The cost-effectiveness of preselection for non-invasive pre-assessment of women referred to full hospital assessment on the basis of a positive screening examination for breast cancer

OBJECTIVE

Evaluation of the Dutch breast cancer screening programme showed that the number of cancers detected could be increased by increasing the referral rate. However, this would lead to a corresponding increase of false-positives, resulting in an increased workload for GPs and hospital specialists, additional financial costs and increased patient anxiety. Our objective is to study the efficiency of a modified assessment strategy which will counter these negative consequences.



STUDY DESIGN

A randomized clinical trial to establish the cost-effectiveness of preselection for non-invasive
pre-assessment in the screening setting in comparison to usual care where all referred women undergo full hospital assessment.



STUDY POPULATION

Women will be randomized at the time of screening, but informed consent will only be sought from referred women (1.5%).



DATA SOURCES

Data on screening and diagnosis are collected from screening and cancer registries. Quality of life, anxiety and patient satisfaction will be measured by questionnaires (EQ-5D, PCQ-DK33, and self-developed items). Monthly diaries will be used to estimate productivity losses and patient time costs.



INTERVENTION

Preselection using a validated reporting system (BI-RADS) leads to fast pre-assessment, only imaging tools, in the screening setting, which would reassure nearly 40% of the initially referred women.



OUTCOME MEASURES

The main outcome parameter is the negative predictive value of preselection, which is interrelated to secondary outcome parameters such as hospital referral rate, detection rate, false-positive rate and the ultimate outcome survival (estimated through a microsimulation model). Additional outcome parameters are quality of life, anxiety, satisfaction, productivity losses, patient time costs and volumes of care.



SAMPLE SIZE / ANALYSIS

A sample size of 1250 referred women ¡¥definite benign after pre-assessment¡¦ ensures a negative predictive value of preselection of 98% or higher and can be met by including eleven screening units. Analyses will be performed according to the intention-to-treat principle.



ECONOMIC EVALUATION

The cost-effectiveness analysis will estimate the cost per QALY gained and cost per percentage increase in satisfaction with screening, or, if effects differ, the incremental cost per QALY gained as the long run modelling outcome.



TIME SCHEDULE

Preparation (1-6 months), data handling (6-36 months), and reporting (30-36 months).

The hypothesis is that stratification of the referred population by the two referring screening radiologists using a validated design (BI-RADS, screening variant) can lead to a fast track workflow with only imaging tools, which would enable us to release nearly 40% of the initially referred women in a complete safe way.

The client must come once to the assessment centre in her region

The modified assessment strategy -i.e. the intervention- can be summarised as follows.

All women referred for further assessment will be classified in three groups, using the BI-RADS classification system. The type of assessment (pre-assessment or full hospital assessment) will differ according to the BI-RADS classification of the screening examination provided by the referring screening radiologists;

- Women with a BI-RADS code 4 or 5 (suspicious for malignancy) will be referred to full hospital assessment, as in current screening practice, since non-invasive pre-assessment will not be sufficient to reach a decision on diagnosis;

- Women with a BI-RADS code 0 (more information needed) will be referred (within 2 working days) for non-invasive pre-assessment in the screening setting consisting of additional or magnification mammograms and breast ultrasound.

The results of pre-assessment will either increase the suspicion that a malignant lesion is present (upgrade to BI-RADS 3, 4, 5 in which case the woman will be referred for full hospital assessment) or virtually exclude this possibility (downgrade to BI-RADS 1, 2, in which case the woman will be classified as 'definite' benign and go home reassured)