Impact case study database

4.1 Malaria

In collaboration with partners in Africa such as the Somalian Ministry of Health and epidemiologists and geographic information system (GIS) experts from the KEMRI Wellcome Trust in Nairobi, Giorgi developed novel MBG methods and applied them for malaria mapping in sub-Saharan Africa which has subsequently informed malaria control policies [S1, S2]. For example, in 2017, Somalia launched a 5-year malaria strategic plan with the aim of sustaining malaria prevalence levels below 1% in the north and increasing access to treatment and vector control in the south. Giorgi drew on [R1] to fit spatio-temporal MBG models to cross-sectional prevalence survey data collected between 2005 and 2015 to identify regions highly likely to have malaria prevalence above or below 1% (Figure 2, left-hand panel). Giorgi also ran the analysis for Sudan [S2] using data from cross-sectional surveys conducted between 2000 and 2016 (Figure 2, right-hand panel).

Figure 2. Left-hand panel: classification of the 18 regions in Somalia based on their estimated probability of malaria prevalence below the 1% threshold, as indicated by the legend. Right-hand panel: classification of areas of Sudan according to their probability of low prevalence (northern Sudan, shades of green) or high prevalence (southern Sudan, shades of red), as indicated by the legend.

The maps shown in Figure 2 allowed the distinction of areas highly likely to have achieved public health policy targets and areas that require additional sampling effort to be classified correctly, which triggered a significant shift in the decision-making process, which was previous informed by maps of predicted malaria prevalence that did not acknowledge the inherent uncertainty. This provided evidence for public health policy decisions in Somalia, Kenya and other countries. In the words of the Head of the Strategic Information Unit at the Global Malaria Programme, World Health Organization, “ Professor Peter Diggle and Dr Emanuele Giorgi’s work has not only been extremely helpful to analysing the spatial and temporal changes of malaria in Somalia and Kenya as well other African countries, but it has also been instrumental in changing policy and practice through the development of national strategic plans for malaria for these countries. This is a clear demonstration of how research evidence can immediately translate in policies and action in countries to reduce the burden of malaria” [S3].

In addition to the research being directly adopted by national programmes for malaria in the worst affected countries in Africa, Giorgi and Diggle’s research and outreach has also supported capacity-building within resource-constrained countries in tackling other public health priorities, leading in particular to impact in Kenya where vaccination policies have been shaped by work on infant mortality undertaken by African statisticians [S3].

4.2 Onchocerciasis

Since 2015, the focus of Lancaster research on onchocerciasis control has been to develop methods for the optimal combination of mobile microscopy and a rapid diagnostic test to determine which areas are and are not safe for mass drug administration. In 2015, Diggle began a formal collaboration with the NTD Centre of the Task Force for Global Health (TFGH). Predictions from geospatial statistical models were validated by in-country data from Nigeria, Cameroon and Gabon. As a result, the prediction algorithm has been incorporated into the LoaScope devices procured by the Gates Foundation and has been considered a viable strategy that the TFGH and its partners are interested in piloting in other global NTD-affected areas [S4]. In January 2020, the TFGH NTD Centre Director of Research stated that Diggle and Giorgi’s work has been “ instrumental in developing a comprehensive Loa loa mapping strategy at the continental scale” [S4]. The theoretical contributions [R1] have been used to develop new analytic tools that avoid the need to test every individual with the LoaScope diagnostic, enabling an estimated resource savings of USD100 million [S4]. In addition, according to the Director of Research, TFGH NTD Centre, due to Diggle and Giorgi’s contributions to the development of these analytic tools, “ *onchocerciasis may be possible to eliminate in areas where Loa Loa is co-endemic (a 10 country region with over 85 million individuals)*” [S4].

4.3 Trachoma

In 2018, Diggle and Giorgi began working with the WHO's Global Trachoma Elimination Programme and used their MBG methods and software [R1, R8], to provide estimates of the burden and fine-scale spatial distribution of Trachomatous trichiasis cases throughout Ethiopia. The importance of this work was summarised by the Trachoma Medical Officer of the Department of Control of Neglected Tropical Diseases (DCNTD) in 2019, “ It will allow highly resource-restricted national trachoma programmes, currently active in more than 50 countries, to better target services to deliver sight-saving operations. These operations are needed by 2.5 million of the world’s poorest and more marginalized individuals” [S5].

The Ethiopian analysis provides an example of the direct use of Diggle and Giorgi’s developed analytical approaches for trachoma control, as confirmed in 2020 by the Trachoma Medical Officer, who stated, “ Geostatistical approaches have been applied to all available current data on trichiasis in Ethiopia, the world’s most heavily trachoma-affected country. They were used to calculate the probability that the TT component of trachoma’s elimination as a public health problem has already been achieved, for each of 482 evaluation units, finding that it had been with probability 0.9 or more in only eight of those evaluation units. In other words, continuing heavy investment in trichiasis surgery programmes is needed in that country” [S5].

4.4 Soil-transmitted helminths (STH)

In 2017, Diggle began working with the Sri Lankan Ministry of Health to provide research techniques to inform their de-worming strategy. The Senior Professor of Parasitology at the University of Kelaniya, responsible for the oversight of the Ministry’s de-worming programme, stated in 2020, “ As a result of using the geostatistical modelling techniques that Prof Diggle have equipped us with, the Ministry of Health in Sri Lanka has revised its approach to deworming for STH infections in the entire country. The design of this new strategy would not have been possible without using the modelling techniques developed by Prof Diggle, and without his input, the revised strategy would have resulted in either overtreatment (with unnecessary use of anthelmintics) or undertreatment, resulting in rebound in the prevalence of STH infections in Sri Lanka” [S6].

All the software developed for these analyses is publicly available as an open source R package, PrevMap, with associated tutorial material [G1, R8]. Between January 2016 and December 2020 the package was downloaded 25,965 times [S7].