Fig. 4: Biomedical data access needs and risks to privacy.
From: Addressing contemporary threats in anonymised healthcare data using privacy engineering
Patient-level care or research may require granular data for laboratory diagnostics, although bedside care often uses only data in ranges rather than precise values (such as left ventricular ejection fraction). Population research and resource planning could use aggregate data, which introduce different and potentially lower risks. Re-identification risks are elevated by self-disclosure on social media and the growth of consumer diagnostics developed by private corporations, which are often not HIPAA-covered entities.
