Журнал биометрии и биостатистики

Detection of Thyroid Cancer Clusters in Algeria

Abstract

Oumelkheir Moussi, Boudrissa N, Bouakline S, Semrouni M and Hasbellaoui F

In the mid-seventies, the Pierre and Marie Curie Center of Algiers (CPMC), was the only structure at the national level to take care of patients with thyroid cancer, it recorded from 15 to 20 cases per year. Today, and although Algeria has ten structures for the management of these patients, CPMC has recorded more than 100 new cases of thyroid cancer per year over the period 2007/2011 This disease is the third most common cancer for women. These observations lead us to ask several questions: • Are there wilayas having an excessive number of cancer’s cases? • Is cases’ concentration abnormally? • Is the spatial distribution of these cases random? Answering these questions is by describing the spatial heterogeneity, that is to say, identifying potential spatial clusters. A cluster is spatial organization defined as an aggregation, a combination of the nearest cases, the proximity being defined in the sense of geographical distance. Various statistical methods have been used to study the spatial heterogeneity. The global methods, for the detection of clusters, the study of the spatial correlation, and the detection of cases tending to "clustering", and the local methods, for the identification of clusters of cases inconsistent under the null hypothesis of no clusters, and the evaluation of their significance level. To study the spatial distribution of the disease, several tests were applied, such as the test of Pearson, the index of Moran, the Tango test, and Pothoff-Whittinghill test to check the hypothesis of constant risk of the incidence of thyroid cancer and study the tendency to "clustering", next the "Two-step clustering", was used for the localization of the clusters, and the scanning Kulldorf test to identify potential clusters, confirm and assess their significance.