# Can GPT-3.5 Generate and Code Discharge Summaries?

arXiv: https://arxiv.org/abs/2401.13512 · HTML: https://mlsanity.com/paper/2401.13512/

## Key facts

- Cited by 13 papers in the map (data through 14 Oct 2026).
- Builds on 24 of its 36 references that are in the map.
- First posted on arXiv 24 Jan 2024.

## Builds on (most cited first, 24 of 24)

- [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://mlsanity.com/paper/1810.04805/) (2018)
- [Training language models to follow instructions with human feedback](https://mlsanity.com/paper/2203.02155/) (2022)
- [LLaMA: Open and Efficient Foundation Language Models](https://mlsanity.com/paper/2302.13971/) (2023)
- [Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks](https://mlsanity.com/paper/2005.11401/) (2020)
- [Distributed Representations of Words and Phrases and their Compositionality](https://mlsanity.com/paper/1310.4546/) (2013)
- [A Survey of Large Language Models](https://mlsanity.com/paper/2303.18223/) (2023)
- [Survey of Hallucination in Natural Language Generation](https://mlsanity.com/paper/2202.03629/) (2022)
- [Large Language Models Encode Clinical Knowledge](https://mlsanity.com/paper/2212.13138/) (2022)
- [Explainable Prediction of Medical Codes from Clinical Text](https://mlsanity.com/paper/1802.05695/) (2018)
- [ICD Coding from Clinical Text Using Multi-Filter Residual Convolutional Neural Network](https://mlsanity.com/paper/1912.00862/) (2019)
- [PLM-ICD: Automatic ICD Coding with Pretrained Language Models](https://mlsanity.com/paper/2207.05289/) (2022)
- [Automated Medical Coding on MIMIC-III and MIMIC-IV: A Critical Review and Replicability Study](https://mlsanity.com/paper/2304.10909/) (2023)
- [Evaluation Measures for Hierarchical Classification: a unified view and novel approaches](https://mlsanity.com/paper/1306.6802/) (2013)
- [Explainable Automated Coding of Clinical Notes using Hierarchical Label-wise Attention Networks and Label Embedding Initialisation](https://mlsanity.com/paper/2010.15728/) (2020)
- [Automated Clinical Coding: What, Why, and Where We Are?](https://mlsanity.com/paper/2203.11092/) (2022)
- [Read, Attend, and Code: Pushing the Limits of Medical Codes Prediction from Clinical Notes by Machines](https://mlsanity.com/paper/2107.10650/) (2021)
- [Mimic-IV-ICD: A new benchmark for eXtreme MultiLabel Classification](https://mlsanity.com/paper/2304.13998/) (2023)
- [DALE: Generative Data Augmentation for Low-Resource Legal NLP](https://mlsanity.com/paper/2310.15799/) (2023)
- [Foresight -- Generative Pretrained Transformer (GPT) for Modelling of Patient Timelines using EHRs](https://mlsanity.com/paper/2212.08072/) (2022)
- [CoPHE: A Count-Preserving Hierarchical Evaluation Metric in Large-Scale Multi-Label Text Classification](https://mlsanity.com/paper/2109.04853/) (2021)
- [HiCu: Leveraging Hierarchy for Curriculum Learning in Automated ICD Coding](https://mlsanity.com/paper/2208.02301/) (2022)
- [BERT for Long Documents: A Case Study of Automated ICD Coding](https://mlsanity.com/paper/2211.02519/) (2022)
- [HieNet: Bidirectional Hierarchy Framework for Automated ICD Coding](https://mlsanity.com/paper/2212.04891/) (2022)
- [An Automatic ICD Coding Network Using Partition-Based Label Attention](https://mlsanity.com/paper/2211.08429/) (2022)

## Built on by (most cited first, 13 of 13)

- [Code Like Humans: A Multi-Agent Solution for Medical Coding](https://mlsanity.com/paper/2509.05378/) (2025)
- [Representation Learning of Structured Data for Medical Foundation Models](https://mlsanity.com/paper/2410.13351/) (2024)
- [MedDCR: Learning to Design Agentic Workflows for Medical Coding](https://mlsanity.com/paper/2511.13361/) (2025)
- [Edinburgh Clinical NLP at SemEval-2024 Task 2: Fine-tune your model unless you have access to GPT-4](https://mlsanity.com/paper/2404.00484/) (2024)
- [Edinburgh Clinical NLP at MEDIQA-CORR 2024: Guiding Large Language Models with Hints](https://mlsanity.com/paper/2405.18028/) (2024)
- [Generating Synthetic Free-text Medical Records with Low Re-identification Risk using Masked Language Modeling](https://mlsanity.com/paper/2409.09831/) (2024)
- [Artificial intelligence to improve clinical coding practice in Scandinavia: a crossover randomized controlled trial](https://mlsanity.com/paper/2410.23725/) (2024)
- [Privacy-Preserving Generation of Clinical Narratives from Medical Terminologies](https://mlsanity.com/paper/2509.10882/) (2025)
- [Don't Pay Attention, PLANT It: Pretraining Attention via Learning-to-Rank](https://mlsanity.com/paper/2410.23066/) (2024)
- [Synthetic Clinical Notes for Rare ICD Codes: A Data-Centric Framework for Long-Tail Medical Coding](https://mlsanity.com/paper/2511.14112/) (2025)
- [Enhancing LLM Medical Coding with Structured External Knowledge](https://mlsanity.com/paper/2605.27377/) (2026)
- [CoLa-ICD: A Knowledge-Enhanced Framework for Long-Tail Automated Medical Coding](https://mlsanity.com/paper/2608.30234/) (2026)
- [Beyond ICD Codes: Semantic IDs for Generative Medical Coding](https://mlsanity.com/paper/2610.19301/) (2026)

## Abstract (arXiv)

Objective: To investigate GPT-3.5 in generating and coding medical documents with ICD-10 codes for data augmentation on low-resources labels. Materials and Methods: Employing GPT-3.5 we generated and coded 9,606 discharge summaries based on lists of ICD-10 code descriptions of patients with infrequent (generation) codes within the MIMIC-IV dataset. Combined with the baseline training set, this formed an augmented training set. Neural coding models were trained on baseline and augmented data and evaluated on a MIMIC-IV test set. We report micro- and macro-F1 scores on the full codeset, generation codes, and their families. Weak Hierarchical Confusion Matrices were employed to determine within-family and outside-of-family coding errors in the latter codesets. The coding performance of GPT-3.5 was evaluated both on prompt-guided self-generated data and real MIMIC-IV data. Clinical professionals evaluated the clinical acceptability of the generated documents. Results: Augmentation slightly hinders the overall performance of the models but improves performance for the generation candidate codes and their families, including one unseen in the baseline training data. Augmented models display lower out-of-family error rates. GPT-3.5 can identify ICD-10 codes by the prompted descriptions, but performs poorly on real data. Evaluators note the correctness of generated concepts while suffering in variety, supporting information, and narrative. Discussion and Conclusion: GPT-3.5 alone is unsuitable for ICD-10 coding. Augmentation positively affects generation code families but mainly benefits codes with existing examples. Augmentation reduces out-of-family errors. Discharge summaries generated by GPT-3.5 state prompted concepts correctly but lack variety, and authenticity in narratives. They are unsuitable for clinical practice.