# Differentially Private De-identification of Dutch Clinical Notes: A Comparative Evaluation

arXiv: https://arxiv.org/abs/2604.21421 · HTML: https://mlsanity.com/paper/2604.21421/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 13 of its 31 references that are in the map.
- First posted on arXiv 23 Apr 2026.

## Builds on (most cited first, 13 of 13)

- [GPT-4 Technical Report](https://mlsanity.com/paper/2303.08774/) (2023)
- [The Llama 3 Herd of Models](https://mlsanity.com/paper/2407.21783/) (2024)
- [Differentially Private Fine-tuning of Language Models](https://mlsanity.com/paper/2110.06500/) (2021)
- [BERTje: A Dutch BERT Model](https://mlsanity.com/paper/1912.09582/) (2019)
- [Synthetic Text Generation with Differential Privacy: A Simple and Practical Recipe](https://mlsanity.com/paper/2210.14348/) (2022)
- [De-identification of Patient Notes with Recurrent Neural Networks](https://mlsanity.com/paper/1606.03475/) (2016)
- [GLiNER: Generalist Model for Named Entity Recognition using Bidirectional Transformer](https://mlsanity.com/paper/2311.08526/) (2023)
- [Leveraging Hierarchical Representations for Preserving Privacy and Utility in Text](https://mlsanity.com/paper/1910.08917/) (2019)
- [Robust Utility-Preserving Text Anonymization Based on Large Language Models](https://mlsanity.com/paper/2407.11770/) (2024)
- [Unlocking the Potential of Large Language Models for Clinical Text Anonymization: A Comparative Study](https://mlsanity.com/paper/2406.00062/) (2024)
- [De-Identification of French Unstructured Clinical Notes for Machine Learning Tasks](https://mlsanity.com/paper/2209.09631/) (2022)
- [An Easy-to-use and Robust Approach for the Differentially Private De-Identification of Clinical Textual Documents](https://mlsanity.com/paper/2211.01147/) (2022)
- [Detection of Adverse Drug Events in Dutch clinical free text documents using Transformer Models: benchmark study](https://mlsanity.com/paper/2507.19396/) (2025)

## Abstract (arXiv)

Protecting patient privacy in clinical narratives is essential for enabling secondary use of healthcare data under regulations such as GDPR and HIPAA. While manual de-identification remains the gold standard, it is costly and slow, motivating the need for automated methods that combine privacy guarantees with high utility. Most automated text de-identification pipelines employed named entity recognition (NER) to identify protected entities for redaction. Although methods based on differential privacy (DP) provide formal privacy guarantees, more recently also large language models (LLMs) are increasingly used for text de-identification in the clinical domain. In this work, we present the first comparative study of DP, NER, and LLMs for Dutch clinical text de-identification. We investigate these methods separately as well as hybrid strategies that apply NER or LLM preprocessing prior to DP, and assess performance in terms of privacy leakage and extrinsic evaluation (entity and relation classification). We show that DP mechanisms alone degrade utility substantially, but combining them with linguistic preprocessing, especially LLM-based redaction, significantly improves the privacy-utility trade-off.