My opinion piece in Pulse+IT in October last year challenging the assumption that clinical coding is an easy picking for AI garnered a lot of interest. As referenced in that article, a national taskforce established by the Health Information Management Association of Australia (HIMAA) consisting of professional association, government representatives, researchers, public sector, private sector, software vendors, clinical coding professionals and clinical governance experts has been meeting over the past 6 months to develop the Australian Clinical Coding Artificial Intelligence (AI) Adoption Guideline.
AI technologies have the potential to improve clinical documentation integrity, produce autonomous coding, conduct clinical coding auditing, and support the management of health information. In an environment of clinical documentation complexity, variation and volume; clinical classification, funding and reporting complexity; human expertise; legislation and standards compliance; privacy and security considerations; risk management; ethics; safety; and quality and efficiency drivers, navigating the path of AI to produce clinical coded data is not straightforward.
The Australian Clinical Coding Artificial Intelligence (AI) Adoption Guideline is intended to provide guidance to healthcare organisations, the clinical coding workforce, software companies, users of coded data, educators, and government agencies, on a principles-based approach to adopting AI technologies related to the clinical coding process.
The Guideline seeks to inform industry of the opportunities and limitations of adopting AI for clinical coding, while offering a framework for appropriate and effective adoption. By promoting industry relevant considerations for AI adoption, the Guideline aims to promote industry alignment of fit-for-purpose technology-enabled clinical coding with clinical coding and clinical governance frameworks.
The Guideline encompasses the integration of AI into the clinical coding process to create and verify clinical coded data. It is intended to guide the development and implementation of AI systems that support the clinical coding workforce to produce accurate and efficient clinical coding. It addresses the people, process and system elements of adopting AI technologies to the end-to-end clinical coding process as described in the Clinical Coding Practice Framework.
In the Australian context of clinical coding and classification licensing, the Guideline identifies the current use cases for clinical coding automation with artificial intelligence while highlighting current known limitations of AI in clinical coding and acknowledging that AI technologies have and will continue to evolve.
A number of important considerations relating to governance, risk management, privacy and security, ethical and safe use, quality improvement, collaboration and partnership, and human expertise outlined in the Guideline are intended to guide the responsible, ethical, and effective use of AI in the production of clinical coded data.
The following categories of principles are further described in the Guideline:
Governance
Effective organisational governance and leadership guides the selection, implementation, and ongoing management of AI systems in clinical coding.
Effective information and data governance policies and practices exist which enables the use of information and data for the purposes of AI in clinical coding
Risk Management
Risk identification and management ensures AI assisted clinical coding is progressed in a way that minimises harm and loss.
Privacy and Security
Data is secure, protected and only accessed and used for authorised purposes.
Ethical and Safe Use
The ethical and safe use of AI in clinical coding is essential to maintaining trust, fairness, and accuracy in health care operations.
Quality Improvement
Robust quality assurance processes continuously monitor, evaluate, and enhance the performance of AI systems in clinical coding.
Collaboration and Partnership
Stakeholders collaborate to produce AI-assisted clinical coding that meets the needs of all parties.
Human Expertise
The clinical coding workforce provides the necessary oversight, validation, and ethical monitoring to ensure that AI systems are used responsibly and effectively.
The draft Australian Clinical Coding Artificial Intelligence (AI) Adoption Guideline opens for consultation from Monday 24th February to Friday 21st March. Links to the draft Guideline and associated consultation will be available via Pulse+IT, HIMAA website and HIMAA LinkedIn.
The taskforce welcomes feedback on the content of the draft Guideline with the goal of publishing the final version of the Australian Clinical Coding Artificial Intelligence (AI) Adoption Guideline in the coming months.
Over to you: Share your view by Commenting below or going to our Poll:
We asked:
It’s great to see HIMAA take steps into it. Actually, this is not a new topic as I saw the computerised coding presentation at the IHTSDO conference in Sydney 15 years ago. Will the current AI storm make the change? Let’s see.
AI is not perfect it should only be part of a tool kit you still need humans to make the final decisions coding is science and human interpretation of facts based on guidelines