Springer Science and Business Media LLC

Following World Health Assembly resolutions 50.36 in 1997 and 56.7 in 2003, the World Health Organization (WHO) committed itself to supporting human African trypanosomiasis (HAT)-endemic countries in their efforts to remove the disease as a public health problem. Mapping the distribution of HAT in time and space has a pivotal role to play if this objective is to be met. For this reason WHO launched the HAT Atlas initiative, jointly implemented with the Food and Agriculture Organization of the United Nations, in the framework of the Programme Against African Trypanosomosis.

The distribution of HAT is presented for 23 out of 25 sub-Saharan countries having reported on the status of sleeping sickness in the period 2000 - 2009. For the two remaining countries, i.e. Angola and the Democratic Republic of the Congo, data processing is ongoing. Reports by National Sleeping Sickness Control Programmes (NSSCPs), Non-Governmental Organizations (NGOs) and Research Institutes were collated and the relevant epidemiological data were entered in a database, thus incorporating (i) the results of active screening of over 2.2 million people, and (ii) cases detected in health care facilities engaged in passive surveillance. A total of over 42 000 cases of HAT and 6 000 different localities were included in the database. Various sources of geographic coordinates were used to locate the villages of epidemiological interest. The resulting average mapping accuracy is estimated at 900 m.

Full involvement of NSSCPs, NGOs and Research Institutes in building the Atlas of HAT contributes to the efficiency of the mapping process and it assures both the quality of the collated information and the accuracy of the outputs. Although efforts are still needed to reduce the number of undetected and unreported cases, the comprehensive, village-level mapping of HAT control activities over a ten-year period ensures a detailed and reliable representation of the known geographic distribution of the disease. Not only does the Atlas serve research and advocacy, but, more importantly, it provides crucial evidence and a valuable tool for making informed decisions to plan and monitor the control of sleeping sickness.

Cancer mortality maps are used by public health officials to identify areas of excess and to guide surveillance and control activities. Quality of decision-making thus relies on an accurate quantification of risks from observed rates which can be very unreliable when computed from sparsely populated geographical units or recorded for minority populations. This paper presents a geostatistical methodology that accounts for spatially varying population sizes and spatial patterns in the processing of cancer mortality data. Simulation studies are conducted to compare the performances of Poisson kriging to a few simple smoothers (i.e. population-weighted estimators and empirical Bayes smoothers) under different scenarios for the disease frequency, the population size, and the spatial pattern of risk. A public-domain executable with example datasets is provided.

The analysis of age-adjusted mortality rates for breast and cervix cancers illustrated some key features of commonly used smoothing techniques. Because of the small weight assigned to the rate observed over the entity being smoothed (kernel weight), the population-weighted average leads to risk maps that show little variability. Other techniques assign larger and similar kernel weights but they use a different piece of auxiliary information in the prediction: global or local means for global or local empirical Bayes smoothers, and spatial combination of surrounding rates for the geostatistical estimator. Simulation studies indicated that Poisson kriging outperforms other approaches for most scenarios, with a clear benefit when the risk values are spatially correlated. Global empirical Bayes smoothers provide more accurate predictions under the least frequent scenario of spatially random risk.

The approach presented in this paper enables researchers to incorporate the pattern of spatial dependence of mortality rates into the mapping of risk values and the quantification of the associated uncertainty, while being easier to implement than a full Bayesian model. The availability of a public-domain executable makes the geostatistical analysis of health data, and its comparison to traditional smoothers, more accessible to common users. In future papers this methodology will be generalized to the simulation of the spatial distribution of risk values and the propagation of the uncertainty attached to predicted risks in local cluster analysis.