Table 1 Policy summary and key outcomes

Background

Quality improvement supported by continuous monitoring and evaluation of performance is vital for cancer screening programs. However, there is a lack of global data to evaluate cancer screening implementation in the real world, especially data from LMICs. This prevents such programs from setting their own benchmarks and compare performance with similar regional programs. IARC initiated the CanScreen5 project, a global cancer screening data repository, to fill this gap.

Main findings and limitations

Eighty-four countries joined the CanScreen5 project and shared qualitative or quantitative (or both) data on breast cancer, cervical cancer and CRC screening programs. Substantial heterogeneity exists between cancer sites and between countries regarding program organization, covering screening policies, protocols, governance, financing mechanisms, systems of invitation and recall, processing of data collected for program monitoring and quality assurance. Overall, organization was better in cancer screening programs from Europe than other continents. For specific cancer sites, CRC screening programs showed better organization. Similarly, considerable heterogeneity existed in program performance as noted through the estimation of harmonized indicators. Examination coverage ranged from 1.7% (Bangladesh) to 85.5% (England, United Kingdom) for breast cancer, from 2.1% (Côte d’Ivoire) to 86.3% (Sweden) for cervical cancer, and from 0.6% (Hungary) to 64.5% (the Netherlands) for CRC screening programs. The proportion advised further assessment following a screening test ranged from 0.6% (Chile) to 14.4% (Republic of Korea) for mammography-based screening and from 1.0% (Mozambique) to 2.8% (Bangladesh) for clinical breast examination-based breast cancer screening programs; from 0.5% (Sri Lanka) to 7.8% for cytology (Uruguay); from 2.3% (Kenya) to 78.8% (Ethiopia) for VIA-based cervical screening programs; and from 2.3% (Calvados, France) to 27.2% (Uruguay) for FIT-based CRC screening programs. Regardless of the screening protocol, further assessment participation rates varied substantially across cancer sites and countries, ranging from 39.7% in Morocco to 100% in the Czech Republic, Denmark and Portugal for breast cancer, from 39.0% in Poland to 100% in Hungary for cervical cancer, and from 33.0% in the Republic of Korea to 97.6% in the Czech Republic for CRC screening programs. The high variability in screening test and protocol, test positivity and further assessment compliance was reflected in the precancer and cancer detection rates.

Policy implications

Substantial heterogeneity in screening program performance revealed by the first batch of data from the CanScreen5 project underscores the need for many such programs to do further in-depth analysis of their performance, identify the scope for improvement and take appropriate measures. To implement corrective actions, program managers need to be aware of the implications of the outcome indicators; for example, a low detection rate may indicate poor performance of the screening or diagnostic tests (or both) or low compliance of screen-positives, but may also be due to low prevalence of disease especially in a frequently screened population. The gap in data collection across the screening continuum in both high- and low-resource settings is a concern. Programs need to build robust information systems to be able to capture screening performance data and use the same for quality improvement. Almost all countries worldwide have invested greatly to strengthen disease surveillance mechanisms (including improvement of health information systems) to mitigate the COVID disease pandemic. Cancer screening programs need to leverage these new developments to improve their own performance and quality.