines: Due to small population sizes of some Pacific Island countries and the lack of tests able to determine such a low level cut-off, the PacELF developed its own monitoring and evaluation (M&E) framework in 2003 and 2004 for assessing interruption of transmission [1]. This framework stipulated a community cluster survey of all ages, and MDA to be stopped if antigen prevalence was <1%. In the PacELF terminology, the baseline survey was A, the sentinel site surveys B, and the stop MDA survey was the C survey. The D survey was the LQAS survey among children as proposed by the GPELF.
Modified 2005 GPELF guidelines: In 2005, the GPELF published the revised M&E guidelines [8] which recommended increased numbers of sentinel and spot check surveys be conducted before the fifth MDA to determine whether the Mf prevalence in all of the sampled sites was <1%. To determine whether MDA should be stopped, the LQAS survey of 3000 children was retained, but the recommended age group was changed to school entrants. However, Vanuatu was already in the process of conducting a C survey in all age groups in 2005 [4], rather than following the 2005 GPELF guidelines.
Modified PacELF 2008 guidelines: In 2007, the PacELF proposed a new surveillance framework that was modified in 2008 [9] for the Pacific countries, including Vanuatu. These guidelines retained the community-based cluster C survey in all ages for the stopping MDA decision with a threshold of 1% antigen prevalence but modified the D survey to a child transmission survey (CTS). The CTS had a target sample size of all children aged 5 in a country (with the exception of PNG), to detect antigen positive children, either in school or community-based surveys. An additional action component included was the tracing of contacts of any cases who were positive for CFA (by ICT) or Mf, by testing surrounding residents within 200 m or the nearest 24 houses to the index child's place of residence. Vanuatu adopted this approach and conducted a nationwide CTS in 2007. The 2008 PacELF framework used in Vanuatu was implemented to some extent in other Pacific countries.
GPELF 2011 guidelines: The modified PacELF framework used until 2008 was then superseded by the new WHO-recommended protocol for transmission assessment surveys (TAS) [10] which was the procedure followed by Vanuatu for subsequent post-MDA surveys after 2010. PacELF followed these guidelines for TAS after 2008. Table 1 shows the summarised timeline of activities between 1997 and 2012, including specific timing of post MDA surveillance and follow-up activities described in this paper.

LF programme timeline
Following the baseline survey in 1997 and 1998, annual MDA rounds were conducted for five consecutive years nationwide during 2000-2004, with reported national coverage rates of 83, 84, 84, 87 and 85% respectively [1,4,11]. Sentinel site B surveys were started in eight villages in 2002 [12]. Spot-check site surveys were also initiated in several villages and hospitals

Sentinel and spot-check site surveys and targeted MDAs
As reported previously and summarised in Table 2 Table 2. All sentinel sites showed a decline in CFA prevalence over time, and five of the six sites had reached 0% CFA prevalence by 2006. The remaining site with persistent positives was Wanur in South Pentecost, Penama province. The locations of the sentinel sites are indicated in Fig. 1.
Some additional villages were sampled as spot-check sites, some of them more than once, from 2002 onwards. These surveys are also compiled in Table 2, with locations of the sites indicated in Fig. 1. The surveys tested all age groups, with additional tests done on 3-6 year olds or 6-7 year olds in certain villages, as shown as footnotes to the table.
It can be noted from Fig. 1 that North Ambrym (1999 census population 3899) and South Pentecost (2000 estimated population 2222) are located close to each other, despite being in different provinces and evaluation units (EUs). North Ambrym was identified as an area of concern due to remaining high prevalence   In 2011, ICT testing (N = 1100) was also conducted at 2 main hospitals (in Santo and Port Vila) on persons presenting for malaria examinations, and zero CFA positives were found. As for migrant screening, the programme tested 101 students who came to study at the University of South Pacific in Port Vila from other Pacific Island countries and found zero CFA positives. The formerly positive persons in all 8

Transmission assessment surveys
The C survey of 2005/2006 is now referred to as TAS 1, although it was done in all ages rather than children aged 6-7 years as recommended by the transmission assessment surveys (TAS) protocol [10]. Results were presented in [4].  (Table 3). There were 25 teams in total and 2 to 6 teams were assigned per province. The teams covered 86.2% of the target number of children in EU1, 52.3% in EU2 and 71.8% in EU3. No CFA positives were found among the 4752 children tested, aged 6-7 years, giving upper 95% confidence intervals for CFA prevalence of 0.2% in EU1, 1.0% in EU2 and 0.3% in EU3 ( Table 3). The survey sample builder software [10] was not used for these surveys, but no positives were detected so the results were below any threshold that would have been generated. Thus, in TAS 2, all three EUs passed.   (Table 4) and no CFA positives were found among the 4480 children tested. Despite the low sampling proportion in EU3, all three EUs passed the TAS 3 (Table 4), as the results were below the critical cut-offs (first integer <0.02N, where N is the estimated population in each EU). The TAS 3 in EU2 was conducted with support from the Task Force for Global Health, and the results have been published in summary form [13].
A TAS 4 was performed only in EU2 (Penama province) in 2012 ( Table 5). The estimated total population of 6-7year-old children was 1034 in the EU. Two ICT positive children were found out of 933 tested. The critical cut-off was 20 and therefore EU2 passed TAS 4 [10]. The results of the TAS 4 have been published in summary form [13].

Morbidity burden estimates
In 2003, the LF programme attempted to identify all persons with LF morbidity in the country by enlisting health staff to investigate during the MDA round. A morbidity survey form was inserted into the 2003 MDA registration books, and nurses were instructed to record any morbidity patients in their area on the forms which were submitted back to the national programme after the MDA. The resulting estimates are shown in Table 6, and the numbers and location by island in Fig. 2. A total of 95 cases were found in 2003, with two thirds being in males; about one third of cases were hydroceles or mixed hydroceles, or mixed limb and breast-related morbidities.  In 2007, during the CTS/TAS 2, an attempt was made to determine an updated estimate and case register of LF morbidity cases with local nurses. The teams carried with them sufficient morbidity kits for all the previously identified cases (care charts, basins, towels and bars of soap) after the team members were trained in morbidity management. At this second round of the assessment, only 32 people with morbidity were followed up (18 males and 10 females, 4 in Sanma province with unknown gender) compared to 95 in 2003 (Table 6); however, it was reported that some of the survey teams did not have time to do thorough morbidity assessment. The cases recorded in 2007 were all above 60 years of age, and again, the majority had limb-related morbidity rather than hydroceles. Decrease in numbers of cases between 2003 and 2007 was seen in all provinces, with the biggest decrease being on Pentecost Island in Penama province (EU2).

Availability of hydrocele surgery and numbers performed
During the period 1998 to 2006, hydrocele surgery was available in Port Vila hospital, and an estimated 10-15 surgeries were performed during that period on hydroceles of stages 2 to 4 [14]. After 2006, this surgery was not available in Vanuatu until a visit by the external surgeon in November 2015. There were at least 23 cases of hydroceles with surgical backlog, of which eight cases were presented for examination. However, out of these 8 patients selected by referring physicians with a diagnosis of hydrocele, all 8 were actually confirmed as inguinal hernia. All were of big size, most of them were non reducible and none was strangled.  While no hydrocelectomy was performed at that time, the fact that all referred cases were hernia indicates the need for further training in the differential diagnosis of hydroceles and poses the question of the real number of remaining patients affected by "hydroceles" out of the remaining 15 (23 minus 8) from Table 4 as part of post-validation activities of LF morbidity management.

Discussion and Conclusions
Vanuatu's national CFA prevalence by ICT before MDA in 1998 was 4.8% (N = 4362) and after MDA in 2005 was 0.2% (N = 7580) [4]. When considered by EU, which is a combination of provinces used to design sampling frames for the C survey/TAS 1 in  would be advisable especially in South Pentecost to ensure that transmission is not persisting any longer.
The dossier summarising all the information about the LF programme in Vanuatu was prepared in 2012 and submitted to WHO in 2013. It was reviewed during the 13th Meeting of the Western Pacific Regional Programme Review Group (RPRG) on neglected tropical diseases (NTDs) in July 2013, which suggested a few modifications. The dossier was updated accordingly and forwarded to WHO headquarters for necessary actions in October 2014. The WHO Strategic and Technical Advisory Group on NTDs endorsed the official process to validate elimination of LF as a public health problem in April 2015. As per the new process, elimination of LF as a public health problem is to be validated by the regional reviewing authority convened by the relevant WHO Regional Office and acknowledged by the WHO Director General. Accordingly, the Regional Dossier Review Group was convened by the WHO WPRO, which reviewed the dossier of Vanuatu and recommended validation of the claim for Vanuatu. An official acknowledgement of accomplishment of LF elimination as a public health problem in Vanuatu was given by the WHO Director General and the WHO WPRO Regional Director during the 67th session of the Regional Committee Meeting held in Manila in October 2016.