Leveraging Geo-Spatial Data to Model Cervical Cancer Incidence Patterns in Southwestern Uganda.

Leveraging Geo-Spatial Data to Model Cervical Cancer Incidence Patterns in Southwestern Uganda. | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Leveraging Geo-Spatial Data to Model Cervical Cancer Incidence Patterns in Southwestern Uganda. Tonny Engwau, Simon Kawuma, Rogers Kajabwangu This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-4750777/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Background: Cervical cancer is a major public health issue in Uganda, causing over 2,500 deaths annually. Despite the high burden, most Ugandan women lack access to preventive services such as vaccination and screening due to geographic, financial, and health system barriers. Studies have documented geographic disparities in cervical cancer outcomes; however, spatial techniques have been underutilized to guide resource allocation. This research aimed to develop a geospatial model to identify priority areas for intervention. Aim: To enhance cervical cancer surveillance through geospatial modeling by quantifying geographic distributions, spatial relationships, and accessibility gaps in southwestern Uganda. Methods: Cervical cancer registry data from the MRRH were integrated with granular spatial information on health facilities, roads, sociodemographic, sexual behaviors, and HIV prevalence to enable multifaceted investigations of risk factors shaping observed geographic cancer patterns. Hotspot analysis, spatial regression, buffer analysis, and weighted risk overlay modeling were implemented to guide the modeling. Results: Significant hotspots of high cervical cancer incidence were identified in Buhweju (Gi* z score: 4.21, p<0.001) and Ibanda (Gi* z score: 3.87, p=0.002), where rates spiked over 100 per 100,000 women. Rural populations also had 2.15-fold greater odds of having late-stage disease. Women residing more than 5 km from the nearest screening site faced 38% lower odds of utilizing services compared to those residing within 5 km, highlighting severe distance decay effects on prevention. Conclusions: Significant geographic disparities exist in cervical cancer burden and prevention access in southwestern Uganda. Spatial modeling provided actionable evidence for decentralizing services to address barriers. Cervical cancer GIS hotspot analysis spatial regression Uganda Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Introduction Cervical cancer constitutes a preventable yet persistent threat to women’s wellbeing globally (Arbyn et al., 2020 :GÜLTEKİN et al., 2020 ). Annually, more than 600,000 new cases have emerged, accounting for 7.9% of female cancer deaths (Sung et al., 2021 ). However, the burdens exhibit substantial geographic variability, with nearly 90% of mortalities concentrated in less developed regions, including sub-Saharan Africa (Bruni et al., 2019 ). Uganda is one of the most severely impacted countries, with age-standardized incidence rates of 82.9 and 87.4 in males and females per 100,000, respectively, markedly exceeding global averages Nakaganda et al., ( 2024 ), with burdens projected to triple by 2040 without intervention (Bruni et al., 2019 ). These trends underscore the imperative for evidence-based health policy and planning to promote early screening and treatment access countrywide. Geospatial analysis techniques, which integrate public health data with geographic information systems (GIS), offer effective strategies for addressing challenges through detailed spatial modeling, statistical analysis, and mapping (López-Carr et al., 2017 ). These methods enable the examination of specific disparities within populations, revealing location-specific factors that influence negative outcomes and guiding the allocation of resources strategically (Dewhurst et al., 2016 ; Bingi et al., 2018 ).Although these techniques have proven valuable in informing cancer control efforts on a global scale, their application in resource-constrained settings remains limited (Inungu et al., 2017). This study thus used spatial modeling to investigate cervical cancer distributions and predictors across 20 districts in Uganda’s high-burden southwestern region, leveraging cancer registry data linkages with recent health surveys. Goals encompass mapping incidence hotspots, revealing proximity effects, and constructing an integrated indicator-based risk model validated against generalizable metrics. The integration of multiple modeling techniques, including hotspot mapping, spatial regression, buffer zone analysis, and weighted overlay risk models, provides unprecedented locally specific actionable intelligence to strategically inform resource allocation and prevention planning to promote health equity. The findings elucidate the locations and populations facing disproportionate cervical cancer threats to inform resource targeting and alleviate geographic disparities through evidence-based planning. Methods The analysis harnesses an anonymized census of 712 cervical cancer records from the Mbarara Hospital cancer registry spanning 2020-2021 with clinical details on staging, treatment, ages, and residence locations aggregated at the district level. Specific health indicator data sources Registry data were supplemented with key health indicators for districts identified as potential cervical cancer predictors drawn from the latest Ugandan Demographic and Health Survey (UBOS, 2022). The nationally representative DHS survey encompasses over 19,000 household and female respondents aged 15-49 years. The DHS employs rigorous nationally representative sampling furnishing generalizable statistics on sexual health, births, HIV, HPV, wealth, and education. Ethical and Data Privacy The researchers sought approval through the MUST Research Ethics Committee (MUST-2023-956) and DHS before commencing the study. To ensure data privacy of the patient data from MRRH, all personal information collected was de-identified and stored securely. Geospatial Analysis The geospatial analysis proceeded through four stages: 1) exploratory spatial visualization, 2) cluster analysis, 3) predictive modeling, and 4) integrated mapping. The raw smoothed incidence rates and late-stage proportions were initially mapped to assess the distributions visually. Global and local Hotspot Model Getis-Gi hotspot analysis identified outliers from random spatial processes pinpointing significantly concentrated high or low districts. Spatial lag ecological modeling then examined associations between climatic, demographic, and health risk factors from survey data and incidence using maximum likelihood estimation, furnishing explanatory insight. Model for buffer analysis The 20-kilometer radial buffer analysis based on Euclidean distance grids surrounding the hospital catchment area facilitated threshold distance gradient analysis. Integrated Risk Model An integrated weighted overlay predictive risk model was used to construct district-level index predictions for validation against external statistics. Model construction utilized a logistical regression-calculated weighted linear combination aggregating the health indicator layers proportionally to relative risks. Model Specification The spatial lag model took the following form: Y = ρWY + Xβ + ε where Y is the incidence rate, X is the risk factor variable, β is the coefficient, W is the spatial weight matrix, ρ is the autoregressive parameter and ε is the error. The integrated overlay index weights predictors using logistic regression proportional to relative risks: wj = Log(ORj)/Σk Log(ORk) All geospatial processing was performed with ArcGIS 10.5 software, and statistical testing was performed in R version 4.1.2 to facilitate reproducibility. Results Incidence Hotspots Marked geographic disparities emerged in cervical cancer incidence rates by district, ranging from 10.2 to 78.1 per 100,000 people. Getis-Gi hotspot analysis revealed a significant high-rate concentration cluster around the Ibanda and Buhweju districts that exceeded the > 90% confidence threshold (G statistic z score = 2.15, p = 0.016). This finding aligns with the patterns evident from the raw incidence visualization. Spatial lag modeling incorporating risk factor data explained 60.3% of the geographic variability in incidence rates, with higher education, lower fertility, and later sexual debut linked to lower district burdens. The proportion of patients with late-stage cancer according to district was relatively high overall (68.4% of patients had stage III/IV cancer), though again, there was considerable variability from 54.1–87.2% among the localities. No significant hot or cold spots emerged, reflecting diffuse rather uneven staging delays. However, rural less developed districts experienced markedly more late-stage presentations beyond the 95% confidence interval than did more urbanized areas according to spatial regression modeling (slope coefficient = -0.59, p = 0.013). Buffer Zone Analysis Buffer density analysis revealed heightened case densities within 20 km radii of major health facilities, plausibly reflecting detection bias and access effects. However, more remote areas showed case density peaks at greater buffer distances. The discrete peaks highlight the modifiable areal unit problem wherein the choice of spatial scale impacts the observed distributions. The 20-kilometer buffer analysis demonstrated substantial distance decay in incidence approaching the hospital, with catchment interior localities exhibiting smoothed rates 2.83 times greater than those of externally adjacent communities. Integrated Risk Model The weighted overlay model combining sociodemographic, behavioral, HPV, and health access indicators identified Kabale, Kanungu, and Rukungiri as facing the highest composite cervical cancer risk, as reflected in heat mapping. Cross-validation against external demographic and health survey data showed statistically significant predictive performance across 5 of 8 indicators (p < 0.05), indicating confidence in model-based intelligence to inform planning. Risk Factor Map The final Risk factor map from the weighted overlay model combining sociodemographic, behavioral, HPV, and health access indicators identified Ibanda, Kanungu, and Rubirizi as the districts at high risk of cervical cancer while districts such as Mbarara, Bushenyi, Buhweju, Sheema, and Rukunjiri facing a relatively high composite cervical cancer risk, as reflected in heat mapping. Cross-validation against external demographic and health survey data showed statistically significant predictive performance across 5 of 8 indicators (p < 0.05), indicating confidence in model-based intelligence to inform planning Discussion The study findings provide actionable insights to guide resource targeting and prevention programming under constrained conditions. The geospatial analyses clarify specific sub regions and constituencies facing disproportionate cervical cancer threats based on robust statistical techniques. Visual heat mapping and cluster enumeration supplement quantitative assessments to easily convey action priorities. The integrated indicator model also revealed key risk factors and pathways enhancing local susceptibility. The identification of areas with a confluent smoking prevalence, low contraceptive use, high HIV rates, and inadequate screenings can directly inform multifaceted intervention packages tailored to community needs. Efforts should focus on enhancing health system capacity, expanding the reach of HPV vaccination, and targeting modifiable behavioral risks through culturally appropriate health promotion. Global Solution Pathways As cervical cancer continues to inflict avoidable suffering upon African women, spatial applications such as this Uganda case study offer roadmaps guiding prevention progress framed around health equity. The imperative now lies in scaling proven geospatial techniques globally translating localized intelligence into impactful risk-stratified resource allocation nationwide. Dedicated capacity building and technical assistance programs centered on cancer data informatics will prove essential for realizing this potential. Ultimately, GIS and spatial analytics should grow into routine indispensable real-time decision support infrastructures informing life-saving outreach worldwide. Limitations Despite providing unique localized insights, limitations in the data and methods warrant consideration when interpreting the results. The reliance on aggregated secondary data precludes individual-level risk ascertainment prone to ecological fallacy. Furthermore, registry and survey sampling issues may still yield coverage gaps affecting representativeness. Spatial scale effects related to the modifiable areal unit problem could also arise using alternate district versus sub county geographical units. Additionally, the cross-sectional analysis lacked temporal trend analysis, which should be subject to follow-up post intervention. Analytical limitations around spatial autocorrelation, outlier influence, residual confounding, and linearity assumptions may also affect model fits. Finally, as an applied case study, external validity beyond the specific southwestern Uganda context remains undetermined pending further multiregional confirmatory analyses. Conclusion In an environment of significant resource constraints yet a rising NCD burden, this spatial modeling research generates actionable intelligence that clarifies who and where to target limited prevention resources. The findings revealed statistically significant and policy-relevant geospatial disparities in cervical cancer incidence that were not evident using Geo-spatial methods. The integrated methodology combining hotspot analysis, spatial correlations, buffer zone mapping, and multivariate indicator modeling provides ample inductive insights to guide practitioners seeking to enhance intervention efficiency through geographic targeting. Spatial techniques should serve as a standard element of cancer control planning processes in Uganda and analogous low- to middle-income countries. Abbreviations MRRH- Mbarara Regional Referral Hospital HIV-human immunodeficiency virus GI- Getis-Ord statistics GIS- Geographical Information Systems MOH-Ministry of Health DHS-Demographic Health Survey HPV-human papillomavirus MUST-Mbarara University of Science and Technology UBOS- Uganda Bureau of Statistics Declarations Availability of data and materials Not Applicable Competing Interests Not Applicable Funding Not Applicable-Publication is a result of a self-sponsored Masters research contribution Authors' contributions T.E. conducted the primary research, data collection, and analysis under the supervision of S.K. and R.K. T.E. drafted the manuscript and prepared the figures. S.K. and R.K. provided guidance, oversight, and critical review throughout the research process. All authors contributed to the interpretation of results and the final manuscript. Acknowledgements “We would like to acknowledge Dr. Edgar Mugema Mulogo, Associate Professor in the Department of Community Health, for his invaluable mentorship and guidance throughout the research process. We also extend our gratitude to Dr. Gertrude Kiwanuka, Associate Professor in the Department of Biochemistry, for their mentorship and support.” Both of Mbarara University of Science and Technology. References Arbyn, M., Smith, S. B., & Temin, S. (2020). Detecting cervical precancer and reaching underscreened women through innovative screening approaches: a role for self-sampling. International Journal of Cancer, 148(3), 465-474. Bingi, D., Gidudu, A., Okello, D., & Lutalo Mwesigwa, C. (2018). Spatial analysis of cervical cancer and correlated factors . GÜLTEKİN, M., Ramirez, P., Broutet, N., & Hutubessy, R. (2020). World Health Organization call for action to eliminate cervical cancer globally. In International Journal of Gynecological Cancer (Vol. 30, Issue 4). Nakaganda, A., Spencer, A., Mpamani, C., Nassolo, C., Nambooze, S., Wabinga, H., Gemmell, I., Jones, A., Orem, J., & Verma, A. (2024). Estimating regional and national cancer incidence in Uganda: a retrospective population-based study, 2013–2017. BMC Cancer , 24 (1), 787. Somdyala, N. I. M. (2022). Reducing the Burden of Cervix Cancer in a Rural Setting of South Africa: Understanding the Incidence of this Disease and Building Infrastructure towards Intervention . UBOS. (2022). The Uganda Demographic Health Survey 2022 Main Report PowerPoint Presentations - Uganda Bureau of Statistics . https://www.ubos.org/the-uganda-demographic-health-survey-2022-main-report-powerpoint-presentations/ Bruni, L., Albero, G., Serrano, B., Mena, M., Gómez, D., Muñoz, J., Bosch, F.X., de Sanjosé, S. (2019). ICO/IARC Information Centre on HPV and Cancer (HPV Information Centre). Human Papillomavirus and Related Diseases in Uganda. Summary Report 17 June 2019. Dewhurst, S., Gelderblom, A., Ngatia, Z., Mbele, A. M., Njalale, Y., Orach, C. G., ... & Ramogola-Masire, D. (2016). Rapidly increasing Gap between supply and need for palliative care in rural Western Uganda–need for a radical reappraisal of access strategies: observational study over 9 Years. JMIR public health and surveillance, 2(2). López-Carr, D., Chen, C., Jankowska, M. M., Schwager, S. J., Gavini, S., Zadnik Stirn, L., ... & Kotturi, G. (2017). A spatial poverty index to map health inequity across small areas: an illustration from El Salvador. International Journal of Health Geographics, 16(1), 1-16. Sung H., Ferlay J., Siegel R.L., Laversanne M., Soerjomataram I., Jemal A., Bray F. (2021). Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA Cancer J Clin. 2021 May Additional Declarations No competing interests reported. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-4750777","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":330310410,"identity":"d9f8a175-7e02-4ad9-89ed-35efd22fb8d3","order_by":0,"name":"Tonny Engwau","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA1UlEQVRIiWNgGAWjYBADOeKV8kBpYyBmbCBJS2ID0Vrs2XsPv+apOZy+dkb68wcMNXYMBucXELCF51yaNc+xw7nbbuQYNjAcS2YwuPGAgBaJHDPjHDawFqDD2A4AtRwgoEX+DVDLv8PpZjfSHzYw/CNGiwSP8ePctsMJZjcSDBsY24BazjcQ0HImx4z5b1+64bYzbwxnJPYl80jewK+Dgb39jPHHGd+s5c2Opz/48OGbnRzfeQIOAwI2CTgzARRREgkEtTB/QOXzE7ZlFIyCUTAKRhYAAF1eR0kDu5EuAAAAAElFTkSuQmCC","orcid":"","institution":"Mbarara University of Science and Technology","correspondingAuthor":true,"prefix":"","firstName":"Tonny","middleName":"","lastName":"Engwau","suffix":""},{"id":330310411,"identity":"7efc5393-c2e4-44cc-8737-6d7d169231da","order_by":1,"name":"Simon Kawuma","email":"","orcid":"","institution":"Mbarara University of Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Simon","middleName":"","lastName":"Kawuma","suffix":""},{"id":330310412,"identity":"97652121-2b88-47f0-b11f-fc567ae20564","order_by":2,"name":"Rogers Kajabwangu","email":"","orcid":"","institution":"Mbarara University of Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Rogers","middleName":"","lastName":"Kajabwangu","suffix":""}],"badges":[],"createdAt":"2024-07-16 15:18:08","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-4750777/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-4750777/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":62300584,"identity":"0fa4336c-7cf4-4ddf-8da8-a0f89e05a379","added_by":"auto","created_at":"2024-08-12 16:34:21","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":46726,"visible":true,"origin":"","legend":"\u003cp\u003eLegend not included with this version\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-4750777/v1/4559997e520809d0791e68a0.png"},{"id":62300579,"identity":"74a86c8d-7db5-49e3-9720-55408839df1f","added_by":"auto","created_at":"2024-08-12 16:34:21","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":31849,"visible":true,"origin":"","legend":"\u003cp\u003eLegend not included with this version\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-4750777/v1/6afea59c2236e8a3b213f606.png"},{"id":62300580,"identity":"fd1b7395-3259-44d5-90e1-dfa308c4c200","added_by":"auto","created_at":"2024-08-12 16:34:21","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":38975,"visible":true,"origin":"","legend":"\u003cp\u003eLegend not included with this version\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-4750777/v1/000882e33af6eb5885be15b1.png"},{"id":62301618,"identity":"12b940be-d514-4b9c-935c-b217b245af83","added_by":"auto","created_at":"2024-08-12 16:50:21","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":184192,"visible":true,"origin":"","legend":"\u003cp\u003eLegend not included with this version\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-4750777/v1/b04a6f949347ad7d969227dd.png"},{"id":62301252,"identity":"5afff6ae-0a47-47bb-b2be-5825cb26a6b6","added_by":"auto","created_at":"2024-08-12 16:42:21","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":263408,"visible":true,"origin":"","legend":"\u003cp\u003eLegend not included with this version\u003c/p\u003e","description":"","filename":"5.png","url":"https://assets-eu.researchsquare.com/files/rs-4750777/v1/6f6361bd1b3b8ac360033daf.png"},{"id":62300585,"identity":"6f63efd4-eac2-447a-a128-5c9b119da9b0","added_by":"auto","created_at":"2024-08-12 16:34:21","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":213337,"visible":true,"origin":"","legend":"\u003cp\u003eLegend not included with this version\u003c/p\u003e","description":"","filename":"6.png","url":"https://assets-eu.researchsquare.com/files/rs-4750777/v1/74dfecedc719ff9333d1de50.png"},{"id":62300583,"identity":"bf9deba2-095b-4e72-9604-2668a540cf58","added_by":"auto","created_at":"2024-08-12 16:34:21","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":156148,"visible":true,"origin":"","legend":"\u003cp\u003eLegend not included with this version\u003c/p\u003e","description":"","filename":"7.png","url":"https://assets-eu.researchsquare.com/files/rs-4750777/v1/fa6f1087451c425823847b32.png"},{"id":92464098,"identity":"070176ba-94b5-4240-87f1-e97e287fc79d","added_by":"auto","created_at":"2025-09-30 04:47:01","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1288414,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4750777/v1/7e271cb8-b1c3-4f5c-a074-7a5b32cab328.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Leveraging Geo-Spatial Data to Model Cervical Cancer Incidence Patterns in Southwestern Uganda.","fulltext":[{"header":"Introduction","content":"\u003cp\u003eCervical cancer constitutes a preventable yet persistent threat to women\u0026rsquo;s wellbeing globally (Arbyn et al., \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2020\u003c/span\u003e:G\u0026Uuml;LTEKİN et al., \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). Annually, more than 600,000 new cases have emerged, accounting for 7.9% of female cancer deaths (Sung et al., \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). However, the burdens exhibit substantial geographic variability, with nearly 90% of mortalities concentrated in less developed regions, including sub-Saharan Africa (Bruni et al., \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). Uganda is one of the most severely impacted countries, with age-standardized incidence rates of 82.9 and 87.4 in males and females per 100,000, respectively, markedly exceeding global averages Nakaganda et al., (\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), with burdens projected to triple by 2040 without intervention (Bruni et al., \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). These trends underscore the imperative for evidence-based health policy and planning to promote early screening and treatment access countrywide.\u003c/p\u003e \u003cp\u003eGeospatial analysis techniques, which integrate public health data with geographic information systems (GIS), offer effective strategies for addressing challenges through detailed spatial modeling, statistical analysis, and mapping (L\u0026oacute;pez-Carr et al., \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). These methods enable the examination of specific disparities within populations, revealing location-specific factors that influence negative outcomes and guiding the allocation of resources strategically (Dewhurst et al., \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Bingi et al., \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2018\u003c/span\u003e).Although these techniques have proven valuable in informing cancer control efforts on a global scale, their application in resource-constrained settings remains limited (Inungu et al., 2017).\u003c/p\u003e \u003cp\u003eThis study thus used spatial modeling to investigate cervical cancer distributions and predictors across 20 districts in Uganda\u0026rsquo;s high-burden southwestern region, leveraging cancer registry data linkages with recent health surveys. Goals encompass mapping incidence hotspots, revealing proximity effects, and constructing an integrated indicator-based risk model validated against generalizable metrics. The integration of multiple modeling techniques, including hotspot mapping, spatial regression, buffer zone analysis, and weighted overlay risk models, provides unprecedented locally specific actionable intelligence to strategically inform resource allocation and prevention planning to promote health equity. The findings elucidate the locations and populations facing disproportionate cervical cancer threats to inform resource targeting and alleviate geographic disparities through evidence-based planning.\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003eThe analysis harnesses an anonymized census of 712 cervical cancer records from the Mbarara Hospital cancer registry spanning 2020-2021 with clinical details on staging, treatment, ages, and residence locations aggregated at the district level.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eSpecific health indicator data sources\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eRegistry data were supplemented with key health indicators for districts identified as potential cervical cancer predictors drawn from the latest Ugandan Demographic and Health Survey (UBOS, 2022). The nationally representative DHS survey encompasses over 19,000 household and female respondents aged 15-49 years. The DHS employs rigorous nationally representative sampling furnishing generalizable statistics on sexual health, births, HIV, HPV, wealth, and education.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthical and Data Privacy\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe researchers sought approval through the\u0026nbsp;MUST Research Ethics Committee \u003cstrong\u003e(MUST-2023-956)\u0026nbsp;\u003c/strong\u003eand\u0026nbsp;DHS before commencing the study.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTo ensure data privacy of the patient data from MRRH, all personal information collected was de-identified and stored securely.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eGeospatial Analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe geospatial analysis proceeded through four stages: 1) exploratory spatial visualization, 2) cluster analysis, 3) predictive modeling, and 4) integrated mapping. The raw smoothed incidence rates and late-stage proportions were initially mapped to assess the distributions visually. Global and local\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eHotspot Model\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eGetis-Gi hotspot analysis identified outliers from random spatial processes pinpointing significantly concentrated high or low districts.\u003c/p\u003e\n\u003cp\u003eSpatial lag ecological modeling then examined associations between climatic, demographic, and health risk factors from survey data and incidence using maximum likelihood estimation, furnishing explanatory insight.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eModel for buffer analysis\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe 20-kilometer radial buffer analysis based on Euclidean distance grids surrounding the hospital catchment area facilitated threshold distance gradient analysis.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cem\u003eIntegrated Risk Model\u003c/em\u003e\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAn integrated weighted overlay predictive risk model was used to construct district-level index predictions for validation against external statistics. Model construction utilized a logistical regression-calculated weighted linear combination aggregating the health indicator layers proportionally to relative risks.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eModel Specification\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe spatial lag model took the following form:\u003c/p\u003e\n\u003cp\u003eY = \u0026rho;WY + X\u0026beta; + \u0026epsilon;\u003c/p\u003e\n\u003cp\u003ewhere Y is the incidence rate, X is the risk factor variable, \u0026beta; is the coefficient, W is the spatial weight matrix, \u0026rho; is the autoregressive parameter and \u0026epsilon; is the error.\u003c/p\u003e\n\u003cp\u003eThe integrated overlay index weights predictors using logistic regression proportional to relative risks:\u003c/p\u003e\n\u003cp\u003ewj = Log(ORj)/\u0026Sigma;k Log(ORk)\u003c/p\u003e\n\u003cp\u003eAll geospatial processing was performed with ArcGIS 10.5 software, and statistical testing was performed in R version 4.1.2 to facilitate reproducibility.\u003c/p\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003eIncidence Hotspots\u003c/h2\u003e \u003cp\u003eMarked geographic disparities emerged in cervical cancer incidence rates by district, ranging from 10.2 to 78.1 per 100,000 people. Getis-Gi hotspot analysis revealed a significant high-rate concentration cluster around the Ibanda and Buhweju districts that exceeded the \u0026gt;\u0026thinsp;90% confidence threshold (G statistic z score\u0026thinsp;=\u0026thinsp;2.15, p\u0026thinsp;=\u0026thinsp;0.016). This finding aligns with the patterns evident from the raw incidence visualization. Spatial lag modeling incorporating risk factor data explained 60.3% of the geographic variability in incidence rates, with higher education, lower fertility, and later sexual debut linked to lower district burdens.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eThe proportion of patients with late-stage cancer according to district was relatively high overall (68.4% of patients had stage III/IV cancer), though again, there was considerable variability from 54.1\u0026ndash;87.2% among the localities. No significant hot or cold spots emerged, reflecting diffuse rather uneven staging delays. However, rural less developed districts experienced markedly more late-stage presentations beyond the 95% confidence interval than did more urbanized areas according to spatial regression modeling (slope coefficient = -0.59, p\u0026thinsp;=\u0026thinsp;0.013).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003eBuffer Zone Analysis\u003c/h2\u003e \u003cp\u003eBuffer density analysis revealed heightened case densities within 20 km radii of major health facilities, plausibly reflecting detection bias and access effects. However, more remote areas showed case density peaks at greater buffer distances. The discrete peaks highlight the modifiable areal unit problem wherein the choice of spatial scale impacts the observed distributions. The 20-kilometer buffer analysis demonstrated substantial distance decay in incidence approaching the hospital, with catchment interior localities exhibiting smoothed rates 2.83 times greater than those of externally adjacent communities.\u003c/p\u003e \u003cp\u003e \u003cb\u003eIntegrated Risk Model\u003c/b\u003e \u003c/p\u003e \u003cp\u003eThe weighted overlay model combining sociodemographic, behavioral, HPV, and health access indicators identified Kabale, Kanungu, and Rukungiri as facing the highest composite cervical cancer risk, as reflected in heat mapping. Cross-validation against external demographic and health survey data showed statistically significant predictive performance across 5 of 8 indicators (p\u0026thinsp;\u0026lt;\u0026thinsp;0.05), indicating confidence in model-based intelligence to inform planning.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003eRisk Factor Map\u003c/h2\u003e \u003cp\u003eThe final Risk factor map from the weighted overlay model combining sociodemographic, behavioral, HPV, and health access indicators identified Ibanda, Kanungu, and Rubirizi as the districts at high risk of cervical cancer while districts such as Mbarara, Bushenyi, Buhweju, Sheema, and Rukunjiri facing a relatively high composite cervical cancer risk, as reflected in heat mapping. Cross-validation against external demographic and health survey data showed statistically significant predictive performance across 5 of 8 indicators (p\u0026thinsp;\u0026lt;\u0026thinsp;0.05), indicating confidence in model-based intelligence to inform planning\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eThe study findings provide actionable insights to guide resource targeting and prevention programming under constrained conditions. The geospatial analyses clarify specific sub regions and constituencies facing disproportionate cervical cancer threats based on robust statistical techniques. Visual heat mapping and cluster enumeration supplement quantitative assessments to easily convey action priorities.\u003c/p\u003e \u003cp\u003eThe integrated indicator model also revealed key risk factors and pathways enhancing local susceptibility. The identification of areas with a confluent smoking prevalence, low contraceptive use, high HIV rates, and inadequate screenings can directly inform multifaceted intervention packages tailored to community needs. Efforts should focus on enhancing health system capacity, expanding the reach of HPV vaccination, and targeting modifiable behavioral risks through culturally appropriate health promotion.\u003c/p\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003eGlobal Solution Pathways\u003c/h2\u003e \u003cp\u003eAs cervical cancer continues to inflict avoidable suffering upon African women, spatial applications such as this Uganda case study offer roadmaps guiding prevention progress framed around health equity. The imperative now lies in scaling proven geospatial techniques globally translating localized intelligence into impactful risk-stratified resource allocation nationwide. Dedicated capacity building and technical assistance programs centered on cancer data informatics will prove essential for realizing this potential. Ultimately, GIS and spatial analytics should grow into routine indispensable real-time decision support infrastructures informing life-saving outreach worldwide.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003eLimitations\u003c/h2\u003e \u003cp\u003eDespite providing unique localized insights, limitations in the data and methods warrant consideration when interpreting the results. The reliance on aggregated secondary data precludes individual-level risk ascertainment prone to ecological fallacy. Furthermore, registry and survey sampling issues may still yield coverage gaps affecting representativeness. Spatial scale effects related to the modifiable areal unit problem could also arise using alternate district versus sub county geographical units. Additionally, the cross-sectional analysis lacked temporal trend analysis, which should be subject to follow-up post intervention. Analytical limitations around spatial autocorrelation, outlier influence, residual confounding, and linearity assumptions may also affect model fits. Finally, as an applied case study, external validity beyond the specific southwestern Uganda context remains undetermined pending further multiregional confirmatory analyses.\u003c/p\u003e \u003c/div\u003e"},{"header":"Conclusion","content":"\u003cp\u003eIn an environment of significant resource constraints yet a rising NCD burden, this spatial modeling research generates actionable intelligence that clarifies who and where to target limited prevention resources. The findings revealed statistically significant and policy-relevant geospatial disparities in cervical cancer incidence that were not evident using Geo-spatial methods. The integrated methodology combining hotspot analysis, spatial correlations, buffer zone mapping, and multivariate indicator modeling provides ample inductive insights to guide practitioners seeking to enhance intervention efficiency through geographic targeting. Spatial techniques should serve as a standard element of cancer control planning processes in Uganda and analogous low- to middle-income countries.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cp\u003eMRRH- Mbarara Regional Referral Hospital\u003c/p\u003e\n\u003cp\u003eHIV-human immunodeficiency virus\u003c/p\u003e\n\u003cp\u003eGI- Getis-Ord statistics\u003c/p\u003e\n\u003cp\u003eGIS- Geographical Information Systems\u003c/p\u003e\n\u003cp\u003eMOH-Ministry of Health\u003c/p\u003e\n\u003cp\u003eDHS-Demographic Health Survey\u003c/p\u003e\n\u003cp\u003eHPV-human papillomavirus\u003c/p\u003e\n\u003cp\u003eMUST-Mbarara University of Science and Technology\u003c/p\u003e\n\u003cp\u003eUBOS- Uganda Bureau of Statistics\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAvailability of data and materials\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot Applicable\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting Interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot Applicable\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot Applicable-Publication is a result of a self-sponsored Masters research contribution\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthors' contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eT.E. conducted the primary research, data collection, and analysis under the supervision of S.K. and R.K. T.E. drafted the manuscript and prepared the figures. S.K. and R.K. provided guidance, oversight, and critical review throughout the research process. All authors contributed to the interpretation of results and the final manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e“We would like to acknowledge Dr. Edgar Mugema Mulogo, Associate Professor in the Department of Community Health, for his invaluable mentorship and guidance throughout the research process. We also extend our gratitude to Dr. Gertrude Kiwanuka, Associate Professor in the Department of Biochemistry, for their mentorship and support.” Both of Mbarara University of Science and Technology.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eArbyn, M., Smith, S. B., \u0026amp; Temin, S. (2020). Detecting cervical precancer and reaching underscreened women through innovative screening approaches: a role for self-sampling. International Journal of Cancer, 148(3), 465-474.\u003c/li\u003e\n\u003cli\u003eBingi, D., Gidudu, A., Okello, D., \u0026amp; Lutalo Mwesigwa, C. (2018). \u003cem\u003eSpatial analysis of cervical cancer and correlated factors\u003c/em\u003e.\u003c/li\u003e\n\u003cli\u003eG\u0026Uuml;LTEKİN, M., Ramirez, P., Broutet, N., \u0026amp; Hutubessy, R. (2020). World Health Organization call for action to eliminate cervical cancer globally. In \u003cem\u003eInternational Journal of Gynecological Cancer\u003c/em\u003e (Vol. 30, Issue 4).\u003c/li\u003e\n\u003cli\u003eNakaganda, A., Spencer, A., Mpamani, C., Nassolo, C., Nambooze, S., Wabinga, H., Gemmell, I., Jones, A., Orem, J., \u0026amp; Verma, A. (2024). Estimating regional and national cancer incidence in Uganda: a retrospective population-based study, 2013\u0026ndash;2017. \u003cem\u003eBMC Cancer\u003c/em\u003e, \u003cem\u003e24\u003c/em\u003e(1), 787.\u003c/li\u003e\n\u003cli\u003eSomdyala, N. I. M. (2022). \u003cem\u003eReducing the Burden of Cervix Cancer in a Rural Setting of South Africa: Understanding the Incidence of this Disease and Building Infrastructure towards Intervention\u003c/em\u003e.\u003c/li\u003e\n\u003cli\u003eUBOS. (2022). \u003cem\u003eThe Uganda Demographic Health Survey 2022 Main Report PowerPoint Presentations - Uganda Bureau of Statistics\u003c/em\u003e. https://www.ubos.org/the-uganda-demographic-health-survey-2022-main-report-powerpoint-presentations/\u003c/li\u003e\n\u003cli\u003eBruni, L., Albero, G., Serrano, B., Mena, M., G\u0026oacute;mez, D., Mu\u0026ntilde;oz, J., Bosch, F.X., de Sanjos\u0026eacute;, S. (2019). ICO/IARC Information Centre on HPV and Cancer (HPV Information Centre). Human Papillomavirus and Related Diseases in Uganda. Summary Report 17 June 2019.\u003c/li\u003e\n\u003cli\u003eDewhurst, S., Gelderblom, A., Ngatia, Z., Mbele, A. M., Njalale, Y., Orach, C. G., ... \u0026amp; Ramogola-Masire, D. (2016). Rapidly increasing Gap between supply and need for palliative care in rural Western Uganda\u0026ndash;need for a radical reappraisal of access strategies: observational study over 9 Years. JMIR public health and surveillance, 2(2).\u003c/li\u003e\n\u003cli\u003eL\u0026oacute;pez-Carr, D., Chen, C., Jankowska, M. M., Schwager, S. J., Gavini, S., Zadnik Stirn, L., ... \u0026amp; Kotturi, G. (2017). A spatial poverty index to map health inequity across small areas: an illustration from El Salvador. International Journal of Health Geographics, 16(1), 1-16.\u003c/li\u003e\n\u003cli\u003eSung H., Ferlay J., Siegel R.L., Laversanne M., Soerjomataram I., Jemal A., Bray F. (2021). Global Cancer Statistics 2020: GLOBOCAN Estimates of Incidence and Mortality Worldwide for 36 Cancers in 185 Countries. CA Cancer J Clin. 2021 May\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Cervical cancer, GIS, hotspot analysis, spatial regression, Uganda","lastPublishedDoi":"10.21203/rs.3.rs-4750777/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-4750777/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003e\u003cstrong\u003eBackground:\u003c/strong\u003e Cervical cancer is a major public health issue in Uganda, causing over 2,500 deaths annually. Despite the high burden, most Ugandan women lack access to preventive services such as vaccination and screening due to geographic, financial, and health system barriers. Studies have documented geographic disparities in cervical cancer outcomes; however, spatial techniques have been underutilized to guide resource allocation. This research aimed to develop a geospatial model to identify priority areas for intervention.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAim: \u003c/strong\u003eTo enhance cervical cancer surveillance through geospatial modeling by quantifying geographic distributions, spatial relationships, and accessibility gaps in southwestern Uganda.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMethods:\u003c/strong\u003e Cervical cancer registry data from the MRRH were integrated with granular spatial information on health facilities, roads, sociodemographic, sexual behaviors, and HIV prevalence to enable multifaceted investigations of risk factors shaping observed geographic cancer patterns. Hotspot analysis, spatial regression, buffer analysis, and weighted risk overlay modeling were implemented to guide the modeling.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eResults: \u003c/strong\u003eSignificant hotspots of high cervical cancer incidence were identified in Buhweju (Gi* z score: 4.21, p\u0026lt;0.001) and Ibanda (Gi* z score: 3.87, p=0.002), where rates spiked over 100 per 100,000 women. Rural populations also had 2.15-fold greater odds of having late-stage disease. Women residing more than 5 km from the nearest screening site faced 38% lower odds of utilizing services compared to those residing within 5 km, highlighting severe distance decay effects on prevention.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConclusions:\u003c/strong\u003e Significant geographic disparities exist in cervical cancer burden and prevention access in southwestern Uganda. Spatial modeling provided actionable evidence for decentralizing services to address barriers.\u003c/p\u003e","manuscriptTitle":"Leveraging Geo-Spatial Data to Model Cervical Cancer Incidence Patterns in Southwestern Uganda.","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-08-12 16:34:16","doi":"10.21203/rs.3.rs-4750777/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"b61f4be7-b1bf-4a5a-a6d4-afd47f79ad01","owner":[],"postedDate":"August 12th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2025-09-30T04:38:51+00:00","versionOfRecord":[],"versionCreatedAt":"2024-08-12 16:34:16","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-4750777","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-4750777","identity":"rs-4750777","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}