Adaptation of the European Guidelines on Breast Cancer Screening to the Tunisian setting using the GRADE-ADOLOPMENT approach

Article type
Authors
Kahale LA1, Ouertatani H2, Ben Brahem A2, Ben Hamouda M2, Sar Parkinsion Z3, Akl EA1
1American University of Beirut
2L'instance National d' Accréditation en Santé
3European Commission
Abstract
Background: guidelines are the cornerstone of high-quality evidence-based practice. Guidelines could be
1) developed from scratch, (i.e. de novo);
2) adopted from one setting to another as is; or
3) adapted from one setting to another taking into consideration contextual factors.

The development of a de novo guideline is a demanding process in terms of time, human, and financial resources while the adoption from one setting to another as is may be inappropriate due to different contextual factors. Adapting guidelines takes contextual factors in consideration, while cutting required resources.

Objectives: to describe the process used for the adaptation of European Guidelines on Breast Cancer Screening to the Tunisian setting.

Methods: we applied the GRADE-ADOLOPMENT methodology using GRADEpro-GDT software, particularly in relation to:
1) defining guideline scope and topics;
2) identifying the source guideline;
3) determining groups and roles;
4) training the working group and guideline panelists on guideline development and adaptation using GRADE methodology;
5) prioritizing questions and outcomes;
6) searching for local data;
7) reviewing evidence tables prepared and shared by the European Guideline;
8) formulating and grading strength of recommendations.

Results: the Tunisian Guideline Panel considered nine screening questions from the European Guidelines on Breast Cancer Screening and Diagnosis. The panel dropped three questions on tomosynthesis screening, as it is not used in Tunisia as a screening tool. They changed the Guideline perspective from that of a population to that of an individual. For all questions, the panelists prioritized the same outcomes as the European Guideline; however they changed the rating of importance for two outcomes:
1) “all-cause mortality”: from “not important” to “important”;
2) “overdiagnosis”: from “critical” to “important”.

Despite the lack of data from the Tunisian context, the panelists assumed a lower incidence of breast cancer but a higher risk of breast cancer mortality in Tunisia compared to Europe. The panel did not modify the overall certainty of the evidence for any recommendation. However they changed the strength of one recommendation from “conditional against” to “conditional for either” mainly due to very low certainty of evidence, large costs and unclear cost-effectiveness.

Conclusions: this process illustrates both the feasibility of the GRADE-ADOLOPMENT approach and the importance of consideration of local data. It also highlights the value of collaboration with the organization that developed the source guideline.

Patient or healthcare consumer involvement: patient representatives were part of the guideline panel.