State-wise forecasting of cancer incidence in India using a moderated exponential regression model incorporating healthcare system determinants

Dileep Kumar; Navin Upadhyay; Himanshu Pandey

doi:10.18203/2394-6040.ijcmph20261025

Authors

Dileep Kumar Department of Obstetrics and Gynecology BRD Medical College Gorakhpur, Uttar Pradesh, India
Navin Upadhyay Department of Mathematics and Statistics, Deen Dayal Upadhyaya Gorakhpur University, Gorakhpur, Uttar Pradesh, India
Himanshu Pandey Department of Mathematics and Statistics, Deen Dayal Upadhyaya Gorakhpur University, Gorakhpur, Uttar Pradesh, India

DOI:

https://doi.org/10.18203/2394-6040.ijcmph20261025

Keywords:

Cancer, Moderated exponential, Regression

Abstract

Background: Cancer incidence in India varies substantially across states and union territories, posing a significant and growing public health challenge. Accurate projections of future cancer burden are essential for effective health planning, resource allocation, and strengthening health systems. However, conventional forecasting approaches largely rely on historical incidence trends and often ignore regional differences in health-care capacity, which influence disease detection and reporting. Variations in infrastructure, oncology services, screening coverage, and health expenditure can significantly affect observed cancer incidence. This study proposed a moderated forecasting framework that integrates key state-level health-care indicators to generate more realistic and policy-relevant projections.

Methods: A moderated exponential regression model was developed using state-wise cancer incidence data from the National Cancer Registry Programme (ICMR-NCRP) for 2019-2022. The model incorporated multiple health system moderators, including cancer hospitals, oncologist availability, screening coverage, CHC/PHC density, health infrastructure index, per-capita health expenditure, and hospital-bed availability. Three projection scenarios were constructed: short-term (2026) using health-care facility density, medium-term (2030) incorporating infrastructure and workforce variables, and long-term (2045) including all seven moderators.

Results: The moderated models revealed substantial inter-state variation in projected cancer incidence. States with stronger health systems exhibited slower increases, whereas those with weaker infrastructure showed more rapid growth. Model fit and predictive accuracy improved significantly compared to conventional exponential models.

Conclusions: Incorporating health-care system factors enhances the reliability of cancer incidence projections. The proposed framework offers a robust, evidence-based tool for policy planning and targeted health system strengthening in India.

Metrics

Metrics Loading ...

References

Carmer SG. Computational aspects of exponential regression. Commun Stat Theory Methods. 1974;3(3):229-45.

Breslow NE, Day NE, Heseltine E. Statistical methods in cancer research: Volume I- the analysis of case-control studies. IARC Scientific Publications No. 32. International Agency for Research on Cancer. 1980.

Lee ET, Wang J. (2003). Statistical methods for survival data analysis. 3rd edn. Wiley; 2003. DOI: https://doi.org/10.1002/0471458546

Abedi H, Sadeghi M, Boroumand MA, Nezami N. Application of survival models in oncology research: a comparative study. Cancer Epidemiol. 2019;60:80-7.

Chhatwal J, Alagoz O, Burnside ES. Logistic regression and decision-analytic modelling of breast cancer diagnosis. Med Decision Making. 2009;29(4):414-26.

Falgreen S, Dybkær K, Young KH, Xu-Monette ZY, El-Galaly TC, Laursen MB, et al. Predicting response to multidrug regimens in cancer patients using cell line experiments and regularised regression models. BMC Cancer. 2015 Apr 8;15:235.. DOI: https://doi.org/10.1186/s12885-015-1237-6

Prataviera F, Rotnitzky A, Robins J. Exponentiated regression models for epidemiological applications. Biometr J. 2019;61(2):302-20.

Ramires TG, Ortega EMM, Cordeiro GM. Power exponential regression models: Properties and applications. Commun Stat Theory Methods. 2018;47(6):1344-62.

Tanaka H, Ishibuchi H, Yoshikawa M. Exponential possibility regression and its applications. Fuzzy Sets Syst. 1995;69(1):51-60. DOI: https://doi.org/10.1016/0165-0114(94)00179-B

Suleiman M, Ahmed Z, Rahman A. Simulation-assisted nonlinear models for cancer data forecasting. Stat Modell. 2025;25(1):90-112.

Abaoud AA, Almuqrin A. Heavy‐tailed regression models for biomedical data: simulation‐driven evaluation and applications. J Appl Stat. 2024;51(3):512-30.

Almuqrin A. Simulation-based nonlinear regression methods for medical prediction models. Commun Stat Simulat Comput. 2024;53(1):145-62.