EQUAL CARE® Certification

Ensure that foundational research includes a gender representation that reflects the prevalence in the population. Companies must provide data demonstrating this balance in their studies, showing that both genders are adequately represented in clinical trials. Deviation up to 25 percentage point will be accepted since exact recruitment can be associated with disproportional efforts in resource and time. In the future we aim to lower this standard.

Companies must: Ensure gender representation in their training data matches the population prevalence as closely as possible. Provide detailed information on the origin and quality of the training data. Maintain transparency in data preprocessing and any transformations applied. Investigate and document potential biases in the data and their impacts on the model.

Companies must: Ensure gender representation in their validation data matches the population prevalence as closely as possible. Provide detailed information on the origin and quality of the validation data. Maintain transparency in data preprocessing and any transformations applied. Investigate and document potential biases in the data and their impacts on the model.

Assess and document the inclusion of gender-specific needs as input variables for the model's decisions. Companies must demonstrate how these needs influence model outcomes.

Validate the model's performance as a whole and separately by gender. Companies need to provide evidence of this validation process and its results.

Ensure the explainability of the hyperparameters used, the training dataset, and the training process. Companies must provide clear documentation of these aspects.

Provide explanations of the model and its processes in a manner understandable to non-experts. Companies should use clear and accessible language in their documentation.

Describe the cost model associated with the application for users, healthcare professionals, and organizations. Companies must provide information, contact or reference.

Provide evidence of any possible side effects or the absence thereof. Companies need to document and present findings related to side effects.

Ensure that the algorithm complies with relevant regulatory requirements and ethical guidelines. Companies must provide documentation of their compliance measures. References to GDPR, HIIPA, MDR, FDA, ISO 13485 and others.

Appropriate documentation and evidence to facilitate the certification process.