Epidemiology and control of malaria in Papua New Guinea : from small-scale heterogeneity to large-scale surveillance and targeted response

Rodríguez-Rodríguez, Daniela. Epidemiology and control of malaria in Papua New Guinea : from small-scale heterogeneity to large-scale surveillance and targeted response. 2019, Doctoral Thesis, University of Basel, Faculty of Science.


Official URL: http://edoc.unibas.ch/diss/DissB_13514

Downloads: Statistics Overview


Papua New Guinea (PNG), with a total estimated population of 8.8 million by 2019, has great environmental and cultural diversity which is mirrored by a complex malaria epidemiology. The geographic landscape in PNG is very diverse and in places extremely rugged. Malaria is endemic across most parts of Papua New Guinea and heterogeneous levels of endemicity characterize different areas of the country, from areas with intense transmission to unstable transmission areas with low levels of endemicity and even areas with “anophelism sans malaria”. Heterogeneity in endemicity has been attributed to factors within the human, the vector and the parasite. For instance, it has been documented that abundance of alternative hosts such as dogs and pigs together with historic and current control have given rise to significant small-scale heterogeneities in morbidity.
In 2004, control efforts were re-intensified with funding from the Global Fund to Fight AIDS, Tuberculosis and Malaria. Countrywide campaigns distributed free LLINs at the household level and, starting late 2011, improved diagnosis by microscopy and RDTs together with the introduction of ACT have been provided progressively at more public and church-run health facilities. In addition the programme was complemented by advocacy and behaviour change campaigns. As a result, the prevalence of malaria decreased from 11.1% (95% confidence interval, CI: 8.5–14.3) in 2008–2009 to 5.1% (95% CI 3.6–7.4) in 2010–2011 and 0.9% (95% CI 0.6–1.5) in 2013–2014, an unprecedented reduction in PNG. In 2017, the latest national survey registered prevalence levels higher than those in 2010/11. In only three years, the estimated number of malaria infections across PNG increased 8.6-fold to 7.1% (95% CI 5.0, 10.1). Four different species of human malaria have been identified in PNG. Of these four, the two dominant species are P. falciparum and P. vivax. Overall, Plasmodium falciparum has remained the dominant species over P. vivax, but their distribution has not been even across the country. In addition, substantial heterogeneity in the prevalence of malaria across PNG has been consistently found over the years with marked differences even between nearby villages.
The aim of this work is to provide a better understanding of the heterogeneous malaria transmission and the dynamics of Plasmodium, humans and interventions rolled-out by the Papua New Guinea National Malaria Control Program. This work comprises two major components: 1) a retrospective analysis of incidence of malaria cases in selected sentinel health facilities including a visualization of trends over time in different Sentinel Health Facilities (SHFs), and 2) a cross-sectional malaria survey complemented by a community based qualitative behavioural study.
The retrospective analysis of incidence found that malaria incidence in different sites initially ranged from 20 to 115/1,000 population; subsequent trends varied by site. Overall, LLIN distributions had a cumulative effect, reducing the number of malaria cases with each round (incidence rate ratio range 0.12 to 0.53 in five sites). No significant reduction was associated with ACT introduction. Plasmodium falciparum remained the dominant parasite in all sentinel health facilities from 2010 to 2014. Resurgence was observed in one site in which a shift to early and outdoor biting of anophelines had previously been documented. LLINs distributions, but not ACT, were associated with reductions of malaria cases in a range of settings, but sustainability of the gains appear to depend on local factors. Malaria programmes covering diverse transmission settings such as PNG must consider local heterogeneity when choosing interventions and ensure continuous monitoring of trends.
The visualization of incidence trends and other information (net use and residence of patients) extracted from a routinely implemented surveillance system proved useful to inform local malaria control programs to better target interventions. The visualization approach added a geospatial component to health facility data in order to understand differences in malaria burden between villages and identify communities that would benefit from targeted interventions or investigations. However, a functional simple tool for calculating and mapping malaria case incidence at district or sub-district level (e.g. eNHIS or similar) is required to operationalize the approach, along with the capacity, policies, and mechanisms required to implement targeted response action at the respective operational level.
The qualitative behavioural investigation identified seven behavioural groups (or demographic groups exhibiting similar behaviours) and highlighted the substantial amount of time spent outdoors or in non-secure structures when ‘indoors’ as a major risk of exposure. Between 4pm and 8am, all age groups in both study sites were likely to be exposed to mosquito bites across all types of activities; sleeping under a LLIN was the exception. Such findings highlight the potential of ‘outdoor biting’ to hamper malaria control and elimination efforts if not addressed appropriately since people spent a remarkable amount of time outdoors without protection from mosquito biting. Targeting groups, places and activities in order to prevent outdoor biting in the early hours of the evening and the morning seems crucial towards elimination.
This work also reveals spatial heterogeneity in the prevalence distribution of malaria and LLIN use between study sites. Malaria prevalence in the Mugil area was 3.7 fold higher than in the Lemakot area. Interestingly, LLIN-use was 2.4 times higher in the Mugil area compared to the Lemakot area. Spatial heterogeneity of malaria was also observed at a village and households level. Prevalence between villages ranged from 0.8% to 19.5% and between households from 0% to 66.6%. In the Mugil area identified risk factor related to behavioural groups (adult women were at lower risk and school children at higher risk) and housing (screened windows and traditional houses were associated with lower exposure) while in the Lemakot area LLIN ownership was a predictor for infection. The identification of site-specific risk factors provides evidence to potentially inform complementary interventions in a local scale that target specific groups or areas.
Heterogeneity of malaria trends was consistent throughout this work. The retrospective analysis and the cross-sectional malaria survey identified: i) heterogeneous effects of malaria interventions across the country, and ii) a heterogeneous distribution of malaria cases over space and time. The cross-sectional malaria survey highlighted varying prevalence between study sites and between neighbouring villages within sites. These findings emphasize the need for locally informed strategies toward improved control. Some communities could still benefit from improved LLIN ownership and use, whereas others might need to complement control with alternatives to LLINs. Targeted interventions in areas of higher transmission has been proposed by modelling and some field studies as opposed to untargeted community-based approaches, but the evidence comparing their effectiveness is scarce. Future research in PNG could address this gap and compare the effect of different control strategies that combine targeted and untargeted interventions.
In addition, outdoor and earlier biting of Anopheles species has been identified as a threat to LLINs effectiveness in PNG and other settings. Studies in PNG have described a shift in mosquito biting to earlier hours following the first LLIN distribution (the peak exposure time to infective bites shifted from later than 9pm in 2008 to between 6 and 7 pm in 2011). Our results identified and increase in the number of cases in one site by 2014 despite consistently high LLIN ownership and use in the area. The behavioural investigation identified potential exposure to mosquito bites on the amount of time spent outdoors (when not asleep) or in non-protected structures. Therefore, it is possible that the reduced efficacy of LLINs in synergy with human behaviour and ACT stock outs led to the observed increase especially in places with historically high mosquito densities.
During the course of this work, malaria elimination from PNG by 2030 became less likely than when it was originally envisioned in the National Malaria Strategic Plan 2014-2018. The resurgence in malaria is likely to worsen unless malaria control is re-intensified and maintained. Structuring programmes in response to evidence of the local malaria burden together with an analysis of transmission will enable adapting the strategy to the local context and optimize the use of resources. However a strong and functional surveillance & response system is needed to monitor the local burden and inform control efforts. Evidence in this study documented reasonable high LLIN ownership across study sites; however LLINs use can be improved in some areas. RDTs and ACTs were not always available in the health facilities therefore efforts need to be made to assure availability especially in areas with higher transmission. Since outdoor biting was consistently identified as an exposure risk and specific groups and areas at higher risk were also identified targeted complementary interventions could be explored and piloted in PNG. Further studies could address the current evidence gap on the effectiveness of targeted interventions.
Advisors:Utzinger, Jürg and Hetzel, Manuel Wolf-Werner and Reeder, John
Faculties and Departments:09 Associated Institutions > Swiss Tropical and Public Health Institute (Swiss TPH) > Former Units within Swiss TPH > Health Impact Assessment (Utzinger)
UniBasel Contributors:Rodriguez-Rodriguez, Daniela
Item Type:Thesis
Thesis Subtype:Doctoral Thesis
Thesis no:13514
Thesis status:Complete
Number of Pages:1 Online-Ressource (xix, 189 Seiten)
Identification Number:
edoc DOI:
Last Modified:10 Mar 2020 05:30
Deposited On:09 Mar 2020 14:14

Repository Staff Only: item control page