Evaluation of the Slovenian Breast Cancer Screening Programme: Years of Life Gained and Avoided Deaths

TL;DR

A study in Slovenia found that breast cancer screening saves lives and adds years to life, helping policymakers understand its benefits.

Contribution

A novel approach was used to estimate years of life gained and deaths avoided in breast cancer screening.

Findings

Invited women had a 4.3 percentage point higher survival probability than non-invited women after 10 years.

Screening resulted in 22 avoided deaths per 100,000 women and 114 years of life gained.

Abstract

Public health interventions should evaluate their effect on the population. Measuring outcomes in organized cancer screenings is not straightforward, since expected reduction in mortality has drawbacks such as long observation time, quality of data, and the relationship between the intervention and the outcome. Researchers are exploring the use of survival, which is available sooner, but has limitations such as lead time bias, length time bias, and over-detection. Slovenian researchers have used a novel approach addressing the known biases comparing the survival between invited and not-invited women and providing estimations of years of life gained and deaths avoided for an easier interpretation for policy-makers. The results show that after 10 years the invited women have a 4.3 percentage points higher survival probability than the not invited, resulting in 22 avoided deaths per 100,000, amounting to a total of 114 years of life gained. The standardized results allow cross-country comparisons.

Background: Most commonly, mortality is used for evaluation of organized breast cancer screening programmes. Survival analyses are also being employed to provide more timely assessment. A novel approach has been introduced for calculating survival using Slovenian data. Methods: Breast cancer cases from the Slovenian Cancer Registry between 2008–2021 in women aged 50 to 72 were included, supplemented by data on the screening history from the screening registry. The comparison was made in two scenarios—invited or not invited to the screening. Survival, years of life gained, and number of avoided deaths over 10 years were calculated and standardized. The comparison was adjusted for confounding covariates, and a biological tumour growth model was used to adjust the survival of the invited group for lead time bias and over-detection. Results: A total of 9392 breast cancer cases were included in the study. Estimated median lead time for screen-detected cases was 1.23 years. The survival probability accounting for lead time bias and covariates of cancers patients invited to the programme was 4.3 percentage points higher than the not invited (0.81 vs. 0.76). Standardized to 100,000 women in the 2-year period, this would result in 22 avoided deaths. On average, the invited cancer patient lived 0.22 years longer, which amounts to a total of 114 years of life gained. Conclusions: The results show the overall benefit of the breast cancer screening programme. The standardized results enable cross-country comparisons. The calculation of years of life gained and avoided deaths can provide additional opportunities for communicating the results to non-expert populations.

Figures