American patients of minority groups

Affecté par des incidents

Incident 1721 Rapport
NarxCare’s Risk Score Model Allegedly Lacked Validation and Trained on Data with High Risk of Bias

2020-07-01

NarxCare's algorithm assessing a patient’s overdose risk allegedly did not undergo peer-reviewed validation studies, and considered sensitive data with high risk of biases towards women and Black patients such as experience of sexual abuse and criminality.

Entités associées

Appriss

Incidents impliqués en tant que développeur et déployeur

Incident 172
1 Rapport
NarxCare’s Risk Score Model Allegedly Lacked Validation and Trained on Data with High Risk of Bias

American physicians

Affecté par des incidents

Incident 172
1 Rapport
NarxCare’s Risk Score Model Allegedly Lacked Validation and Trained on Data with High Risk of Bias

American pharmacists

Affecté par des incidents

Incident 172
1 Rapport
NarxCare’s Risk Score Model Allegedly Lacked Validation and Trained on Data with High Risk of Bias

American patients

Affecté par des incidents

Incident 172
1 Rapport
NarxCare’s Risk Score Model Allegedly Lacked Validation and Trained on Data with High Risk of Bias