A malaria incidence model was developed. Data from 1996 to 1998 from the state of Amazonas were used, including: annual temperature (Max & Min), rainfall, chemical vector control variables such as the index of population protected by indoor house insecticide spraying (IHIS) and the index of population protected (IPP) by outdoor spraying, malaria incidence from previous years, and population size. We used a Poisson multiple regression model. Results: After testing a number of variable combinations, the best to predict the incidence of 1998 included the 1998 population, incidence for 1997, IHIS for 1998, squared of the IPP for 1998, and rainfall for 1998. The pseudo R2, the deviance and the relative difference between the observed vs. expected cases were used to assess the predictability of the model for internal evaluation. The best model yielded the following results: Pseudo R2=99.64% (Best possible: 100%), Deviance=8.51 (2 d.f.) (Best possible: 2), average of the difference between observed and expected cases: 10.02 cases or 3.65% difference (Best possible result: 0%). The proposed model demonstrated to have high predictability as it was able to closely predict the number of malaria cases. Unfortunately not enough solid data was available to conduct an external evaluation. Solid data is necessary to continue the development of the proposed model that will eventually be used to better understand the epidemiology of malaria. The model could also serve as an aid for malaria control decision-making as it could simulate various environmental, epidemiological and/or control situations.
Learning Objectives: At the end of the presentation the participant will be able to 1. Describe the basic epidemiologic features of malaria. 2. Define the descriptive epidemiology of malaria in Amazonas, Venezuela. 3. Understand the impact of environmental and vector control variables in the incidence of malaria. 4. Understand how to use a simple Poisson model to predict the incidence of malaria based on epidemiologic, environmental, and vector control variables. 5. Understand how to use the proposed model as an aid for planning control measures for malaria.
Keywords: Research, Epidemiology
Presenting author's disclosure statement:
Organization/institution whose products or services will be discussed: None
I do not have any significant financial interest/arrangement or affiliation with any organization/institution whose products or services are being discussed in this session.