The development of stochastic models for relative risk estimation in constructing pneumonia disease mapping in Malaysia
Pneumonia is one of the leading causes of death for infectious diseases especially in developing countries. Conventionally, its spread is only being monitored based on the total number of cases recorded without considering geographical distribution. Alternatively, disease mapping can be constructed...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | eng eng |
| Published: |
2022
|
| Subjects: | |
| Online Access: | https://etd.uum.edu.my/9775/1/permission%20to%20deposit-grant%20the%20permission-903660.pdf https://etd.uum.edu.my/9775/2/s903660_01.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Pneumonia is one of the leading causes of death for infectious diseases especially in developing countries. Conventionally, its spread is only being monitored based on the total number of cases recorded without considering geographical distribution. Alternatively, disease mapping can be constructed based on the relative risk that includes a geographical distribution. A good disease mapping relies on the accuracy of relative risk estimated from the best-fitted statistical model. Therefore, this study aims to develop an alternative method in estimating the pneumonia relative risk based on four stochastic models: Susceptible-Infected-Carriers (SIC), Susceptible-Infected Recovered (SIR), Susceptible-Carrier-Infected-Recovered (SCIR), and Susceptible- Vaccinated-Carrier-Infected-Recovered (SVCIR). These estimated relative risks are then compared with those of the existing methods: Standardized Mortality Ratio (SMR), Poisson-gamma and Besag, York and Mollie (BYM) models. There are four phases in this study. Firstly, four deterministic models that are suitable for pneumonia disease transmission are selected, from which the stochastic models are developed. Next, these four stochastic models are applied to estimate the relative risk for pneumonia disease by analyzing pneumonia data in Malaysia from the year 2010 until the year 2019. The performance of these four stochastic models and existing methods is evaluated by comparing their relative risk values. Finally, the pneumonia risk maps are then constructed based on the relative risk values obtained. Findings show that there is a large gap in relative risk estimation values when using the stochastic SVCIR model compared to other models. The relative risk values when using stochastic SVCIR model decrease from high-risk level to medium risk level and from medium risk level to low-risk level. This situation occurs since stochastic SVCIR model allows for the spatial correlation between the areas and includes extra information in the model such as vaccination and carrier components. Application of the models on the Malaysian data shows that Putrajaya is identified as the highest risk of contracting pneumonia. This is because Putrajaya is the smallest area with the highest population growth rate in Malaysia. In conclusion, these stochastic models demonstrate better performance compared to the conventional models. Furthermore, these models are applicable to other infectious diseases with similar transmission characteristics. The disease mapping may assist the government in prioritizing areas that need further attention in gearing towards a sustainable health system. |
|---|