Identifying Rural Hotspots for Head and Neck Cancer Using the Bayesian Mapping Approach

TL;DR

This study uses Bayesian mapping to identify rural areas in Queensland with higher rates of head and neck cancer, aiming to improve resource allocation for prevention and treatment.

Contribution

The study is the first to apply Bayesian mapping to identify head and neck cancer hotspots in rural and remote Australian communities.

Findings

22 rural and remote local government areas in Queensland showed significantly higher head and neck cancer risks.

Four LGAs had the highest mortality rates for head and neck cancer.

A rising trend in head and neck cancer incidence was observed from 1982 to 2018.

Abstract

Cancers often tend to occur at a higher rate in rural and remote communities due to various reasons. Identifying these cancer hotspots will assist in adequate resourcing of such hotspots. This study was conducted to identify head and neck cancer hotspots in Queensland (QLD), Australia, based on the historical data collected by the cancer register and employing a specialized Bayesian mapping approach. The findings of this study suggested that many rural and remote regions in QLD experience significantly higher head and neck cancer incidence rates and death rates when compared to the QLD state average rates and their surrounding regions. Additionally, a generalized increasing trend of head and neck cancers was noted across the studied period (1982–2018). Although the precise reasons for this increasing trend over time are unclear, a range of factors, such as distance from the tertiary hospitals, lack of awareness of risk factors, behavioral and lifestyle factors, and delayed diagnosis, may be considered to contribute. Our study successfully utilized a robust method to identify head and neck cancer hotspots and will aid in supporting the community and healthcare providers in the region with additional resources to prevent and manage cancer.

Background: The Bayesian mapping approach has not been used to identify head and neck cancer hotspots in Australia previously. This study aims to identify rural communities at risk of head and neck cancer (HNC) for targeted prevention programs. Methods: This study included data from 23,853 cases recorded in the Queensland Cancer Register between 1982 and 2018. Outcomes for mapping included incidence, overall mortality, 3-year mortality, and 5-year mortality. Local government areas (LGAs) with a general population aged 15 years and above (according to 2016 census data from the Australian Bureau of Statistics) were utilized for mapping. Results: Of the 59 LGAs with higher-than-average risk, 22 predominantly rural and remote LGAs showed statistically significant higher risks of head and neck cancer occurrence. Estimated median standardized mortality ratios (SMRs) ranged from 0.57 to 3.44 and were higher than the state average in 38 LGAs. Four LGAs had the highest mortality—the Shires of Quilpie, Yarrabah, Murweh, and Hinchinbrook. Conclusions: Whilst reasons for some LGAs exhibiting higher HNC are unknown, Bayesian mapping highlights these rural and remote regions as worthy of further investigation. In conclusion, the Bayesian disease mapping approach is effective in identifying high-risk communities for HNC. Findings from this study will aid in designing targeted screening and interventions for the prevention and management of head and neck cancer in regional and remote communities through support services such as a cancer navigator.

Figures