{"id":125070,"date":"2026-09-03T10:17:35","date_gmt":"2026-09-03T10:17:35","guid":{"rendered":"https:\/\/www.europesays.com\/ch\/125070\/"},"modified":"2026-09-03T10:17:35","modified_gmt":"2026-09-03T10:17:35","slug":"spatiotemporal-dynamics-of-breast-cancer-screening-across-half-a-million-invitations-in-geneva-switzerland","status":"publish","type":"post","link":"https:\/\/www.europesays.com\/ch\/125070\/","title":{"rendered":"Spatiotemporal dynamics of breast cancer screening across half a million invitations in Geneva, Switzerland"},"content":{"rendered":"<p>DataBreast cancer screening participation data<\/p>\n<p>We analyzed data from 589,879 invitations (135,772 women) from the Geneva canton\u2019s BCS program (1999\u20132020), targeting women aged 50 to 74<a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 21\" title=\"Suisse contre le cancer - wwwliguecancerch L. Normes de qualit&#xE9; pour le d&#xE9;pistage organis&#xE9; du cancer du sein en Suisse. (2014).\" href=\"http:\/\/www.nature.com\/articles\/s43856-026-01451-7#ref-CR21\" id=\"ref-link-section-d123504962e655\" rel=\"nofollow noopener\" target=\"_blank\">21<\/a>,<a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 22\" title=\"Ligue contre le cancer. Le D&#xE9;pistage Du Cancer Du Sein Par Mammographie. &#010;                  https:\/\/doi.org\/10.2019\/021490014241&#010;                  &#010;                 (2017).\" href=\"http:\/\/www.nature.com\/articles\/s43856-026-01451-7#ref-CR22\" id=\"ref-link-section-d123504962e658\" rel=\"nofollow noopener\" target=\"_blank\">22<\/a> (Supplementary Fig.\u00a0<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"supplementary material anchor\" href=\"http:\/\/www.nature.com\/articles\/s43856-026-01451-7#MOESM1\" rel=\"nofollow noopener\" target=\"_blank\">1<\/a>). The first invitation is sent when women turn 50, with subsequent invitations issued every two years after a screening mammogram or following the previous invitation if no response was received. Each invitation record includes the date of invitation, the participation status (Yes\/No), residential address, and age group. For women who actually participated in mammography screening (i.e., participants), additional information on the chosen screening center was recorded. Additional details on the screening program\u2019s historical context are provided in Supplementary Note\u00a0<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"supplementary material anchor\" href=\"http:\/\/www.nature.com\/articles\/s43856-026-01451-7#MOESM1\" rel=\"nofollow noopener\" target=\"_blank\">1<\/a>. The study was conducted in collaboration with the Geneva Foundation for Cancer Screening, who provided access to anonymized invitation records from their administrative database. Ethical approval was obtained from the Cantonal Research Ethics Commission of Geneva, Switzerland (reference number 2023-02097, 15.01.2024). The requirement for informed consent was waived by the ethics committee under Article 34 of the Swiss Human Research Act. This waiver was granted due to the retrospective nature of the study using coded data from over 100,000 participants across 2003\u20132020, where obtaining individual consent would be impracticable.<\/p>\n<p>We identified a systematic underestimation of participation in the data covering the early implementation phase of the BCS program (1999\u20132003), these years were therefore excluded from analysis. The final dataset comprised 482,318 invitations (118,232 women) covering the period 2003\u20132020.<\/p>\n<p>Residential addresses were geocoded to obtain spatial coordinates for subsequent spatiotemporal analysis using a string-matching algorithm against the Swiss Buildings database (RegBL)<a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 23\" title=\"OFS. Registre f&#xE9;d&#xE9;ral des b&#xE2;timents et des logements. Published 2017. Accessed 28 April 2020. &#010;                  https:\/\/www.housing-stat.ch\/index_fr.html&#010;                  &#010;                .\" href=\"http:\/\/www.nature.com\/articles\/s43856-026-01451-7#ref-CR23\" id=\"ref-link-section-d123504962e674\" rel=\"nofollow noopener\" target=\"_blank\">23<\/a>. The algorithm first standardizes addresses by removing special characters, expanding abbreviations (e.g., \u201cch.\u201d to \u201cchemin\u201d), and extracting street numbers. It then employs a two-step matching process<a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 24\" title=\"De Ridder, D., Arzel, B., Joost, S. &amp; Guessous, I. Software: Spatiotemporal dynamics of breast cancer screening across half a million invitations in Geneva, Switzerland. &#010;                  https:\/\/doi.org\/10.5281\/ZENODO.18482901&#010;                  &#010;                . (2026).\" href=\"http:\/\/www.nature.com\/articles\/s43856-026-01451-7#ref-CR24\" id=\"ref-link-section-d123504962e678\" rel=\"nofollow noopener\" target=\"_blank\">24<\/a>: quick geocoding matches addresses within the same postal code using sequence similarity ratios, and if match quality is insufficient, slow geocoding expands the search to canton level. When street numbers were missing or not found, street-level geocoding was performed using the centroid of the street. Similarity scores are calculated using the SequenceMatcher algorithm from the difflib package (v3.13.7)<a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 25\" title=\"difflib &#x2014; Helpers for computing deltas &#x2014; Python 3.11.4 documentation. Accessed 30 June 2023. &#010;                  https:\/\/docs.python.org\/3\/library\/difflib.html&#010;                  &#010;                .\" href=\"http:\/\/www.nature.com\/articles\/s43856-026-01451-7#ref-CR25\" id=\"ref-link-section-d123504962e682\" rel=\"nofollow noopener\" target=\"_blank\">25<\/a> in Python (v3.11), with matches above an 80% threshold being classified as acceptable quality. This threshold was selected to optimize the balance between false positives (incorrect matches passing the threshold) and false negatives (existing matches not being identified). Manual review and cleaning were conducted for addresses near the 80% threshold to correct false negatives (valid matches below threshold) and false positives (invalid matches above threshold). This completely offline procedure avoids the need to send information online (e.g., via APIs such as Google Maps) and therefore ensures improved\u00a0data protection.<\/p>\n<p>Exclusion criteria were applied to align with national age guidelines and to ensure geocoding accuracy (details in Supplementary Note\u00a0<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"supplementary material anchor\" href=\"http:\/\/www.nature.com\/articles\/s43856-026-01451-7#MOESM1\" rel=\"nofollow noopener\" target=\"_blank\">2<\/a>).<\/p>\n<p>Analytical data structure<\/p>\n<p>Our analysis employed both individual-level and area-level approaches. Individual-level invitation records were used for descriptive analyses of participation trends across demographic and socioeconomic characteristics (Supplementary Table\u00a0<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"supplementary material anchor\" href=\"http:\/\/www.nature.com\/articles\/s43856-026-01451-7#MOESM1\" rel=\"nofollow noopener\" target=\"_blank\">1<\/a>).<\/p>\n<p>For spatial and spatiotemporal analyses, individual records were aggregated to neighborhood-level participation rates. Geocoded invitations were consolidated into biennial periods and aggregated in 2121 inhabited neighborhoods to ensure uniform geographic units over time using Swiss neighborhoods, developed by MicroGIS<a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 26\" title=\"microgis.ch:: DONNEES STATISTIQUES. Accessed August 6, 2020. &#010;                  http:\/\/microgis.ch\/donnees\/donnees-statistiques&#010;                  &#010;                .\" href=\"http:\/\/www.nature.com\/articles\/s43856-026-01451-7#ref-CR26\" id=\"ref-link-section-d123504962e706\" rel=\"nofollow noopener\" target=\"_blank\">26<\/a>. These neighborhoods are defined as homogeneous units corresponding to urban blocks. Importantly, these neighborhoods do not have a fixed spatial scale but are delineated based on population density and land-use patterns. We used biennial periods for consistency with participation trends, labeling each period by the even-numbered year (e.g., 2003 and 2004 as \u201c2004\u201d). This spatial aggregation enables detection of geographic clustering patterns and linkage with area-level covariates, but transforms the analytical unit from individual screening decisions to contextual participation patterns at the neighborhood level.<\/p>\n<p>Individuals located outside populated neighborhoods (1394 out of 482,318 invitations, 0.29%) were assigned to the nearest inhabited neighborhood using the nearest_points function from the Shapely library in Python. These cases likely reflect temporal changes in population distribution, as the population data is contemporaneous to BCS invitations that span 2003\u20132020. Some individuals may have resided in areas that were populated during the earlier study periods but are no longer inhabited according to the more recent neighborhood data. Those at intersections of multiple neighborhoods were randomly assigned to prevent duplication.<\/p>\n<p>For each neighborhood and biennial period, we calculated raw participation rates as the ratio between the number of participants who attended mammography screening (within the cantonal BCS program) and the number of invitations sent, multiplied by 100. We also computed mean values for other covariates within each spatial unit. To address unequal variances due to the disparate distribution of individuals, we used Spatial Empirical Bayes smoothing (SEBS) with 8 nearest neighbors<a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 26\" title=\"microgis.ch:: DONNEES STATISTIQUES. Accessed August 6, 2020. &#010;                  http:\/\/microgis.ch\/donnees\/donnees-statistiques&#010;                  &#010;                .\" href=\"http:\/\/www.nature.com\/articles\/s43856-026-01451-7#ref-CR26\" id=\"ref-link-section-d123504962e719\" rel=\"nofollow noopener\" target=\"_blank\">26<\/a> implemented using the smoothing module from the Exploratory Spatial Data Analysis (ESDA, v2.7.0) package within Python Spatial Analysis Library (PySAL, v25.1)<a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 27\" title=\"Waller, L.A. &amp; Gotway, C.A. Applied spatial statistics for public health data. Accessed 1 July 2004. &#010;                  https:\/\/doi.org\/10.1002\/0471662682&#010;                  &#010;                .\" href=\"http:\/\/www.nature.com\/articles\/s43856-026-01451-7#ref-CR27\" id=\"ref-link-section-d123504962e723\" rel=\"nofollow noopener\" target=\"_blank\">27<\/a>. The 8-neighbor parameter provides conservative smoothing to stabilize rates in neighborhoods with limited invitations while minimizing spatial information borrowing.<\/p>\n<p>Socioeconomic status<\/p>\n<p>Socioeconomic status was determined using a neighborhood-level socioeconomic deprivation index, as individual-level SES data were not available through the BCS program. We used the same Swiss neighborhoods<a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 26\" title=\"microgis.ch:: DONNEES STATISTIQUES. Accessed August 6, 2020. &#010;                  http:\/\/microgis.ch\/donnees\/donnees-statistiques&#010;                  &#010;                .\" href=\"http:\/\/www.nature.com\/articles\/s43856-026-01451-7#ref-CR26\" id=\"ref-link-section-d123504962e735\" rel=\"nofollow noopener\" target=\"_blank\">26<\/a>, which represent the finest operational level for socioeconomic analysis. The procedure to create the deprivation index was based on the methodology from Lallou\u00e9 et al.<a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 28\" title=\"Lallou&#xE9;, B. et al. A statistical procedure to create a neighborhood socioeconomic index for health inequalities analysis. Int. J. Equity Health 12, 21 (2013).\" href=\"http:\/\/www.nature.com\/articles\/s43856-026-01451-7#ref-CR28\" id=\"ref-link-section-d123504962e739\" rel=\"nofollow noopener\" target=\"_blank\">28<\/a>, with adaptation to data available at the neighborhood scale in the canton of Geneva as described in our previous work<a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 18\" title=\"De Ridder, D. et al. Socioeconomically disadvantaged neighborhoods face increased persistence of SARS-CoV-2 clusters. Front. Public Health 8, 626090 (2021).\" href=\"http:\/\/www.nature.com\/articles\/s43856-026-01451-7#ref-CR18\" id=\"ref-link-section-d123504962e743\" rel=\"nofollow noopener\" target=\"_blank\">18<\/a>. The deprivation index was standardized to a 0\u20131 scale using min-max normalization and inverted such that higher values represent greater deprivation (1 = highest deprivation, 0 = lowest deprivation), ensuring consistent interpretation across all analyses.<\/p>\n<p>Physical accessibility to cancer screening centers<\/p>\n<p>Physical accessibility to cancer screening centers was assessed by computing the density of screening centres within a 5000-m radius for each individual, factoring in a linear decay function and street network distances<a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 29\" title=\"OpenStreetMap. Accessed 15 July 2024. &#010;                  https:\/\/www.openstreetmap.org\/#map=8\/46.825\/8.224&#010;                  &#010;                \" href=\"http:\/\/www.nature.com\/articles\/s43856-026-01451-7#ref-CR29\" id=\"ref-link-section-d123504962e755\" rel=\"nofollow noopener\" target=\"_blank\">29<\/a> using the Python package Pandana<a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 30\" title=\"Foti, F., Waddell, P., Luxen, D. A Generalized Computational Framework for Accessibility: From the Pedestrian to the Metropolitan Scale. 2012. Accessed 25 March 2020. &#010;                  https:\/\/onlinepubs.trb.org\/onlinepubs\/conferences\/2012\/4thITM\/Papers-A\/0117-000062.pdf&#010;                  &#010;                .\" href=\"http:\/\/www.nature.com\/articles\/s43856-026-01451-7#ref-CR30\" id=\"ref-link-section-d123504962e759\" rel=\"nofollow noopener\" target=\"_blank\">30<\/a>. The 5-km radius was selected to capture local accessibility, primarily via walking, cycling, public transit and shorter-distance private vehicle trips in Geneva\u2019s dense urban context, while the linear decay function ensures closer centers receive appropriate priority weighting. This approach acknowledges that women may access specialized healthcare services across substantial distances using available local transportation options. Screening center locations, operational periods, and program participation status were obtained from Geneva\u2019s organized BCS program, with openings and closures incorporated to reflect temporal changes in accessibility within the organized screening network.<\/p>\n<p>Ethical approval<\/p>\n<p>The study was approved by the Cantonal Research Ethics Commission of Geneva, Switzerland (2023-02097).<\/p>\n<p>Spatiotemporal analysesHot spot analysis &#8211; Getis-Ord Gi* statistic<\/p>\n<p>We analyzed the spatial dependence of breast screening participation rates per neighborhood to identify geographic patterns among beneficiaries. Spatial dependence, rooted in Tobler\u2019s first law of geography, suggests that nearby observations are more likely to exhibit similar attributes due to shared environmental, social, and economic conditions<a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 31\" title=\"Tobler, W. R. A computer movie simulating urban growth in the detroit region. Econ. Geogr. 46, 234 (1970).\" href=\"http:\/\/www.nature.com\/articles\/s43856-026-01451-7#ref-CR31\" id=\"ref-link-section-d123504962e784\" rel=\"nofollow noopener\" target=\"_blank\">31<\/a>.<\/p>\n<p>Using the Getis-Ord Gi* statistic, we identified spatial clusters\u2014or \u201chot spots\u201d and \u201ccold spots\u201d\u2014where local (i.e., within a specified spatial lag) screening rates were significantly higher or lower than the global average<a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 32\" title=\"Getis, A. &amp; Ord, J. K. The analysis of spatial association by use of distance statistics. Geogr. Anal. 24, 189&#x2013;206 (1992).\" href=\"http:\/\/www.nature.com\/articles\/s43856-026-01451-7#ref-CR32\" id=\"ref-link-section-d123504962e791\" rel=\"nofollow noopener\" target=\"_blank\">32<\/a>. Our analytical approach focused directly on local cluster identification to address our applied public health objectives of identifying specific geographic areas with high and low participation for targeted intervention, as Getis-Ord Gi can identify statistically significant local clustering patterns even when global spatial autocorrelation may not be significant.<\/p>\n<p>The spatial lag refers to the neighborhood definition used to calculate local statistics\u2014specifically, the set of neighborhoods whose values are included in the local calculation for each focal neighborhood. For hot spot identification, we evaluated spatial patterns using K-nearest neighbors (KNN) weight matrices: K\u2009=\u20098 and K\u2009=\u200916 nearest neighborhoods for each spatial unit. The spatial relationship conceptualization was based on neighborhood centroids using Euclidean distance calculations. We applied binary weights where neighbors were assigned a value of 1 and non-neighbors 0, with no additional row standardization required due to equal neighbor counts across all spatial units.<\/p>\n<p>We selected KNN weights over contiguity-based weights, as many units lack sufficient contiguous neighbors. We also chose KNN over distance-based weights to ensure each neighborhood has a consistent number of neighbors for spatial lag calculations, providing methodological consistency despite varying distances to neighbors across different density contexts.<\/p>\n<p>Statistical significance was assessed using Monte Carlo permutation testing with 9999 permutations to empirically derive pseudo-p-values. Hot spots and cold spots were classified using multiple significance thresholds (p\u2009&lt;\u20090.001, p\u2009&lt;\u20090.01, p\u2009&lt;\u20090.05, p\u2009&lt;\u20090.1) to provide graduated significance levels, with higher statistical confidence represented by more intense colors in mapping visualizations. The spatial patterns were more pronounced with k\u2009=\u200916 neighbors, hence results from this neighborhood definition were retained for final analysis.<\/p>\n<p>Space-time pattern mining\u2013Emerging hot spot analysis<\/p>\n<p>To discern shifts in the patterns of participation in the screening program, we utilized a space-time pattern mining (STPM) technique also known as emerging hot spot analysis<a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 33\" title=\"How Emerging Hot Spot Analysis works&#x2014;ArcGIS Pro | Documentation. Accessed 13 June, 2023. &#010;                  https:\/\/pro.arcgis.com\/en\/pro-app\/latest\/tool-reference\/space-time-pattern-mining\/learnmoreemerging.htm&#010;                  &#010;                .\" href=\"http:\/\/www.nature.com\/articles\/s43856-026-01451-7#ref-CR33\" id=\"ref-link-section-d123504962e834\" rel=\"nofollow noopener\" target=\"_blank\">33<\/a>. This method is designed to spot evolving temporal trends in spatial data, including the emergence, intensification, reduction, and sporadic occurrence of hot and cold spots. This method is mainly available through the ArcGIS\u00ae pro software<a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 34\" title=\"ESRI. Desktop GIS Software | Mapping Analytics | ArcGIS Pro. Published online 2024. Accessed 28 April 2024. &#010;                  https:\/\/www.esri.com\/en-us\/arcgis\/products\/arcgis-pro\/overview&#010;                  &#010;                .\" href=\"http:\/\/www.nature.com\/articles\/s43856-026-01451-7#ref-CR34\" id=\"ref-link-section-d123504962e838\" rel=\"nofollow noopener\" target=\"_blank\">34<\/a>, it can also be implemented in Python<a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 35\" title=\"Van Rossum, G. &amp; Drake, F.L. Python 3 Reference Manual. CreateSpace; 2009.\" href=\"http:\/\/www.nature.com\/articles\/s43856-026-01451-7#ref-CR35\" id=\"ref-link-section-d123504962e842\" rel=\"nofollow noopener\" target=\"_blank\">35<\/a>.<\/p>\n<p>This analysis first requires creating a 3D space-time cube composed of bins defined by space and time parameters. The bin size was set at the neighborhood scale, which enables the detection of very local urban patterns and allows a convenient spatial linkage with multiple sociodemographic descriptors (See section \u201cData \u2013 Socioeconomic status\u201d). The biennial time parameter was chosen to align with the invitation schedule of the BCS program. The resultant 3D space-time cube was composed of a maximum of nine-time intervals and 2121 spatial bins (i.e., neighborhoods), yielding up to 19,089 unique space-time bins.<\/p>\n<p>We utilized the Getis-Ord Gi* statistic to evaluate each bin within the space-time cube and employed a false discovery rate (FDR) correction during this process to account for multiple hypothesis testing. Like the Getis-Ord Gi* analysis, each bin was assigned a Z-score, a p-value, and a category. Overall trends were evaluated using the Mann-Kendall trend test, a non-parametric test used to identify statistically significant trends. Based on the p-values, z-scores, and Mann-Kendall statistics, each location was then categorized into a maximum of 17 predefined categories (1 no pattern, 8 cold spots, 8 hot spots) (Table\u00a0<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"table anchor\" href=\"http:\/\/www.nature.com\/articles\/s43856-026-01451-7#Tab1\" rel=\"nofollow noopener\" target=\"_blank\">1<\/a>).<\/p>\n<p>Table 1 Classification of spatiotemporal patterns into 17 categories using space-time pattern mining (emerging hot spot analysis)<\/p>\n<p>Pattern classification followed ESRI\u2019s established emerging hot spot analysis methodology, including the standard 90% threshold parameter that determines temporal consistency requirements for certain categories. The terminology used in pattern categories follows ESRI\u2019s official framework to ensure methodological consistency and reproducibility. We implemented one modification: the Historical pattern threshold was reduced from 90 to 50% of time steps to better capture meaningful historical patterns in our dataset context.<\/p>\n<p>Some less populated neighborhoods did not appear in all nine biennial intervals, likely due to the intermittent presence of women within the target age range (Supplementary Fig.\u00a0<a data-track=\"click\" data-track-label=\"link\" data-track-action=\"supplementary material anchor\" href=\"http:\/\/www.nature.com\/articles\/s43856-026-01451-7#MOESM1\" rel=\"nofollow noopener\" target=\"_blank\">2<\/a>).<\/p>\n<p>All maps were created using Python (matplotlib, geopandas) with base geographic layers obtained from the Geneva Territory Information System (SITG), including administrative boundaries of the canton of Geneva, municipalities, lake contours, and rivers<a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 36\" title=\"Catalogue SITG. Accessed 9 December 2025. &#010;                  https:\/\/sitg.ge.ch\/&#010;                  &#010;                .\" href=\"http:\/\/www.nature.com\/articles\/s43856-026-01451-7#ref-CR36\" id=\"ref-link-section-d123504962e1076\" rel=\"nofollow noopener\" target=\"_blank\">36<\/a>.<\/p>\n<p>Statistical analyses<\/p>\n<p>We employed the eXtreme Gradient Boosting (XGBoost) method, an established machine learning algorithm for regression applications, to assess the relative importance of selected area-level determinants in predicting neighborhood-level BCS participation rates. XGBoost (xgboost package, v3.0.5) was selected for its\u00a0ability to capture complex non-linear relationships and interactions between predictors while maintaining interpretability through SHapley Additive exPlanations (SHAP) values<a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 37\" title=\"Li, Z. Extracting spatial effects from machine learning model using local interpretation method: An example of SHAP and XGBoost. Comput. Environ. Urban Syst. 96, 101845 (2022).\" href=\"http:\/\/www.nature.com\/articles\/s43856-026-01451-7#ref-CR37\" id=\"ref-link-section-d123504962e1089\" rel=\"nofollow noopener\" target=\"_blank\">37<\/a>,<a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 38\" title=\"Chen, T. &amp; Guestrin, C. XGBoost: a scalable tree boosting system. Published online 2016. &#010;                  https:\/\/doi.org\/10.1145\/2939672.2939785&#010;                  &#010;                .\" href=\"http:\/\/www.nature.com\/articles\/s43856-026-01451-7#ref-CR38\" id=\"ref-link-section-d123504962e1092\" rel=\"nofollow noopener\" target=\"_blank\">38<\/a>.<\/p>\n<p>The analysis used neighborhood-level participation rates from the final biennial period (2019\u20132020) as the outcome variable. This contemporary cross-sectional approach focuses on identifying current determinants most relevant for intervention planning, given that our primary predictor variables are measured at a single time point or exhibit minimal temporal variation. In the 2019\u20132020 period, the analytical dataset comprised 1983 inhabited neighborhoods with screening invitation data.<\/p>\n<p>Predictor variables included the neighborhood socioeconomic deprivation index, density of screening centers, age group composition of invited women (proportions in five-year categories: 50\u201354, 55\u201359, 60\u201364, 65\u201369, 70\u201374 years), and spatial coordinates to capture residual geographic effects not explained by measured covariates.<\/p>\n<p>Data were randomly partitioned into training (80%) and test (20%) sets. Hyperparameters were optimized using Bayesian optimization with five-fold cross-validation to minimize root mean squared error using the hyperopt package (v0.2.7)<a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 39\" title=\"Bergstra, J., Yamins, D. &amp; Cox, D. Making a science of model search: hyperparameter optimization in hundreds of dimensions for vision architectures. 28, 115&#x2013;123. Accessed 1 October 1, 2025. &#010;                  https:\/\/proceedings.mlr.press\/v28\/bergstra13.html&#010;                  &#010;                 (2013).\" href=\"http:\/\/www.nature.com\/articles\/s43856-026-01451-7#ref-CR39\" id=\"ref-link-section-d123504962e1105\" rel=\"nofollow noopener\" target=\"_blank\">39<\/a>. The optimized model was trained on the training set with early stopping to prevent overfitting, then evaluated on the held-out test set to assess generalization performance.<\/p>\n<p>Model interpretation employed SHAP values using the shap package (v0.48.0), which quantify each predictor\u2019s contribution to participation rate predictions while accounting for interactions between variables<a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 39\" title=\"Bergstra, J., Yamins, D. &amp; Cox, D. Making a science of model search: hyperparameter optimization in hundreds of dimensions for vision architectures. 28, 115&#x2013;123. Accessed 1 October 1, 2025. &#010;                  https:\/\/proceedings.mlr.press\/v28\/bergstra13.html&#010;                  &#010;                 (2013).\" href=\"http:\/\/www.nature.com\/articles\/s43856-026-01451-7#ref-CR39\" id=\"ref-link-section-d123504962e1113\" rel=\"nofollow noopener\" target=\"_blank\">39<\/a>. SHAP values provide both variable importance rankings and directional effects, enabling identification of key determinants and their relationships with screening participation<a data-track=\"click\" data-track-action=\"reference anchor\" data-track-label=\"link\" data-test=\"citation-ref\" aria-label=\"Reference 40\" title=\"Auchincloss, A. H., Gebreab, S. Y., Mair, C. &amp; Diez Roux, A. V. A review of spatial methods in epidemiology, 2000-2010. Annu. Rev. Public Health 33, 107&#x2013;122 (2012).\" href=\"http:\/\/www.nature.com\/articles\/s43856-026-01451-7#ref-CR40\" id=\"ref-link-section-d123504962e1117\" rel=\"nofollow noopener\" target=\"_blank\">40<\/a>.<\/p>\n","protected":false},"excerpt":{"rendered":"DataBreast cancer screening participation data We analyzed data from 589,879 invitations (135,772 women) from the Geneva canton\u2019s BCS&hellip;\n","protected":false},"author":2,"featured_media":125071,"comment_status":"","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"footnotes":"","_share_on_mastodon":"0"},"categories":[7],"tags":[207,25381,2839,584,65,4090,61068],"class_list":["post-125070","post","type-post","status-publish","format-standard","has-post-thumbnail","category-geneva","tag-breast-cancer","tag-cancer-screening","tag-epidemiology","tag-general","tag-geneva","tag-medicine-public-health","tag-preventive-medicine"],"share_on_mastodon":{"url":"https:\/\/pubeurope.com\/@ch\/117206597606536977","error":""},"_links":{"self":[{"href":"https:\/\/www.europesays.com\/ch\/wp-json\/wp\/v2\/posts\/125070","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.europesays.com\/ch\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.europesays.com\/ch\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.europesays.com\/ch\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/www.europesays.com\/ch\/wp-json\/wp\/v2\/comments?post=125070"}],"version-history":[{"count":0,"href":"https:\/\/www.europesays.com\/ch\/wp-json\/wp\/v2\/posts\/125070\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.europesays.com\/ch\/wp-json\/wp\/v2\/media\/125071"}],"wp:attachment":[{"href":"https:\/\/www.europesays.com\/ch\/wp-json\/wp\/v2\/media?parent=125070"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.europesays.com\/ch\/wp-json\/wp\/v2\/categories?post=125070"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.europesays.com\/ch\/wp-json\/wp\/v2\/tags?post=125070"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}