CaRDO: A Tool for Empowering Cancer Registries to Publish Population-Level Cancer Statistics Online

Abstract

Introduction:

Population-based cancer registries are critical for monitoring cancer burden, informing policy, and guiding cancer control strategies. However, many registries face challenges in effectively communicating their data, due to limited resources and technical capacity. There is a growing need for accessible tools that support the visualization and dissemination of cancer statistics in engaging and interpretable formats.

Methods:

CaRDO (Cancer Registry Data Online) was developed by the Viertel Cancer Research Centre at Cancer Council Queensland as a free, secure, and user-friendly R package for creating interactive dashboards using population-level cancer data. Built on the R Shiny framework, CaRDO enables users with minimal programming experience to generate dashboards entirely offline. The tool incorporates standard epidemiological methods, including age-standardized rates, lifetime risk estimation, and segmented regression for trend analysis, while maintaining strong privacy safeguards.

Results:

CaRDO allows users to display cancer incidence and mortality data by sex, cancer type, year, age group, and measure (counts or rates), with visualizations of underlying temporal trends. All data processing occurs locally, and dashboards can be customized with plots that can be downloaded for use in reports or presentations. Example datasets and detailed guidance are provided to support implementation. The current version includes core dashboard elements, with flexibility for future expansion.

Conclusions:

CaRDO provides a timely and practical solution for enhancing the visibility and utility of cancer registry data. By removing financial and technical barriers, it supports under-resourced registries in fulfilling their essential role in cancer control through transparent, datadriven public health communication.

Introduction

Population-based cancer registries are essential for understanding the impact that cancer has at a population level. By systematically collecting standardized data, cancer registries provide a critical platform for tracking cancer incidence, survival, and mortality; monitoring trends over time; and identifying disparities in cancer indicators across population subgroups and geographical areas. They also provide a unique data-driven evidence base to support the development, implementation, and evaluation of cancer control efforts, from prevention and screening strategies to broader health system planning.^1,2

The global cancer burden is rising rapidly. In 2022, there were an estimated 20 million new cancer diagnoses and 9.7 million cancer deaths. By 2050, annual cancer diagnoses are projected to reach 35 million, a 77% increase.³ While outcomes have improved for some population groups and cancer types, significant disparities persist.³ At the same time, academic research faces growing challenges, including funding shortfalls, public mistrust, pressure for translational research, and disconnects between evidence and policy.⁴ In this context, it is more important than ever to effectively communicate the value of routine cancer surveillance through cancer registries.⁴

Making information derived from cancer registry data more accessible to policymakers, healthcare professionals, and the public is one way to raise the profile of cancer registries and highlight their value. Cancer registries should be viewed not just as data repositories, but as essential tools for guiding evidence-based decision making. Yet a major barrier is the lack of affordable, user-friendly tools for presenting cancer statistics in an engaging and accessible format. Many registries operate with limited resources, making it difficult to invest in proprietary software or specialized expertise. As a result, their ability to produce high-quality outputs that inform policy and engage stakeholders is often constrained.

Several software packages—such as Power BI, Tableau, and others—are purpose-built for data visualization and dashboard development. These tools offer powerful capabilities and are widely used across industry, including health analytics. However, they often require licensing costs and technical expertise, which can pose challenges for smaller or under-resourced cancer registries. While well-suited for building interactive dashboards, they are not specifically designed to meet the unique needs of population-based cancer registries aiming to communicate broad cancer trends to non-technical audiences.

In addition to these general-purpose platforms, there are more specialized tools designed for clinical and molecular cancer research.^5,6,7 However, tools focused on visualizing and communicating population-level cancer outcomes in accessible formats remain limited.

Florensa and colleagues⁸ (2023) proposed a comprehensive workflow for visualizing cancer registry data using R Shiny dashboards. Their approach enhances data accessibility and transparency, supporting evidence-based decision-making. However, it relies on backend infrastructure—such as APIs and databases—to connect registry data to the dashboard application. This requirement may exceed the technical capacity of smaller or under-resourced registries.

To help bridge the gap in accessible online reporting tools for cancer registries, the Viertel Cancer Research Centre at Cancer Council Queensland has developed Cancer Registry Data Online (CaRDO). CaRDO is a free R package designed to support the creation of interactive dashboards using population-level cancer data. It was built specifically for registries with limited technical capacity—requiring no programming skills and ensuring that sensitive data remain offline.

Methods

CaRDO (Cancer Registry Data Online) was developed by the Viertel Cancer Research Centre at Cancer Council Queensland to provide a free, secure, and user-friendly solution for visualizing population-level cancer statistics. The tool was designed to support cancer registries with limited technical capacity, enabling them to create interactive dashboards without requiring programming expertise or online access to sensitive data.

Technical Framework

CaRDO is implemented as an R package using the R Shiny⁹ framework from the R programming language.¹⁰ It allows users to build interactive web-based dashboards that display cancer statistics in a clear and engaging format. The package is freely available and only requires installation of R Studio, which is also free. All data processing occurs locally on the user's computer, ensuring that sensitive information remains secure and offline.

Broadly, CaRDO works by standardizing the user's data structure and variable names, allowing the prebuilt R Shiny application to perform the required calculations on the supplied data. The package, including the software and its documentation, is only distributed via GitHub, a free online code-hosting platform, which allows users to install CaRDO directly from within R. The full source code is publicly available at https://github.com/ for readers and users wishing to explore the technical side of CaRDO in more detail. Importantly, GitHub is not used to store or process user data; all analyses are performed locally on the user's computer.

User Workflow

CaRDO was designed to be intuitive and accessible to users with minimal or no experience in R programming. The dashboard creation process is guided entirely through a graphical user interface, with no coding required. The workflow involves four main steps (see Figure 1): (1) loading data into R Studio; (2) specifying variables of interest; (3) customizing dashboard details; and (4) generating the final dashboard on the local computer.

Figure 1.

Steps Involved in Generating a Statistics Dashboard Using CaRDO. For further details, see The CaRDO Handbook, Chapter 4, CaRDO Example

Some initial data preparation is required prior to these steps, and detailed guidance is provided in the accompanying The CaRDO Handbook. The handbook also outlines best practices for data formatting, aggregation, and privacy protection. In addition, CaRDO has built-in error handling to identify common user issues—such as mis-specified variable names or invalid data types—and provides clear feedback to users to help them identify and correct these errors.

The content of the CaRDO dashboard depends on the scope of the data available. For example, the simplest dashboard users can create provides a display of cancer incidence count data. This requires details about cancer type, counts (number of diagnoses), diagnosis year, age at diagnosis (in 5-year age groups), and sex. This can then be expanded to calculate incidence rates, if the corresponding population data are available. If mortality data are available, they can be incorporated into the dashboard using the current version of CaRDO.

Statistical Methods

When counts and populations were provided, age-standardized rates were calculated using the direct method, applying age-specific rates to the world standard population across 5-year age groups. Lifetime risk estimates were derived using cumulative risk formulas, assuming constant age-specific rates up to age 85.

Trends in cancer incidence and mortality were modeled using segmented regression techniques. Breakpoints were identified using structural change detection methods, and piecewise models comprised of linear fits were applied to each segment.

Sex-specific cancers were identified algorithmically by assessing the presence of sex-coded data across cancer types. Where data are available for only one sex, the cancer type is classified as sex-specific, and so the person-specific population count is not used as the denominator.

Privacy and Data Governance

CaRDO was developed with strong privacy safeguards to ensure compliance with data governance standards. All data used in CaRDO are stored and analyzed locally; no data are transmitted or stored online. For incidence and mortality dashboards, only aggregated data are required, eliminating the need for unit record data. Even when sensitive data are used, CaRDO functions analogously to other secure data management software, importing and processing data entirely within the RStudio environment.

To further protect privacy, CaRDO includes several built-in features including (a) 5-year aggregation options to reduce the risk of identifying individuals in registries with small case numbers, (b) user-defined suppression thresholds to automatically exclude small cell counts from display, and (c) offline operation to ensure data never leave the user's device.

Development and Testing

The software was built using the following R packages: shiny software package for R,⁹ tidyverse,¹¹ bslib,¹² shinyWidgets,¹³ plotly,¹⁴ markdown,¹⁵ strucchange,¹⁶ and segmented.¹⁷

CaRDO was developed iteratively with input from internal and external stakeholders, including cancer researchers, cancer registrars, clinical coders, and data custodians, both within Australia and internationally. Feedback was collected through individual discussions, either in person or through videoconference and/or email, with the intent to refine usability and ensure clarity of the documentation and relevance to privacy standards and registry practices. Before deployment, CaRDO was tested using a range of internal and external datasets, including both real-world registry data and purpose-built fictitious datasets designed to test data security and suitability for sparse datasets. As this version of CaRDO represents the first iteration, ongoing user feedback is actively encouraged to inform future development and refinement.

Publishing the Dashboard

The focus of CaRDO is to produce a web-based dashboard that is stored on the local computer. CaRDO does not automatically publish the data online, thus providing the user with greater security while using CaRDO and more flexibility in terms of how they want to publish the dashboard online. There are several options available for publishing this dashboard, and their respective strengths and limitations have been summarized elsewhere.¹⁸

Example Datasets

To help users implement and test CaRDO, example datasets containing fictional incidence, mortality, and population data are provided to demonstrate the required data structure.

Results

An example of the dashboard that can be generated by CaRDO is provided in Figure 2. Depending on the data provided, the dashboard can display population-level cancer data by sex, cancer type, year, age group, and measure (rates or counts) for both incidence and mortality, and details of the underlying temporal trends.

Figure 2.

Example “Diagnosis” Page Dashboard Created Using CaRDO with Statistics Based on Fictional Data

All data displayed in the dashboard, as well as the figures themselves, are available for users to download, enabling straightforward incorporation into reports, presentations, or other materials.

Discussion

The development of CaRDO aligns with a growing recognition of the vital role cancer registries play in public health. Cancer registries are not merely data repositories—they provide a critical foundation essential to appropriately understand the impact of cancer within a population, guide policy, and drive improvements in cancer outcomes. Yet despite their importance, registries often operate under resource constraints and so face challenges in communicating their value to stakeholders. CaRDO directly addresses this gap by providing a free, user-friendly tool that enables registries to transform their confidential complex data into accessible, public-facing interactive dashboards.

By removing financial and technical barriers, CaRDO provides the potential to empower under-resourced registries to enhance the visibility and impact of their data. This democratization of data visualization supports broader efforts to champion cancer registries and raise awareness of their contributions to cancer control. In doing so, CaRDO helps registries better engage with policymakers, healthcare professionals, and the public—key audiences whose support is essential for sustaining and expanding registry operations.

The current version of CaRDO has been designed to include the core elements that might typically be expected in a standard online dashboard for a population-based cancer registry. Utilizing the inherent flexibility of R Shiny, there are many avenues for expansion, such as including various measures of survival and more details about trends and age-specific patterns, along with many options for presenting these statistics visually. The Queensland Cancer Statistics Online (QCSOL)¹⁹ provides one example of the possibilities that could be included with further development of the underlying CaRDO package (see Figure 3).

Figure 3.

Collage of Dashboards from Queensland Cancer Statistics Online (QCSOL), Demonstrating Potential for Future Expansion of CaRDO

The CaRDO software and associated materials are freely available under the Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License (CC BY-NC-SA 4.0). This license permits users to copy, redistribute, and adapt the materials for non-commercial purposes, provided appropriate credit is given and any derivative works are shared under the same license. This open licensing model supports collaborative development and equitable access, particularly for low-resource cancer registries. Organizations seeking customizations or enhancements beyond the base product are encouraged to contact the developers to discuss potential syntax development on a cost-recovery basis.

Conclusion

CaRDO provides a secure, accessible, and adaptable solution for visualizing population-level cancer data. Its intuitive interface and open-source design make it especially valuable for registries with limited resources, enabling them to produce interactive dashboards without programming expertise or compromising data privacy. Built on a flexible R Shiny framework, CaRDO supports a range of statistical outputs and can be expanded to meet evolving registry needs. As cancer burden continues to rise and the need for transparent, evidence-based public health responses grows, CaRDO provides a timely and impactful tool to help cancer registries fulfill their essential role in cancer control.

While CaRDO is designed to be user-friendly, users are welcome to contact statistics@cancerqld.org.au for general enquiries or support. If your organization requires customizations or enhancements to the base product, please contact us to discuss further. Feedback from users is actively encouraged, as understanding how CaRDO is adopted and used is critical to demonstrating its impact and supporting future development.