Emanuele Giorgi

The geographic distribution of onchocerciasis in the 20 participating countries of the African Programme for Onchocerciasis Control: (2) pre-control endemicity levels and estimated number infected

BackgroundThe original aim of the African Programme for Onchocerciasis Control (APOC) was to control onchocerciasis as a public health problem in 20 African countries. In order to identify all high risk areas where ivermectin treatment was needed to achieve control, APOC used Rapid Epidemiological Mapping of Onchocerciasis (REMO). REMO involved spatial sampling of villages to be surveyed, and examination of 30 to 50 adults per village for palpable onchocercal nodules. REMO has now been virtually completed and we report the results in two articles. A companion article reports the delineation of high risk areas based on expert analysis. The present article reports the results of a geostatistical analysis of the REMO data to map endemicity levels and estimate the number infected.MethodsA model-based geostatistical analysis of the REMO data was undertaken to generate high-resolution maps of the predicted prevalence of nodules and of the probability that the true nodule prevalence exceeds the high risk threshold of 20%. The number infected was estimated by converting nodule prevalence to microfilaria prevalence, and multiplying the predicted prevalence for each location with local data on population density. The geostatistical analysis included the nodule palpation data for 14,473 surveyed villages.ResultsThe generated map of onchocerciasis endemicity levels, as reflected in the prevalence of nodules, is a significant advance with many new endemic areas identified. The prevalence of nodules was > 20% over an area of 2.5 million km2 with an estimated population of 62 million people. The results were consistent with the delineation of high risk areas of the expert analysis except for borderline areas where the prevalence fluctuated around 20%. It is estimated that 36 million people would have been infected in the APOC countries by 2011 if there had been no ivermectin treatment.ConclusionsThe map of onchocerciasis endemicity levels has proven very valuable for onchocerciasis control in the APOC countries. Following the recent shift to onchocerciasis elimination, the map continues to play an important role in planning treatment, evaluating impact and predicting treatment end dates in relation to local endemicity levels.

Mapping the Risk of Snakebite in Sri Lanka - A National Survey with Geospatial Analysis.

Background There is a paucity of robust epidemiological data on snakebite, and data available from hospitals and localized or time-limited surveys have major limitations. No study has investigated the incidence of snakebite across a whole country. We undertook a community-based national survey and model based geostatistics to determine incidence, envenoming, mortality and geographical pattern of snakebite in Sri Lanka. Methodology/Principal Findings The survey was designed to sample a population distributed equally among the nine provinces of the country. The number of data collection clusters was divided among districts in proportion to their population. Within districts clusters were randomly selected. Population based incidence of snakebite and significant envenoming were estimated. Model-based geostatistics was used to develop snakebite risk maps for Sri Lanka. 1118 of the total of 14022 GN divisions with a population of 165665 (0.8%of the country’s population) were surveyed. The crude overall community incidence of snakebite, envenoming and mortality were 398 (95% CI: 356–441), 151 (130–173) and 2.3 (0.2–4.4) per 100000 population, respectively. Risk maps showed wide variation in incidence within the country, and snakebite hotspots and cold spots were determined by considering the probability of exceeding the national incidence. Conclusions/Significance This study provides community based incidence rates of snakebite and envenoming for Sri Lanka. The within-country spatial variation of bites can inform healthcare decision making and highlights the limitations associated with estimates of incidence from hospital data or localized surveys. Our methods are replicable, and these models can be adapted to other geographic regions after re-estimating spatial covariance parameters for the particular region.

Dynamic monitors of brain function: a new target in neurointensive care unit.

IntroductionSomatosensory evoked potential (SEP) recordings and continuous electroencephalography (EEG) are important tools with which to predict Glasgow Outcome Scale (GOS) scores. Their combined use may potentially allow for early detection of neurological impairment and more effective treatment of clinical deterioration.MethodsWe followed up 68 selected comatose patients between 2007 and 2009 who had been admitted to the Neurosurgical Intensive Care Unit of Treviso Hospital after being diagnosed with subarachnoid haemorrhage (51 cases) or intracerebral haemorrhage (17 cases). Quantitative brain function monitoring was carried out using a remote EEG-SEP recording system connected to a small amplification head box with 28 channels and a multimodal stimulator (NEMO; EBNeuro, Italy NeMus 2; EBNeuro S.p.A., Via P. Fanfani 97/A - 50127 Firenze, Italy). For statistical analysis, we fit a binary logistic regression model to estimate the effect of brain function monitoring on the probability of GOS scores equal to 1. We also designed a proportional odds model for GOS scores, depending on amplitude and changes in both SEPs and EEG as well as on the joint effect of other related variables. Both families of models, logistic regression analysis and proportional odds ratios, were fit by using a maximum likelihood test and the partial effect of each variable was assessed by using a likelihood ratio test.ResultsUsing the logistic regression model, we observed that progressive deterioration on the basis of EEG was associated with an increased risk of dying by almost 24% compared to patients whose condition did not worsen according to EEG. SEP decreases were also significant; for patients with worsening SEPs, the odds of dying increased to approximately 32%. In the proportional odds model, only modifications of Modified Glasgow Coma Scale scores and SEPs during hospitalisation statistically significantly predicted GOS scores. Patients whose SEPs worsened during the last time interval had an approximately 17 times greater probability of a poor GOS score compared to the other patients.ConclusionsThe combined use of SEPs and continuous EEG monitoring is a unique example of dynamic brain monitoring. The temporal variation of these two parameters evaluated by continuous monitoring can establish whether the treatments used for patients receiving neurocritical care are properly tailored to the neurological changes induced by the lesions responsible for secondary damage.

Model-Based Geostatistics for Prevalence Mapping in Low-Resource Settings

ABSTRACT In low-resource settings, prevalence mapping relies on empirical prevalence data from a finite, often spatially sparse, set of surveys of communities within the region of interest, possibly supplemented by remotely sensed images that can act as proxies for environmental risk factors. A standard geostatistical model for data of this kind is a generalized linear mixed model with binomial error distribution, logistic link, and a combination of explanatory variables and a Gaussian spatial stochastic process in the linear predictor. In this article, we first review statistical methods and software associated with this standard model, then consider several methodological extensions whose development has been motivated by the requirements of specific applications. These include: methods for combining randomized survey data with data from nonrandomized, and therefore potentially biased, surveys; spatio-temporal extensions; and spatially structured zero-inflation. Throughout, we illustrate the methods with disease mapping applications that have arisen through our involvement with a range of African public health programs.

Mapping the geographical distribution of podoconiosis in Cameroon using parasitological, serological, and clinical evidence to exclude other causes of lymphedema

Background Podoconiosis is a non-filarial elephantiasis, which causes massive swelling of the lower legs. It was identified as a neglected tropical disease by WHO in 2011. Understanding of the geographical distribution of the disease is incomplete. As part of a global mapping of podoconiosis, this study was conducted in Cameroon to map the distribution of the disease. This mapping work will help to generate data on the geographical distribution of podoconiosis in Cameroon and contribute to the global atlas of podoconiosis. Methods We used a multi‐stage sampling design with stratification of the country by environmental risk of podoconiosis. We sampled 76 villages from 40 health districts from the ten Regions of Cameroon. All individuals of 15-years old or older in the village were surveyed house-to-house and screened for lymphedema. A clinical algorithm was used to reliably diagnose podoconiosis, excluding filarial-associated lymphedema. Individuals with lymphoedema were tested for circulating Wuchereria bancrofti antigen and specific IgG4 using the Alere Filariasis Test Strips (FTS) test and the Standard Diagnostics (SD) BIOLINE lymphatic filariasis IgG4 test (Wb123) respectively, in addition to thick blood films. Presence of DNA specific to W. bancrofti was checked on night blood using a qPCR technique. Principal findings Overall, 10,178 individuals from 4,603 households participated in the study. In total, 83 individuals with lymphedema were identified. Of the 83 individuals with lymphedema, two were found to be FTS positive and all were negative using the Wb123 test. No microfilaria of W. bancrofti were found in the night blood of any individual with clinical lymphedema. None were found to be positive for W. bancrofti using qPCR. Of the two FTS positive cases, one was positive for Mansonella perstans DNA, while the other harbored Loa loa microfilaria. Overall, 52 people with podoconiosis were identified after applying the clinical algorithm. The overall prevalence of podoconiosis was found to be 0.5% (95% [confidence interval] CI; 0.4–0.7). At least one case of podoconiosis was found in every region of Cameroon except the two surveyed villages in Adamawa. Of the 40 health districts surveyed, 17 districts had no cases of podoconiosis; in 15 districts, mean prevalence was between 0.2% and 1.0%; and in the remaining eight, mean prevalence was between 1.2% and 2.7%. Conclusions Our investigation has demonstrated low prevalence but almost nationwide distribution of podoconiosis in Cameroon. Designing a podoconiosis control program is a vital next step. A health system response to the burden of podoconiosis is important, through case surveillance and morbidity management services.

Estimating the number of cases of podoconiosis in Ethiopia using geostatistical methods.

Background: In 2011, the World Health Organization recognized podoconiosis as one of the neglected tropical diseases. Nonetheless, the magnitude of podoconiosis and the geographical distribution of the disease is poorly understood. Based on a nationwide mapping survey and geostatistical modelling, we predict the prevalence of podoconiosis and estimate the number of cases across Ethiopia. Methods: We used nationwide data collected in Ethiopia between 2008 and 2013. Data were available for 141,238 individuals from 1,442 communities in 775 districts from all nine regional states and two city administrations. We developed a geostatistical model of podoconiosis prevalence among adults (individuals aged 15 years or above), by combining environmental factors. The number of people with podoconiosis was then estimated using a gridded map of adult population density for 2015. Results: Podoconiosis is endemic in 345 districts in Ethiopia: 144 in Oromia, 128 in Southern Nations, Nationalities and People’s [SNNP], 64 in Amhara, 4 in Benishangul Gumuz, 4 in Tigray and 1 in Somali Regional State. Nationally, our estimates suggest that 1,537,963 adults (95% confidence intervals, 290,923-4,577,031 adults) were living with podoconiosis in 2015. Three regions (SNNP, Oromia and Amhara) contributed 99% of the cases. The highest proportion of individuals with podoconiosis resided in the SNNP (39%), while 32% and 29% of people with podoconiosis resided in Oromia and Amhara Regional States, respectively. Tigray and Benishangul Gumuz Regional States bore lower burdens, and in the remaining regions, podoconiosis was almost non-existent. Conclusions: The estimates of podoconiosis cases presented here based upon the combination of currently available epidemiological data and a robust modelling approach clearly show that podoconiosis is highly endemic in Ethiopia. Given the presence of low cost prevention, and morbidity management and disability prevention services, it is our collective responsibility to scale-up interventions rapidly.

Impact of metric and sample size on determining malaria hotspot boundaries

The spatial heterogeneity of malaria suggests that interventions may be targeted for maximum impact. It is unclear to what extent different metrics lead to consistent delineation of hotspot boundaries. Using data from a large community-based malaria survey in the western Kenyan highlands, we assessed the agreement between a model-based geostatistical (MBG) approach to detect hotspots using Plasmodium falciparum parasite prevalence and serological evidence for exposure. Malaria transmission was widespread and highly heterogeneous with one third of the total population living in hotspots regardless of metric tested. Moderate agreement (Kappa = 0.424) was observed between hotspots defined based on parasite prevalence by polymerase chain reaction (PCR)- and the prevalence of antibodies to two P. falciparum antigens (MSP-1, AMA-1). While numerous biologically plausible hotspots were identified, their detection strongly relied on the proportion of the population sampled. When only 3% of the population was sampled, no PCR derived hotspots were reliably detected and at least 21% of the population was needed for reliable results. Similar results were observed for hotspots of seroprevalence. Hotspot boundaries are driven by the malaria diagnostic and sample size used to inform the model. These findings warn against the simplistic use of spatial analysis on available data to target malaria interventions in areas where hotspot boundaries are uncertain.

Geostatistical Methods for Disease Mapping and Visualisation Using Data from Spatio-temporally Referenced Prevalence Surveys: Geostatistical Methods for Prevalence Mapping

In this paper, we set out general principles and develop geostatistical methods for the analysis of data from spatio-temporally referenced prevalence surveys. Our objective is to provide a tutorial guide that can be used in order to identify parsimonious geostatistical models for prevalence mapping. A general variogram-based Monte Carlo procedure is proposed to check the validity of the modelling assumptions. We describe and contrast likelihood-based and Bayesian methods of inference, showing how to account for parameter uncertainty under each of the two paradigms. We also describe extensions of the standard model for disease prevalence that can be used when stationarity of the spatio-temporal covariance function is not supported by the data. We discuss how to define predictive targets and argue that exceedance probabilities provide one of the most effective ways to convey uncertainty in prevalence estimates. We describe statistical software for the visualisation of spatio-temporal predictive summaries of prevalence through interactive animations. Finally, we illustrate an application to historical malaria prevalence data from 1 334 surveys conducted in Senegal between 1905 and 2014.

Household air pollution, chronic respiratory disease and pneumonia in Malawian adults: A case-control study

Background: Four million people die each year from diseases caused by exposure to household air pollution. There is an association between exposure to household air pollution and pneumonia in children (half a million attributable deaths a year); however, whether this is true in adults is unknown. We conducted a case-control study in urban Malawi to examine the association between exposure to household air pollution and pneumonia in adults. Methods: Hospitalized patients with radiologically confirmed pneumonia (cases) and healthy community controls underwent 48 hours of ambulatory and household particulate matter (µg/m 3) and carbon monoxide (ppm) exposure monitoring. Multivariate logistic regression, stratified by HIV status, explored associations between these and other potential risk factors with pneumonia. Results: 145 (117 HIV-positive; 28 HIV-negative) cases and 253 (169 HIV-positive; 84 HIV-negative) controls completed follow up. We found no evidence of association between household air pollution exposure and pneumonia in HIV-positive (e.g. ambulatory particulate matter adjusted odds ratio [aOR] 1.00 [95% CI 1.00–1.01, p=0.141]) or HIV-negative (e.g. ambulatory particulate matter aOR 1.00 [95% CI 0.99–1.01, p=0.872]) participants. Chronic respiratory disease was associated with pneumonia in both HIV-positive (aOR 28.07 [95% CI 9.29–84.83, p<0.001]) and HIV-negative (aOR 104.27 [95% CI 12.86–852.35, p<0.001]) participants. Conclusions: We found no evidence that exposure to household air pollution is associated with pneumonia in Malawian adults. In contrast, chronic respiratory disease was strongly associated with pneumonia.

On the computation of multivariate scenario sets for the skew- t and generalized hyperbolic families

The problem of computing multivariate scenarios sets for skewed distributions is motivated by the potential use of such sets in the stress testing of insurance companies and banks. Multivariate scenario sets based on the notion of half-space depth (HD) are considered and the notion of expectile depth (ED) is introduced. These depth concepts facilitate the definition of convex scenario sets, which generalize the concepts of quantiles and expectiles to higher dimensions. In the case of elliptical distributions the scenario sets coincide with the regions encompassed by the contours of the density function. In the context of multivariate skewed distributions, the equivalence of depth contours and density contours does not hold in general. Two parametric families that account for skewness and heavy tails are analysed: the generalized hyperbolic and the skew- t distributions. By making use of a canonical form representation, where skewness is completely absorbed by one component, it is shown that the HD contours of these distributions are near-elliptical; in the case of the skew-Cauchy distribution the HD contours are exactly elliptical. A measure of multivariate skewness as a deviation from angular symmetry is proposed. This measure is shown to explain the quality of the elliptical approximation for the HD contours.