scispace - formally typeset
Search or ask a question
Journal ArticleDOI

A Tale of Two Parasites: Statistical Modelling to Support Disease Control Programmes in Africa

01 Feb 2020-Statistical Science (Institute of Mathematical Statistics)-Vol. 35, Iss: 1, pp 42-50
TL;DR: A model-based geostatistical analysis of Loa loa prevalence survey data can be used to map the predictive probability that each location in the region of interest meets a WHO policy guideline for safe mass distribution of ivermectin and describe two applications: one is to data from Cameroon that assesses prevalence using traditional blood-smear microscopy; the other is to Africa-wide data that uses a low-cost questionnaire-based method.
Abstract: Vector-borne diseases have long presented major challenges to the health of rural communities in the wet tropical regions of the world, but especially in sub-Saharan Africa. In this paper, we describe the contribution that statistical modelling has made to the global elimination programme for one vector-borne disease, onchocerciasis. We explain why information on the spatial distribution of a second vector-borne disease, Loa loa, is needed before communities at high risk of onchocerciasis can be treated safely with mass distribution of ivermectin, an antifiarial medication. We show how a model-based geostatistical analysis of Loa loa prevalence survey data can be used to map the predictive probability that each location in the region of interest meets a WHO policy guideline for safe mass distribution of ivermectin and describe two applications: one is to data from Cameroon that assesses prevalence using traditional blood-smear microscopy; the other is to Africa-wide data that uses a low-cost questionnaire-based method. We describe how a recent technological development in image-based microscopy has resulted in a change of emphasis from prevalence alone to the bivariate spatial distribution of prevalence and the intensity of infection among infected individuals. We discuss how statistical modelling of the kind described here can contribute to health policy guidelines and decision-making in two ways. One is to ensure that, in a resource-limited setting, prevalence surveys are designed, and the resulting data analysed, as efficiently as possible. The other is to provide an honest quantification of the uncertainty attached to any binary decision by reporting predictive probabilities that a policy-defined condition for action is or is not met.

Summary (2 min read)

Introduction

  • Vector-borne diseases have long presented major challenges to the health of rural communities in the wet tropical regions of the world, but especially in sub-Saharan Africa.
  • The authors explain why information on the spatial distribution of a second vector-borne disease, Loa loa, is needed before communities at high risk of onchocerciasis can be treated safely with mass distribution of ivermectin, an antifiarial medication.
  • One is to ensure that, in a resource-limited setting, prevalence surveys are designed, and the resulting data analysed, as efficiently as possible.
  • The authors describe how a recent technological development in image-based microscopy has resulted in a change of emphasis from prevalence alone to the bivariate spatial distribution of prevalence and the intensity of infection amongst infected individuals.

1 Problem statement

  • Vector-borne diseases have long presented major challenges to the health of rural communities in the wet tropical regions of the world, but especially in sub-Saharan Africa.
  • As a consequence, in 2012 the World Health Organization recognized that elimination is the appropriate target for onchocerciasis in Africa, where 99% of the world’s population at risk live.
  • The presence of Loa loa has stymied the elimination effort in areas where the two parasites are co-endemic, as there has not been an agreed upon strategy for how to deliver safe treatment with ivermectin.
  • To give more robust inferences about the effects of the environmental risk-factors and, more importantly in this application, to give predictions of prevalence at unsampled locations that are both more accurate than can be obtained using only the available environmental risk-factors and more honest in their associated measures of uncertainty, also known as Its purpose is two-fold.

3 Scaling up: the RAPLOA method

  • The conclusion from the Diggle, Thomson et al (2007) analysis is that the intensity of datacollection in Cameroon is insufficient to delineate the whole of the country into “safe” and “unsafe” areas for MDA.
  • But expanding even this intensity of data collection to all of the APOC countries would have been unaffordable.
  • Declaring a positive RAPLOA result as three ”yes” answers gives village-level prevalence estimates that correlate well with estimates based on blood smear microscopy, at a fraction of the cost.

4 Joint analysis of prevalence and intensity of infection:

  • The LoaScope A major limitation of prevalence mapping as a way of identifying areas that are and are not considered to be safe for MDA with ivermectin is that village-level prevalence is at best a proxy for the proportion of village inhabitants who are heavily infected with Loa loa, and therefore at risk of experiencing an SAE when given ivermectin.
  • This implies that it is possible to predict the proportion of highly infected, and therefore high-risk, individuals in a village, given only the presence or absence of Loa loa microfilariae in blood smears from a sample of village inhabitants.
  • Model-based predictions are, however, less robust than empirical ones, leading to an understandable reluctance to adopt them for routine use.
  • The practical significance of this is that it enables individual-level testing to identify high-risk individuals at the point of care and in real-time, without the need for trained microscopists.

5 Combining information from multiple diagnostics

  • When different diagnostics have been used in the same geographical region, statistical models that can combine information from multiple diagnostics have the potential to make better use of all available data sources.
  • Amoah et al (2018) introduce a geostatistical framework for combining data from multiple diagnostics and apply this to prevalence data obtained by blood-smear microscopy and by RAPLOA.
  • S1(x) and S2(x) are a pair of independent Gaussian processes that are used to account for unexplained spatial variation specific to each diagnostic.
  • Against this, its practical utility in low-resource settings may be constrained by computational and data-storage requirements.

6 Translating research into policy: decision-making un-

  • The expansion of ivermectin MDA into areas where onchocerciasis sequelae are less prevalent brings into sharp focus the ethical dilemma of an acceptable SAE risk.
  • Test-and-not-treat undoubtedly gives the best protection against the occurrence of SAEs, which are not only catastrophic for the individuals concerned but harm the acceptability of ivermectin to their communities.
  • Firstly, given data and an agreed definition of “safe” at an agreed geographical scale, statistical modelling in conjunction with predictive infererence can deliver an honest assessment of the probability that the safety threshold has been reached.
  • Secondly, when the available data are insufficent to resolve the safety issue, statistical design can inform where addtional data should be collected.

7 Conclusion

  • At the time of writing, the “Loa loa problem” persists.
  • Statistical modelling and inference can inform, but not resolve, the problem.
  • Balancing the remote risk of causing severe damage to a relatively small number of individuals against the likelihood of protecting much larger numbers against life-changing disease is not something that can be done using only statistical or, more broadly, “scientific” arguments.
  • Social context must also be taken into account and it is imperative that the affected countries be a leading voice in these deliberations.
  • Nonethless, in the authors’ opinion it is essential that any debate aimed at achieving consensus is informed by the use of statistical methods that make the best possible use of relevant data whilst acknowledging their necessary limitations.

Did you find this useful? Give us your feedback

Figures (6)

Content maybe subject to copyright    Report

A Tale of Two Parasites: statistical modelling to support
disease control programmes in Africa
Peter J Diggle
(Lancaster University and Health Data Research UK)
Emanuele Giorgi
(Lancaster University),
Julienne Atsame (Ministry of Health, Gabon),
Sylvie Ntsame Ella (Ministry of Health, Gabon),
Kisito Ogoussan (Task Force for Global Health)
and
Katherine Gass
(Task Force for Global Health)
August 26, 2019
Abstract
Vector-borne diseases have long presented major challenges to the health of rural communities
in the wet tropical regions of the world, but especially in sub-Saharan Africa. In this paper
we describe the contribution that statistical modelling has made to the global elimination
programme for one vector-borne disease, onchocerciasis.
We explain why information on the spatial distribution of a second vector-borne disease, Loa
loa, is needed before communities at high risk of onchocerciasis can be treated safely with
mass distribution of ivermectin, an antifiarial medication.
We show how a model-based geostatistical analysis of Loa loa prevalence survey data can be
used to map the predictive probability that each location in the region of interest meets a WHO
policy guideline for safe mass distribution of ivermectin and describe two applications: one
is to data from Cameroon that assesses prevalence using traditional blood-smear microscopy;
the other is to Africa-wide data that uses a low-cost questionnaire-based method.
We describe how a recent technological development in image-based microscopy has resulted
in a change of emphasis from prevalence alone to the bivariate spatial distribution of preva-
lence and the intensity of infection amongst infected individuals. We discuss how statistical
1

modelling of the kind described here can contribute to health policy guidelines and decision-
making in two ways. One is to ensure that, in a resource-limited setting, prevalence surveys
are designed, and the resulting data analysed, as efficiently as possible. The other is to provide
an honest quantifi
cation of the uncertainty attached to any binary decision by reporting predictive probabilities
that a policy-de
fined condition for action is or is not met.
Vector-borne diseases have long presented major challenges to the health of rural communities
in the wet tropical regions of the world, but especially in sub-Saharan Africa. In this paper
we describe the contribution that statistical modelling has made to the global elimination
programme for one vector-borne disease, onchocerciasis.
We explain why information on the spatial distribution of a second vector-borne disease, Loa
loa, is needed before communities at high risk of onchocerciasis can be treated safely with
mass distribuiton of ivermectin, an antiflarial medication.
We show how a model-based geostatistical analysis of Loa loa prevalence survey data can be
used to map the predictive probability that each location in the region of interest meets a
WHO policy guideline for safe mass distribution of ivermectin and describe two applications:
one to data from Cameroon that assesses prevalence using traditional blood-smear microscopy;
one to Africa-wide data that uses a low-cost questionnaire-based method.
We describe how a recent technological development in image-based microscopy has resulted in
a change of emphasis from prevalence alone to the bivariate spatial distribution of prevalence
and the intensity of infection amongst infected individuals.
We discuss how statistical modelling of the kind described here can contribute to health
policy guidelines and decision-making in two ways. One is to ensure that, in a resource-
limited setting, prevalece surveys are designed, and the resulting data analysed, as efficiently
as possible. The other is to provide an honest quantification of the uncertainy attached to
any binary decision by reporting predictive probabilities that a policy-defined condition for
action is or is not met.
1 Problem statement
Vector-borne diseases have long presented major challenges to the health of rural communities
in the wet tropical regions of the world, but especially in sub-Saharan Africa. In this paper
we describe the contribution that statistical modelling has made to the global elimination
programme for one vector-borne disease, onchocerciasis.
Onchocerciasis was not always targeted for elimination. When the International Task Force
for Disease Eradication met in 1993, onchocerciasis was not among the six diseases considered
eradicable at the time (Centers for Disease Control and Prevention, 1993). Indeed, the On-
chocerciasis Control Program (OCP), and the African Programme for Onchocerciasis Control
2

(APOC) that succeeded it, focused on controlling significant disease of the skin and eyes so that
onchocerciasis would no longer be a public health problem (Amazigo, 2008). That strategy for
control relied on mass drug administration (MDA) with ivermectin, an antifilarial medication
that kills larval-stage parasites (microfilariae) before they can either cause significant disease
in infected people or be transmitted to the blackfly vectors of infection. Ivermectin, under its
trade name MECTIZAN
R
, has been provided free of charge since 1987 by its manufacturer,
Merck & Co Inc, to control onchocerciasis in all affected countries world-wide
1
. Even so, the
successful execution of a control programme using mass drug administration with ivermectin
(henceforth, MDA) faced formidable practical difficulties due to the size and inaccessibility of
the populations at risk and the low national income levels of many of the affected countries.
To address this challenge and ensure that even the most remote populations can be reached,
APOC developed a novel delivery strategy known as Community Directed Treatment with
Ivermectin (CDTI, Homeida et al, 2002), which has since been used by onchocerciasis pro-
grammes with great success.
The past two decades have seen tremendous success for onchocercaisis programmes. In the
Americas, where the infection is more focal and political will is strong, onchocerciasis has
now been eliminated in 11 of 13 foci of infection over 6 countries (World Health Organisation,
2013, 2014, 2015; Thiele et al, 2016). In Africa, over a dozen annual rounds of MDA led to
the interruption of transmission in selected foci in several countries including Mali, Senegal,
Uganda, Ethiopia, Nigeria and Sudan (World Health Organisation, 2018). Collectively, these
experiences demonstrated that the tools to detect and eliminate onchocerciasis infection exist
and can be used successfully by country programmes. As a consequence, in 2012 the World
Health Organization recognized that elimination is the appropriate target for onchocerciasis
in Africa, where 99% of the world’s population at risk live. This elimination effort is now
being coordinated by WHO’s regional office in Africa through their Expanded Special Project
for Elimination of Neglected Tropical Diseases (ESPEN, World Health Organisation, 2012).
In the midst of this great progress, there remains one insidious obstacle to the global elimina-
tion of onchocerciasis, namely the presence of Loa loa. Towards the end of the last century,
reports emerged of severe, sometimes fatal, reactions to ivermectin experienced by some in-
dividuals who were heavily infected with a second vector-borne parasitic infection, Loa loa
(Gardon et al, 1997; Boussinesq et al, 1998; Boussinesq et al, 2003). This discovery threat-
ened to de-rail the MDA progamme that until then had been regarded as safe. Because Loa
loa was not considered to be a major public health problem in its own right, there was limited
understanding of its geographical distribution in areas endemic for onchocerciasis, or of the
numbers of highly infected individuals considered to be at risk of experiencing serious adverse
reactions (henceforth, SAEs) to ivermectin. The presence of Loa loa has stymied the elimina-
tion effort in areas where the two parasites are co-endemic, as there has not been an agreed
upon strategy for how to deliver safe treatment with ivermectin. The remainder of the paper
describes how statistical modelling has been, and continues to be, used to address the Loa
loa problem,” and hence to inform the operational policy around MDA for the elimination of
onchocerciasis.
1
See: https://investors.merck.com/news/press-release-details/2017/Merck-Commemorates-30
-Years-of-MECTIZAN- Donation- Program-Progress/default.aspx
3

2 Mapping Loaloa parasitological prevalence
To make safe ivermectin treatment decisions for onchocerciasis in areas where Loa loa is po-
tentially co-endemic, it is first necessary to map the areas where heavily infected individuals,
that is individuals harboring an abundant number of Loa loa parasites, are likely to be found.
A general feature of the epidemiology of vector-borne diseases is that heavily infected individ-
uals are more likely to be found in high-prevalence areas. For evidence in the current context,
see Boussinesq et al (2001). For this reason, and because it would have been infeasible to
test individuals for their levels of infection with Loa loa parasites before administering iverem-
ctin, Thomson et al (2000) used data from Loa loa prevalence surveys in conjunction with
satellite-derived images of environmental variables to map the geographical variation in Loa
loa prevalence. They fitted a logistic regression model to data on the presence/absence of Loa
loa parasites under microscopic examination of blood smears for “14,305 individuals sampled
from more than 100 villages” in five Loa loa-endemic west and central African countries. As
explanatory variables, they used data at 1km pixel resolution on forest cover, land use type
and soil type. Their model explained approximately 50% of the variation in prevalence over
the sampled villages. Thomson et al (2004) also developed a logistic regression model for
Loa loa prevalence using data for 14,225 individuals from 94 villages in Cameroon together
with a wider range of explanatory variables; their model explained approximately 60% of the
variation in prevalence over the sampled villages.
Variation in prevalence of an environmentally driven disease typically obeys Tobler’s so-called
first law of geography, which states that “everything is related to everything else, but near
things are more related than distant things” (Tobler, 1970). The statistical expression of this
is that empirical prevalence data usually exhibit spatial correlation. To the extent that the
relevant environmental risk-factors can be measured, this spatial correlation can be removed
by covariate adjustment, but this typically leaves an unexplained component of geographical
variation that manifests itself in residual spatial correlation. One way to account for this
residual spatial correlation is to extend the logistic regression model by including a latent,
spatially correlated process in the linear predictor (Breslow and Clayton, 1993; Diggle, Moyeed
and Tawn, 1998). The resulting model for P (x), the prevalence of disease at location x, is
log[P (x)/{1 P (x)}] = d(x)
0
β + S(x), (1)
where d(x) is a vector of covariates associated with x and S(x) is a latent, spatially continuous
stochastic process. Conditional on the prevalence surface P(x) throughout the region of
interest, the numbers Y
i
of infected individuals out of m
i
tested at locations x
i
: i = 1, ..., n
are independent, binomially distributed random variables with binomial probabilities P (x
i
)
and denominators m
i
.
Diggle, Thomson et al (2007) used this approach to refine the predictions in Thomson et
al (2004), using prevalence data for 21,938 individuals from 168 villages in Cameroon and
Nigeria. The addition of the stochastic term S(x) on the right hand side of (??) helps to
explain the geographical variation in Loa loa prevalence. Its purpose is two-fold: to give more
robust inferences about the effects of the environmental risk-factors and, more importantly in
this application, to give predictions of prevalence at unsampled locations that are both more
4

Figure 1: Observed and predicted prevalences of Loa loa: (a) spatial model from Diggle,
Thomson et al (2007); (b) logistic regression model from Thomson et al (2004). Figure
reproduced from Diggle, Thomson et al, (2007).
accurate than can be obtained using only the available environmental risk-factors and more
honest in their associated measures of uncertainty. Diggle, Thomson et al (2007) examined
the residuals from a standard logistic regression model to suggest a suitable model for S(x).
This led them to specify S(x) to be a Gaussian process with covariance function
Cov{S(x), S(x
0
)} = σ
2
exp(−||x x
0
||) + τ
2
I(x = x
0
), (2)
where || · || denotes distance and I(·) is the indicator function. This model recognises two
sources of extra-binomial variation in the data, a spatial component σ
2
and a non-spatial
component τ
2
. Diggle, Thomson et al (2007) fitted the model using an MCMC implementation
of Bayesian inference, with the ratio τ
2
2
held fixed at 0.4 (owing to a limitation of the
software available to the authors at the time), an improper uniform prior for σ
2
and the
regression parameters β, and a proper uniform prior for the spatial correlation range parameter
φ U(0, c) with c = 100km. The explanatory variables in the model were elevation and two
satellite-derived variables associated with greenness of vegetation. The rationale for choosing
these was that the vector, a Chrysops fly species, can breed successfully in hot, wet conditions;
in equatorial regions, elevation acts as a proxy for minimum temperature.
The minimum mean square error predictor for P (x) is its conditional expectation given the
data. The improvement in accuracy of prediction over the logistic regression model of Thomson
et al (2004) is summarised by the two scatterplots of observed against predicted prevalence
shown in Figure ??. A conventional measure of the associated uncertainty would be the
conditional variance of P (x) given the data. Diggle, Thomson et al (2007) argue that a more
useful way to summarise uncertainty is to map selected points of the predictive distribution
of each P (x), i.e. its conditional distribution given the data. This was particularly relevant
5

References
More filters
Journal ArticleDOI
TL;DR: A Computer Movie Simulating Urban Growth in the Detroit Region as discussed by the authors was made to simulate urban growth in the city of Detroit, Michigan, United States of America, 1970, 1970.
Abstract: (1970). A Computer Movie Simulating Urban Growth in the Detroit Region. Economic Geography: Vol. 46, PROCEEDINGS International Geographical Union Commission on Quantitative Methods, pp. 234-240.

7,533 citations


"A Tale of Two Parasites: Statistica..." refers background in this paper

  • ...Variation in prevalence of an environmentally driven disease typically obeys Tobler’s so-called first law of geography, which states that “everything is related to everything else, but near things are more related than distant things” (Tobler, 1970)....

    [...]

Journal ArticleDOI
TL;DR: In this paper, generalized linear mixed models (GLMM) are used to estimate the marginal quasi-likelihood for the mean parameters and the conditional variance for the variances, and the dispersion matrix is specified in terms of a rank deficient inverse covariance matrix.
Abstract: Statistical approaches to overdispersion, correlated errors, shrinkage estimation, and smoothing of regression relationships may be encompassed within the framework of the generalized linear mixed model (GLMM). Given an unobserved vector of random effects, observations are assumed to be conditionally independent with means that depend on the linear predictor through a specified link function and conditional variances that are specified by a variance function, known prior weights and a scale factor. The random effects are assumed to be normally distributed with mean zero and dispersion matrix depending on unknown variance components. For problems involving time series, spatial aggregation and smoothing, the dispersion may be specified in terms of a rank deficient inverse covariance matrix. Approximation of the marginal quasi-likelihood using Laplace's method leads eventually to estimating equations based on penalized quasilikelihood or PQL for the mean parameters and pseudo-likelihood for the variances. Im...

4,317 citations


"A Tale of Two Parasites: Statistica..." refers methods in this paper

  • ...One way to account for this residual spatial correlation is to extend the logistic regression model by including a latent, spatially correlated process in the linear predictor (Breslow and Clayton, 1993; Diggle, Moyeed and Tawn, 1998)....

    [...]

Journal ArticleDOI
TL;DR: Empirical surveys aimed at assessing the intensity of infection with L loa microfilariae should be done before ivermectin is distributed for onchocerciasis control in areas where loiasis is endemic.

546 citations


"A Tale of Two Parasites: Statistica..." refers background in this paper

  • ...Towards the end of the last century, reports emerged of severe, sometimes fatal, reactions to ivermectin experienced by some individuals who were heavily infected with a second vector-borne parasitic infection, Loa loa (Gardon et al, 1997; Boussinesq et al, 1998; Boussinesq et al, 2003)....

    [...]

Journal ArticleDOI
TL;DR: The contour map of eye worm prevalence provides the first global map of loiasis based on actual survey data and provides critical information for ivermectin treatment programs among millions of people in Africa.
Abstract: Background: Loiasis is a major obstacle to ivermectin treatment for onchocerciasis control and lymphatic filariasis elimination in central Africa. In communities with a high level of loiasis endemicity, there is a significant risk of severe adverse reactions to ivermectin treatment. Information on the geographic distribution of loiasis in Africa is urgently needed but available information is limited. The African Programme for Onchocerciasis Control (APOC) undertook large scale mapping of loiasis in 11 potentially endemic countries using a rapid assessment procedure for loiasis (RAPLOA) that uses a simple questionnaire on the history of eye worm. Methodology/Principal Findings: RAPLOA surveys were done in a spatial sample of 4798 villages covering an area of 250063000 km centred on the heartland of loiasis in Africa. The surveys showed high risk levels of loiasis in 10 countries where an estimated 14.4 million people live in high risk areas. There was a strong spatial correlation among RAPLOA data, and kriging was used to produce spatially smoothed contour maps of the interpolated prevalence of eye worm and the predictive probability that the prevalence exceeds 40%. Conclusion/Significance: The contour map of eye worm prevalence provides the first global map of loiasis based on actual survey data. It shows a clear distribution with two zones of hyper endemicity, large areas that are free of loiasis and several borderline or intermediate zones. The surveys detected several previously unknown hyperendemic foci, clarified the distribution of loiasis in the Central African Republic and large parts of the Republic of Congo and the Democratic Republic of Congo for which hardly any information was available, and confirmed known loiasis foci. The new maps of the prevalence of eye worm and the probability that the prevalence exceeds the risk threshold of 40% provide critical information for ivermectin treatment programs among millions of people in Africa.

233 citations


"A Tale of Two Parasites: Statistica..." refers background in this paper

  • ...…RAPLOA and bood smear prevalences, whereas Wanji et al (2012) assumed a linear relationship between log-odds-transformed prevalences; see Figure ??. Zoure et al (2011) report on an Africa-wide study of RAPLOA prevalence, covering 381,575 individuals in 4,798 villages over an area of approximately…...

    [...]

Journal ArticleDOI
TL;DR: A mobile phone–based video microscope that automatically quantifies L. loa microfilariae in whole blood and excludes patients from ivermectin-based treatment at the point of care in Loa-endemic regions would allow resumption/expansion of mass drug administration programs for onchocerciasis and lymphatic filariasis in Central Africa.
Abstract: Parasitic helminths cause debilitating diseases that affect millions of people in primarily low-resource settings. Efforts to eliminate onchocerciasis and lymphatic filariasis in Central Africa through mass drug administration have been suspended because of ivermectin-associated serious adverse events, including death, in patients infected with the filarial parasite Loa loa. To safely administer ivermectin for onchocerciasis or lymphatic filariasis in regions coendemic with L. loa, a strategy termed “test and (not) treat” has been proposed whereby those with high levels of L. loa microfilariae (>30,000/ml) that put them at risk for life-threatening serious adverse events are identified and excluded from mass drug administration. To enable this, we developed a mobile phone–based video microscope that automatically quantifies L. loa microfilariae in whole blood loaded directly into a small glass capillary from a fingerprick without the need for conventional sample preparation or staining. This point-of-care device automatically captures and analyzes videos of microfilarial motion in whole blood using motorized sample scanning and onboard motion detection, minimizing input from health care workers and providing a quantification of microfilariae per milliliter of whole blood in under 2 min. To validate performance and usability of the mobile phone microscope, we tested 33 potentially Loa-infected patients in Cameroon and confirmed that automated counts correlated with manual thick smear counts (94% specificity; 100% sensitivity). Use of this technology to exclude patients from ivermectin-based treatment at the point of care in Loa-endemic regions would allow resumption/expansion of mass drug administration programs for onchocerciasis and lymphatic filariasis in Central Africa.

188 citations


"A Tale of Two Parasites: Statistica..." refers background in this paper

  • ...A recent and potentially game-changing technological development has been the invention of the LoaScope, a form of portable microscopy that uses image analysis of the movement of microfilariae in an individual’s blood sample to estimate their level of infection (D’Ambrosio et al, 2015)....

    [...]

Related Papers (5)
Frequently Asked Questions (1)
Q1. What contributions have the authors mentioned in the paper "A tale of two parasites: statistical modelling to support disease control programmes in africa" ?

In this paper the authors describe the contribution that statistical modelling has made to the global elimination programme for one vector-borne disease, onchocerciasis. The authors explain why information on the spatial distribution of a second vector-borne disease, Loa loa, is needed before communities at high risk of onchocerciasis can be treated safely with mass distribution of ivermectin, an antifiarial medication. The authors show how a model-based geostatistical analysis of Loa loa prevalence survey data can be used to map the predictive probability that each location in the region of interest meets a WHO policy guideline for safe mass distribution of ivermectin and describe two applications: one is to data from Cameroon that assesses prevalence using traditional blood-smear microscopy ; the other is to Africa-wide data that uses a low-cost questionnaire-based method. The authors describe how a recent technological development in image-based microscopy has resulted in a change of emphasis from prevalence alone to the bivariate spatial distribution of prevalence and the intensity of infection amongst infected individuals. The authors discuss how statistical