Poisson-Gamma and Spatial-Temporal Models: with Application to Cervical Cancer in Kenya’s Counties

Abstract

In Africa, Cancer is an emerging health problem where in 2012 new cancer cases were about 847,00 and around 519,00 deaths, three quarters of those deaths occurred in sub-Saharan region. In 2018, cancer was ranked as the third leading cause of deaths in Kenya after infectious and cardiovascular diseases. In 2018 cancer incidences were estimated to be 47,887 new cancer cases and 32,987 deaths. According to data from World Health Organization in 2020, cervical cancer is the second most prevalent cancer among women while breast cancer is the first. In this study, data collected by the Nairobi Cancer Registry (NCR) was used to produce spatial-temporal distribution of the cervical cancer in counties in Kenya. The results showed that counties where data was available among them Embu, Meru, Machakos, Mombasa, Nyeri, Kiambu, Kakamega, Nairobi and Bomet respectively had high risk of cervical cancer. Availability of county-based estimates and spatial-temporal distribution of cervical cancer cases will aide development of targeted county strategies, enhance early detection, promote awareness and implementation of universal coverage of major control interventions which will be crucial in reducing and halting the rising burden of the cancer cases in Kenya. In counties where data was not available the model showed relative risks for cervical cancer disease was minute but it was present, therefore spatial temporal models are very appropriate to estimate relative risks of diseases even when there is a small sample (and possibly without a sample) in a given area by borrowing information from other neighboring regions.