DDReg pharma

DELIVER BETTER DATA TO ENSURE FASTER APPROVAL
DDReg Pharma
Talk To Us
Artificial Intelligence

Enhancing medico-regulatory writing with artificial intelligence

Medico-regulatory writing is an important part of regulatory affairs in the Pharmaceutical and Life Sciences industry. Regulatory standards are becoming more stringent in terms of data requirements as a result of ever-changing regulations. So, it is crucial to incorporate the most updated and precise medical information into regulatory applications for pharmaceutical product approval, and to avoid chances of applications getting rejected. However, sifting through large and unstructured data and transforming it into structured data for various formats is a painstaking task that can be relatively time consuming and can lead to delays in submission.
Today, as industries leverage artificial intelligence (AI) and automation to enhance their capabilities and processes, medico-regulatory writing is also keeping in line with the same trend to streamline processes that would lead to faster outputs. Indeed, the use of AI and machine learning (ML) has been beneficial in drug development to make patient enrollment, medical & clinical data analysis, trial & protocol design etc automated and more efficient. In medical writing, AI can be used to tackle challenges associated with traditional peer review processes, bias, lack of transparency, plagiarism, and bad reporting among others. AI tools with automation could manage and change regulatory documents to ensure they are compliant with the required format [1, 2].

Combining NLP and NLG in Medical Regulatory Writing

Natural Language Processing (NLP) is a form of AI that essentially helps computers in comprehending language of humans; natural language generate (NLG) is when computers generate text from unstructured data. This has proven to be successful in the clinical setting in medicine for detecting early-stage chronic diseases and enhanced prediction of patient outcome [3]. The combination of NLP and NLG in medical writing is one that could revolutionize the medico-regulatory writing landscape.

Automatic extraction of data from various structured & unstructured data sources followed by their comprehensive analysis would help to understand and classify the information in its context. The information is then stored in clusters which marks the end of the NLP stage. The NLG stage takes over from this point where reports are generated from the structured and stored data. Essentially, a repository or database of structured medical information is created which can be re-used as necessary, depending on the need of the different regulatory stakeholders.

Currently, medical writers collect data from a variety of regulatory documents & sources that are usually unstructured which is a relatively time-consuming and laborious task. After creating the initial draft, the document undergoes the initial review process that results in an improved draft. After the final review, the document is finalized and ready for use. There may be additional review processes involved depending on the volume and complexity of the document. The use of NLP and NLG AI tools can streamline the traditional approach in medical writing by reducing multiple reviews; the initial draft is automatically generated followed by automatic review to produce the final draft version of the document. This would in turn accelerate submission processes and lead to rapid market access for pharmaceutical products.  

Challenges in adopting automation & AI

On one hand, AI technologies can significantly improve and transform industries such as Pharmaceutical, Life Sciences, Health etc by automating processes and delivering more personalized, real-world outputs. However, there are some challenges associated with adopting AI technology. From a resource perspective, one key concern is AI technology replacing the jobs of the people that would normally be conducting regulatory tasks and activities. It is important to note that the purpose of AI tools is not to replace resources but more to optimize time and effort that professionals would spend on time-consuming tasks by spending it on other activities & operations that would bring more business value [4].

Inputs drive the outputs. The output of the AI technology will only be as good as the data that is being fed into the system. It is important to ensure that the data sources that are used for medico-regulatory writing are of high quality. Raw data can be associated with errors which is why data processing measures should be in place before NLP and NLG processing can begin. This is particularly important when creating a model that can be applied for medico-writing processes. For example, AI systems are programmed based on a training and testing dataset. If there are errors in these datasets, then it could affect the entire model/AI system [5].
Conclusion

Despite some of the concerns, the use of AI technology in various sectors and industries is proving to be beneficial. Automation of processes and activities means that resources can be utilized more effectively leading to a greater return on investment. NLP and NLG technology can automate and streamline key processes in medical regulatory writing resulting in high quality and faster submissions for organizations. Not only would this contribute to efficient review processes and faster submissions but also greater return on investments for businesses. 

References and Further Reading:

[1] Parisis N. Medical writing in the era of artificial intelligence. Medical Writing. 2019 Dec 1;28:4-9.

[2] Harrer S, Shah P, Antony B, Hu J. Artificial intelligence for clinical trial design. Trends in pharmacological sciences. 2019 Aug 1;40(8):577-91.

[3] Locke S, Bashall A, Al-Adely S, Moore J, Wilson A, Kitchen GB. Natural language processing in medicine: a review. Trends in Anaesthesia and Critical Care. 2021 Jun 1;38:4-9.

[4] Rodriguez-Antolin C. Artificial Intelligence Will Not Replace Health Professionals, but the Proper Use of Artificial Intelligence Will Make Health Professionals Better. Cancer Research. 2020 May 15;80(10):1912-3.

[5] Basu K, Sinha R, Ong A, Basu T. Artificial intelligence: How is it changing medical sciences and its future?. Indian Journal of Dermatology. 2020 Sep;65(5):365.