Dr. Ashish Indani

 of this paper, we delved into medical writing; its business use case, the current landscape and challenges surrounding the medical writers while authoring a medical document. The use of AI (Artificial Intelligence) and automation was espoused to streamline the process and reduce repetitive and cumbersome nature of this job. The entire knowledge base, can be summarized into a single document using next-generation technologies of NLP (Natural Language Processing) and NLG (Natural Language Generation), thus saving significant time and effort.

As discussed in the previous paper, summarization is a method where information present in multiple documents is consolidated and presented in a single document. Both summarization techniques, vis-à-vis, extractive and abstractive have their pros and cons, that restricts us from applying them directly to solve the problem.

However, with recent advances in deep learning and artificial intelligence, we can certainly mould this to our benefit. In this part, the method of pure abstraction will be discussed, which can be instrumental in generating new sentences.

Medical Writing Automation (MWA) is a framework that leverages the techniques and algorithms of Natural Language Processing (NLP) and Natural Language Generation (NLG) to generate articles. The AI-generated report includes sections such as introduction, methodology, discussion, study objectives, inclusion-exclusion criteria etc.

MWA provides a pure abstraction method. This method is entirely dependent on AI and NLG. It uses pre-trained and state-of-the-art models to generate new statements. Keywords are collected from individual articles using this methodology. These are terms that can be used to create a storyline of the article. Such keywords are extracted from each article, thereby, resulting in a set of keywords. All keywords are compared, and the ones with the highest priority and frequency of occurrence are selected.

For this task, an AI model that has been pre-trained on large amounts of medical data and conducts text-to-text generation is used. The keywords that are in a sequence are fed into the AI model helping it to construct a sentence based on the input. This sentence, along with the second keyword, is returned to the model that, in turn, generates a new sentence. This procedure is repeated for each keyword, ensuring that the newly created sentence, similar to previous sentences, retains correct context. For this method of generation to perform well, it requires a large amount of well-structured data.

This MWA system comprising of summarization and abstraction techniques can be used as an API or can even be integrated with other service or application as and when required.

As discussed, summarization maintains the proper context of the text generated and is grammatically correct, however, abstraction on the other hand generates new sentences from scratch. But the quality of the sentence and its relevance generated by the method of abstraction depends on the training of the model.

Abstractive summarization techniques have a small edge over extractive summarization; however both are not up to the mark in generating accurate articles. The primary reason for this is the restriction due to current technology on training of deep learning models to specific contexts and subsequent sentence generation. With advancements in technologies, including concepts such as Transfer Learning and Ensemble Learning, we can build a generic model to understand the context, thus enabling generation of summaries mapped to the context. Leveraging pure abstraction technique won’t suffice in case of medical writing as the domain is limitless. And training an AI model on each and every drug, disease and scenario is next to impossible, owing to the limitations of current technology. A new hybrid approach needs to be developed that will combine the advantages of both extractive and abstractive summarization techniques and generate the summary text.

The process of pure abstraction generates new sentences. The AI model used to construct the sentences must be trained on a large quantum of data for the sentences to be grammatically correct. Only if the model is trained on medical data will the created phrases have the right context. In pure abstraction, just the keywords are extracted from articles and sentences are generated basis the data on which the AI model has been trained. In Figure 3, below, it is evident that keywords are extracted from four articles with the PMCIDs mentioned and then fed into the AI model to generate the text.

The quality of the text generated by extractive summarization mimics the quality of sentences penned by humans, as it compares the sentences written by them, ranks them and returns the top-rated sentences as the summary. However, in abstractive generation the quality of the sentence requires a lot of improvement that can be done either by increasing the volume of training material or narrowing down the system to focus only on generating certain topics. This will make it easier for the model to train on limited number of topics.

The summarization techniques can’t replace the human medical writers entirely. They can only give the writers a foundation of where to look and provide them with a certain degree of assistance by providing a summary of the published literatures pertaining to the searched topic while considering the timeline of search. Abstractive generation is usually a trade-off between plagiarism score and quality of the text generated over the extractive generation.

Although extractive summarization produces good results, it cannot be used directly as the final writeup due to levels of plagiarism. So, it can only be used to assist the medical writer by enabling them with a base and instructions as to where and how to begin writing. The users can run the system numerous times on different words to obtain the extractive summaries, that will provide the writer with an overview of what has been written and documented about the search terms with respect to time. This will allow the writer to save a significant amount of time.

The method of generating text by pure abstraction generates unique sentences, but if the model isn’t trained on the context related to the text being generated, it's possible that the sentences generated by this method doesn't always make sense or have a proper context. This is entirely dependent on the sort of data utilised to train the AI model. If the AI model is trained on common English literature, it will be capable of generating accurate grammatical sentences, but will lack the ability to construct medically (domain specific) meaningful sentences.

Also, if the model is trained on data related to medically relevant terms such as diabetes, it will correctly frame sentences based on keywords related to diabetes. However, if the keywords are related to other conditions, the model will not be able to produce the correct output and the generated sentences will be prejudiced and directed towards diabetes. This is the most difficult aspect of our strategy. To overcome this barrier and create an AI model capable of creating statements on any medical condition, a large amount of diversified training data is required, in addition to a constant updation of the training data to ensure that no topic is missed.

Text mathematics approach can be evaluated. This is a hybrid approach to combine both summarization and abstractive methods of generation. Text mathematics allows the text to be analysed using NLP techniques and high priority keywords are extracted from the text. These keywords are used to prepare a story line of the article to be generated. Using n-gram techniques an AI model can be trained to obtain the correct sequence of the keywords extracted. Keywords give out important information that is to be conveyed and once they are in correct sequence, the only remaining task is to complete the sentence by adding fillers and grammar.

The story line prepared will be good enough to convey the idea or information correctly. These keywords arranged in a proper sequence will then be fed into the AI model trained on medical data to generate sentences. This approach is deemed to be better than the approach of just giving in the keywords one by one without any sequence as it’ll help to generate sentences that are more contextually related to each other.

This is the text mathematics approach in which NLP techniques are used to extract keywords from text, generate a sequence of these keywords and then use NLG to generate new sentences. This strategy, once implemented, will produce far superior outcomes than the abstractive or summary methods alone.

As a result, using today's technology, it's best to take a hybrid strategy and generate sections of articles utilising both extractive and abstraction summarization strategies. Although the generated papers will require some human review, this technology will help medical writers be more productive and increase their efficiency many folds. The generated text can be used by the writers to obtain a summary of research work has been done in the searched context and provide the writers with a starting point from where they can build their new document in a seamless manner.

, Former Head Research and Innovation; all with

Articles

Following a discussion around summarization in Part 1, Part 2 discusses the method of pure abstraction in light of recent advances in deep learning and AI.

AI/ML Approaches to Assisted Medical Writing—Part 2

Medical writing is a highly specialized field that encompasses the art and science of content writing, as well as of clinical research. It entails the creation of well-structured scientific materials, such as clinical research papers, healthcare website content, periodicals, journals etc. The audience for these documents range from a layman to a highly skilled medical expert. Medical writers must possess a deep understanding of the subject and be able to comprehend medical concepts and terminology. In addition, they should be familiar with applicable rules on the style and content of certain papers and possess excellent writing abilities. The art of writing documentation for regulatory authorities requesting approval for devices, medicines, and biologics is referred to as regulatory medical writing. This contains information from clinical trials, regulatory submission materials, and post-approval documents, among other things.

 On the other hand, writing documents on devices, biologics, and medicine for educational purposes is known as educational medical writing. In recent times, the demand for medical writers has exponentially increased as the number of drugs discovery and clinical studies undertaken have ramped up.

The scope of our experiment is limited to regulatory medical writing involving updates or amendments to various clinical trial documents, vis-à-vis, Protocol, Investigation Brochure (IB), Informed Consent Forms (ICF), Common Technical Document (CTD) and Clinical Study Report (CSR).

Challenges and need for assisted medical writing

Medical writing involves highly skilled and trained professionals. Every single aspect mentioned in a document must be reviewed by multiple peers to check for authenticity. Authoring a single paper and double-checking its contents therefore becomes an extremely cumbersome and time-consuming process.

Medical writing necessitates the writers to read and understand numerous research papers, documents, and journals to arrive at a conclusion. A machine can scan thousands of articles in matter of seconds, compared to humans who would require days. Disparity and biases in the human mind too exist, that invariably finds its way in the article being produced.

AI has made significant advances in the field of text mining, processing, and generation. AI engines powering these services can comprehend the context and recommend text accordingly. The technology also proves instrumental in intuitive content writing.

A computer if trained properly exhibits no bias. It provides its predictions and suggestions basis its training. With advances in computer technology and rise of AI in the field of NLP and NLG (Natural Language Generation), medical writers can leverage these innovations to assist them in creating medical documentation.

Nowadays, AI is being frequently used in Pharmacovigilance for extracting information from ICSR (Individual Case Study Reports) and further processing of data.

 The machines in these cases are performing the preliminary and repetitive tasks that are time consuming and thus, long working hours of a skilled person is saved allowing him/her to focus on critical activities. There have been attempts to identify the causal relationship between drugs and adverse events by leveraging AI. The key focus of our experiment is to assist the regulatory medical writers and provide them with a launchpad for seamless authoring of regulatory documents. The AI model would subsequently be trained to generate sections of a protocol and help in validating the efforts taken for using AI in the field of Medical Writing.

A medical writer needs to enter medical terms for which he or she wishes to generate the section of a new document These keywords are combined to form a search query, which is then run through PMC (PubMed Central Repository). This search returns several articles/papers that are related to the search terms, post which relevant articles are chosen, and data is extracted from them in a section-wise manner. The search result is generally used as a reference point to build a new regulatory document.

The sections considered in this experiment are introduction, methodology, and discussion pertaining to a protocol. The extracted data is cleaned and pre-processed before NLG and NLP algorithms are applied to it to generate the required section.

Medical data required to generate text in this experiment is sourced from the PMC. Data from the full text articles is extracted in the form of a Data Table, with one article/paper forming a row. As seen in Figure 1 below, the research paper/article sourced from PMC is transformed into a structured format which is processed further.

We have proposed a framework that makes use of the techniques and algorithms of NLP and NLG to generate the articles. The proposed framework for Medical Writing Automation (MWA) uses AI and NLG to generate a summary of various research papers and articles published. The AI-generated reports include sections such as Introduction, Methodology, Discussion, alongside a few other sections such as Study Objectives and Inclusion-Exclusion criteria.

The framework leverages words as input from the user, e.g., Diabetes, Malaria, India etc. and makes use of PMC as its database. To search for papers on PubMed, a precise search phrase must be used, to receive optimized results and avoid unrelated items from appearing.

The model makes a proper MeSH (Medical Subject Headings) query

 and provisions a search on the PMC. The search returns a list of Articles along with their identifiers. These articles form the base that needs to be generated. Several approaches are used to collect data from these full-text articles. This data is stored in a DataFrame (spreadsheet) with columns such as article id, title of the article, introduction portion, methodology section, and discussion section.

The raw data obtained after extraction from the articles can contain HTML elements, unnecessary punctuations, and other unicode characters. To address this, data cleaning and pre-processing follows. The data comprising of date, time, weights, volumes etc is converted into standard units. For instance, if the article has a statement that says 48 hours, the time frame is converted into days.

In an MWA system, user has the options to select the type of text generation, between summarization or pure abstraction. NLP techniques for summarization and generation used here are Extractive and Abstractive summarization.

The approach for extractive summarization involves taking the complete text as input and breaking it down into individual sentences. These sentences are then sorted as per their significance, that is determined using the cosine similarity and textrank algorithms.

 Once the sentences have been organized per importance, user can select the top ranked sentences and delete the remaining ones. When a user selects Extractive summarization as the method for generating text in MWA, a summary of each article in the DataFrame is generated first (in the background, the Introduction, Methodology, and Discussion sections of each article are individually summarised), post which, once the summaries of all relevant articles are obtained, a summary of the entire data, i.e., a summary of all the summaries, is generated. This cascading of summaries ensures that no information is left out of, and that each sentence from the article is equally significant.

Except for the last section, abstractive summarization operates in the same way as extractive summarization. Individual sentences are prioritised according to their value in abstractive summarization, just as they do in extractive summarization. The only difference being, abstractive summarization approaches use AI and machine learning to understand the relationships between words and sentences, and is trained to compare two words and phrases and calculate their similarity.

Primary data analysis and defending research hypothesis

The data extracted from the articles of PMC is in English and contains many fillers and stop words that are necessary, yet insignificant to complete a sentence. Such words can be removed while training the AI model for abstractive summarization. There will also be few frequent words occurring in the sentence, which can be known by the word count or word cloud approach. If the frequency of the word is high, it may alter the model performance in Abstractive generation, thus, it is better to skip these words too.

In summarization techniques, the individual sentences are considered as a single entity. The correlation of words in the sentence calculated by cosine similarity is critical and if there are similar kind of sentences present in the paragraph to be summarized, the summarization algorithm will find it difficult to rank the sentences. For the summarization techniques to work properly, the sentences should be properly separated out. Long tailed sentences have lower accuracy rates in proper summarization. Hence, while processing the data, sentences should be split properly.

Following are the results of using the search phrases ‘Diabetes Mellitus India' with a view to see how the system fared. Extractive Summarization and Abstractive Summarization were used to create the methodology section

In the case where Extractive Summarization was chosen as the mode of generation, the articles selected from the PMC database were with the PMCIDs of 7328526, 7325532, 7216981, 5210443, 5763039, 6357479, 7667845, 5240082, 6676834, 6755825, 5999405, 5471800, 5884366 and 6227383. Summary from each article was generated followed by a combined summary of all the articles.

The generated summary was, ‘the theoretical minimum risk exposure level for BMI at age 7300 days or more was estimated to range from 20 to 25 kg/m2. Estimates of deaths, YLLs, YLDs, and DALYs attributable to each risk factor for diabetes were produced by location, age, sex, and year. For the secondary analysis, results from interaction models exploring the change in clinical parameters over time were examined and compared with new and known diagnoses of hypertension.’

For Abstractive Summarization the articles selected by the system from PMC database were with the PMCIDs of 7328526 and 7325532. The summary generated was, ‘Inclusion/exclusion criteria Adults with T1DM or T2DM who consulted their physicians during the 14 day-long recruitment period were included in the study. In a survey on 307 diabetes patients in Singapore, only 30.6% of patients were found to be vaccinated with influenza vaccine.’

Performing plagiarism check would reveal that the summary generated by the extractive summarization approach would result in a very high plagiarism score. The reason for high plagiarism is justified by the selection of top correlated sentences and producing the summary without changing the text whereas abstractive summarization has a small edge over extractive summarization in that it makes slight alterations to the words and phrases by replacing them with synonyms or matching terms.

Abstractive summarization techniques score over extractive summarization; however both are not the best approaches in generating accurate articles. The current limitation with respect to the deep learning model is that it generates sentences basis the specific context it has been trained for. With the advancements in technologies like Transfer Learning and Ensemble Learning, a generic model can be built to understand the context and generate summary. In addition, there is a third method of generating text besides summarization, that is referred to as Pure Abstraction method. We shall discuss this method in the second part of our article. Stay tuned!

, Former Head Research and Innovation; all with 

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3149406/

https://www.researchgate.net/publication/341233707_Teaching_Pharmacovigilance_to_Machines_Provided_by_Tata_Consultancy_Services

https://web.eecs.umich.edu/~mihalcea/papers/mihalcea.emnlp04.pdf

Experimenting with different methods using AI.

AI/ML Approaches to Assisted Medical Writing—Part 1

In order to achieve enhanced patient safety, life science industry is highly regulated, adding complexity to achieve product compliance.

 Life sciences regulators are responsible for ensuring that food, human and veterinary drugs, medical devices, cosmetics, and other pharmaceutical products meet the applicable regulations.

 Advancement in technology and scientific research has increased the pace at which regulations change. In addition to regional regulatory requirements, pharmaceutical organizations need to comply with international regulations too.

The first challenge in the changing regulatory landscape is the rapid increase in the number of drugs under observation as well as the ever-evolving standards and guidelines by regulatory agencies.

 Consequently, the amount of data that needs to be searched and analyzed has also increased exponentially (see Figure 1 below). Life sciences industry regulations are complex and the amount of data that needs to be analyzed is huge, thereby necessitating humongous manual efforts and time to maintain compliance of the product.

 Information sources for regulatory intelligence

The second challenge is to interpret the information and assess the impact of changing regulations on the project, organization, and other stakeholders. Post collating changes in regulations, product assessment needs to be followed to maintain its compliance against the new regulation. However, the entire procedure to filter regulatory communication for products and process this communication is time-consuming. While it is of utmost importance for individuals and organizations to be informed and act according to these regulations, it can be challenging to do so since there are multiple regulations spanning across different geographies with varying frequency of regulatory updates.

The third challenge is the ability to respond quickly and efficiently to changes in regulatory standards and to support the pace of rapid development. If any communication is missed; impact assessment is delayed impacting compliance of the product with revised or new regulations.

The use of Artificial Intelligence (AI) in the Lifesciences space has tremendous potential and impact with analyzing complex data sets to enhance drug development. AI can handle the quantum increase in data for regulatory domain by processing tedious tasks and increasing the speed and accuracy of certain regulatory functions such as regulatory information processing, generation of impact analysis reports, gathering insights on current trends, etc. One possible solution to this problem is to utilize an intelligent chatbot, that leverages the concept of conversational AI.

A chatbot is a software that utilizes Artificial Intelligence (AI) to simulate human-like conversation through text or speech. Many chatbots have been created and deployed across a wide range of customer-facing domains such as e-commerce, healthcare, education, and retail. In our case, the chatbot will extract data from regulatory websites and provide information and updates tailored to a customer’s request. Using a chatbot provides a user-friendly and efficient platform for customers to interact with and receive responses to their specific requests without having to search through the mass information available online.

With the use of smartphones and digital devices, user preference is shifting towards personalized experiences. Conversational AI is the subbranch of AI that deals with empowering machines to understand, analyze, and process human language.

 As explained in Figure 2 below, chatbots use large datasets, natural language processing, and Machine Learning to understand the context and imitate text and speech responses. Conversational AI assistants (Chatbot) offer human-like conversations and a better experience for users.

A conversational AI can communicate and empathize with users in a compelling way to assist in desired tasks. Chatbots have moved from a traditional rule-based approach like answering FAQs to more engaging and context-based reasoning.

 They offer various features enlisted below:

—Chatbots can easily connect to various systems and retrieve information in seconds that saves a lot of time for users. This boosts efficiency and reduces the burden of work

—One of the key features is its ease of access, that simulates a human-like conversation with text or speech format

—To fulfill users' dynamic needs, chatbots can cater to queries through real-time access to various databases

—Each industry carries its own unique terminologies. Chatbots can be easily trained to understand this contextual knowledge and provide accurate responses

—Based on industry needs chatbots are easily scalable to address varying workloads

There are various chatbot frameworks available such as Dialog flow.

 Microsoft Bot framework, RASA opensource etc. that can be effortlessly customized and configured to solve specific business problems.

How chatbots can help in regulatory domain?

In regulatory space, a regulatory professional performs various tasks such as maintaining files and data of various events for audit purposes and future reference, recommending strategies to achieve necessary compliance in the company, reporting compliance status, etc. A chatbot will help the regulatory professional to expedite the process of intelligence gathering by using Nextgen technologies such as AI, graph database etc.

The power of conversational AI can be utilized for higher productivity and improved user experience. Chatbots can handle significantly higher requests than humans and process complex information to provide results quickly. It can also work 24/7 and lower query resolution time drastically. In addition, chatbots requires lower operational cost as usage charges will be based upon the utilization of services and provide higher productivity.

Most regulatory professionals work with various products and standards across geographies for data mining, cleaning, and insights collection.

 Chatbots can easily fetch data for the regulatory updates from official regulatory bodies such as USA’s Food and Drug Administration (FDA) or EMA. Depending on the availability, chat bot either uses the Open FDA API, or extracts data by web scraping the official website. The retrieved data accumulates into datastore. Often regulatory professions juggle between various data sources to get the right information. Chatbot, on the contrary acts as an easy to use and reliable medium to reduces manual work.

The process flow for chatbot-based regulatory intelligence is shown in Figure 3 below. After a user input is received, the Natural Language Understanding (NLU) module identifies the intent/entities and extracts it from the user message. Entities are saved into slots that act as a memory for the chatbot.

Until all slots are filled for a particular intent, follow-up questions are asked to allocate the slots. When all slots are acquired, the defined action is executed to provide response to user.

When compared to the traditional and archaic regulatory intelligence process, chatbots offers multiple benefits.

 Traditional vs AI driven approach for regulatory intelligence

The regulatory chatbot can bridge the gap between users who seek information and vast amounts of data contained across the world in different regulatory websites. It is not only able to efficiently respond to user queries by searching through regulatory information, but is able to accomplish this in a user-friendly, easy to understand manner. Suffice to say, this offers a key to increasing its usability and versatility across the life sciences domain.

While the chatbot can classify and recognize the specific user query from the input to return updated URLs, its functionality can be enhanced further. Instead of returning the relevant links when, for example, a user searches for ‘safety communications about Infusion pumps in 2020’, the chatbot could make things easier by responding with a summary of information from those links. This would increase convenience for the user multifold since he/she would not have to individually check each website link. They would be able to extract the relevant information by simply reading the chatbot’s summary of each link and can then choose to visit the website for further information if needed.

The usage of chatbot is not relegated just for insight collection but can be enhanced further by enabling access to various cross-functional teams within the organization. This will decrease the mundane, operational jobs by accessing internal systems of the organization.

 As result, surveillance efforts will be reduced, and information will be accessed quickly.

Another area for progress would be to increase the usage of the regulatory body’s API. This would allow a larger variety of questions to be asked by the user and more detailed information to be provided. The usage of API instead of web scraping would also increase code reusability and efficiency. The regulatory chatbot helps in achieving clinical functions such as stacking regulatory changes and creating relevant SOPs for regulatory professionals, identifying recurring patterns in data to improve the process etc. resulting in increased efficiency and productivity.

, Data Scientist, Research and Innovation, 

, Former Head Research and Innovation, 

, Former Data Scientist, Research and Innovation; all with 

https://www.productlifegroup.com/services/regulatory-intelligence/?msclkid=099684dcaf2111ec9c510499e0ab3545

https://www.pharma-iq.com/business-development/articles/regulatory-intelligence-how-ai-will-change-regulatory-operations?msclkid=09966582af2111ecabee610c314e5124

https://www.raps.org/news-and-articles/news-articles/2019/2/regulatory-intelligence-risk-management-and-drug?msclkid=09963fa1af2111ecb8c16383eb3860fd

https://www.raps.org/news-and-articles/news-articles/2019/1/managing-regulatory-intelligence-for-medical-devic

https://www.raps.org/news-and-articles/news-articles/2019/1/proactive-regulatory-intelligence-communication

https://www.linkedin.com/pulse/how-conversational-ai-improving-customer-experience-yosef#:~:text=%20The%20Technologies%20Used%20In%20Conversational%20AI%20,enormous%20volumes%20of%20data.%20Conversational%20AI...%20More%20

https://www.tcs.com/content/dam/tcs/pdf/Industries/Banking%20and%20Financial%20Services/Charting%20the%20Future%20of%20the%20Wealth%20Management%20Industry.pdf

https://medium.com/conversationalai/the-future-of-conversational-ai-16626a81f071

https://blog.rasa.com/integrating-rasa-with-knowledge-bases/

https://www.ijert.org/research/an-analytical-study-and-review-of-open-source-chatbot-framework-rasa-IJERTV9IS060723.pdf

https://medium.com/coinmonks/rasa-python-weather-chatbot-51fc218d346d

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6616181/

Chatbots can benefit regulatory landscape in light of evolving standards and guidelines.

The Future of Regulatory Intelligence With Conversational AI

Documentation is instrumental in proving the creditability of clinical trials. The documents need to be complete and accurate in order to ensure compliance with ethics, GCP, and applicable regulations. A database used for documentation occupies the center stage of the clinical research.

 The emerging concepts of clinical research transformation, such as decentralized clinical trials, patient-centric clinical trials and community-based clinical trials, are rapidly gaining popularity.

 This scientific perspective will welcome another major change post the paper to electronic transformation, predominantly characterized by the direct involvement of participants instead of the classical sponsor-investigator approach. The clinical research will presumably be an activity of mass, rather than an activity led by the medical personnel. As a restructuring effort, any major pharmaceutical companies will require to adopt various secondary changes in the upcoming years. These changes will range from the inclusion of advance data capturing technologies, like the Internet of Things (IoT), to the management of operational documents in the clinical research.

The trial master file (TMF) is a repository of all essential documents related to the management and conduct of clinical trials right from their inception to the completion of archival period, and is critical to any clinical trial management, essentially for ensuring compliance with applicable regulations.

 In the era of electronic processing-led transformation, TMF has transformed data from paper documents to an electronically managed database. eTMF (electronic trial master file) can handle multiple stakeholders from different geographies in real-time.(Figure 1)

A lean and robust system for the management of clinical trial documents is required for maintaining the momentum of a clinical research. Moreover, stringent regulations demand clean, complete, and diligent documentation of the clinical trial data that is meant to be audit and inspection ready at all times. Auditors are required to tally the data across scanned documents and their digital copies for validating the legitimacy and correctness of the documents. Clinical trials consist of documents being updated at different geographical locations by different staff members for meeting regulatory requirements. With the advancement in clinical research, constant expansion in the scope of study, and the complexity of studies, document collection will prove to be a challenge while accessing documents during inspection and synchronization.

When clinical trials involve multiple stakeholders from different organizational sites and geographies, revising documents using diverse software and keeping track of rapid updates in documents are not possible with the current versioning systems used in eTMF. The lack of stepwise versioning in a document creates conflicting versions of the document, leading to the disarray of version control and document parity mismatch in eTMF.

Any lack of oversight affects record-keeping, thus, ensuing costly reiterations, inaccuracy, and even risks to safety of patients.

 Although eTMF promises a secure archival of clinical records, it also encompasses severe complexities of being a cumbersome process.

The eTMF lacks crucial mechanisms that would enable easy identification of revisions made to the content of a document.

 Moreover, it allows files to be downloaded on local computers, which may be susceptible to possible fraudulent activities with a limited scope to identify the origin of these data breaches.

Additionally, searchability in the dashboard of eTMF is unstandardized and not user-friendly. The board search function and the use of non-uniform naming conventions by multiple stakeholders make it difficult to perform a quick search and locate the required files.

Inaccuracies and missing data is unavoidable. The massive amount of source documents generated by various sources such as lab, sponsors, sites, vendors etc. tend to introduce scalability problems leading to unreliable access management, thereby leading to a considerable wastage of time and effort. Document storage should provide reliable ‘role-based’ access control and prevent unauthorized access or modification of the trial data. Document revisions should be made with consensus among all the stakeholders involved during the entire trial period.

A clinical trial involving researchers from sites across diverse locations requires data to be cleaned and checked for inaccuracies before synchronizing it with other site data. Missing any of the requirement should alert the data operator at the corresponding site itself. Continuous trial data, like sensor data, should be automatically captured during the auditing phase.

Dispersed clinical trials lack the element of trust and transparency because of the enormous amount of data being generated and the documents shared with sites. The clinical trial volumes are increasing at a rapid pace with a projected annual growth of 5.7% until 2028.

 This growth is expected to present extra challenges for data collection and data preservation.

 Keeping this in mind, it is foreseen that eTMF will not suffice the need of a database for the fair and robust documentation of clinical trials. However, there does exist a technology which can be a one-stop solution to all these requirements—Blockchain.

Blockchain is a transactional ledger of a growing list of records called blocks. The blocks are chronologically ordered and linked cryptographically to their previous blocks.

 Each block contains hash of the previous block, timestamp, and the transaction data. Addition or alteration of the chain requires consensus from majority of sharing parties called nodes. Like other databases, blockchain is basically governed within an organization by a committee. This committee oversees the blockchain governance rules placed within an organization and regulates the entire blockchain system.

 Some of the salient features of a blockchain system are

Decentralized—The participating nodes are dispersed at multiple locations and connected only through a networking protocol, and yet their collective computational power and storage define the entire blockchain.

Immutable—Requirement of a huge computation power makes it practically impossible to alter blocks once they are added to the blockchain

Version control—Version control of all essential documents and any transactional activity is easily managed and disseminated in real-time. This activity is visible to all stakeholders. 

Access control—Blockchain provides different levels of access control to ensure that sensitive data is not accessed or modified by an unauthorized user.

Transparency—The documents evolve in the blockchain i.e., they are created, processed, and finally archived in the tool itself. This assures trust and transparency amongst various stakeholders pertaining to the data shared on blockchain.

Collaboration—Organizations can easily collaborate with one another through gossip protocol (a tool in blockchain that enables communication with stakeholders through computers).

Public blockchains, private blockchains, consortium blockchains and hybrid blockchains are the four basic forms of a blockchain network. Each of these platforms has its own set of advantages, disadvantages and ideal applications. The implementation team chooses the blockchain type depending on a variety of factors such as process requirements, size, usage and user base. Blockchain is the foundation that can be leveraged using platforms like Ethereum, Hyperledger Fabric, R3 Corda etc. In addition, blockchain has two kinds of licenses i.e., open-source and proprietary. While in the proprietary license, case procurement depends entirely on the service provider, the open-source licensed blockchain doesn’t require procuring. A few eTMF vendors do support blockchain but as far as the established vendors are concerned it is still a long-drawn process for them to move eTMF to a blockchain based solution.

Thus, this technology embraces vital features of healthcare and allied industries for successful documentation that involves multiple stakeholders where documentation is a crucial entity.

How will blockchain help in the digitization of eTMF?

Use of a blockchain network allows distributed storage of the eTMF data on blockchain in a more reliable and secured manner. The chronological ordering in blockchain largely prevents a 

 Blockchain infuses confidence among stakeholders for genuineness and sanctity of the entire clinical trial data stored in an eTMF-blockchain flow.

 It prevents analytical remodeling of the document and ensures document traceability to provide a robust and non-hackable automation tool with shareable parameters.

Blockchain technology enables the following features to enhance the eTMF functionality:

Secure and restricted access—Access management becomes highly secure with different levels of restrictions on users. Furthermore, automation via chain codes help in identifying the flow of data and ensuring that the stakeholders adhere to GCP (good clinical practice) guidelines.

Decentralized network—The decentralization feature of the tool enables a sturdy network to ensure that the stakeholders are well-connected. For e.g., if there is a failure at one node, the data can be fetched instantly from other nodes with maximum guaranteed uptime.

Trust and non-repudiation—Since blockchain ensures the controlled access and security of data, pharmaceutical companies would have a higher level of trust on it. The use of blockchain will ensure that the pharmaceutical companies, investigators, patients, Institutional Review Board (IRB) / Independent Ethics Committee (IEC) and regulatory bodies have absolute confidence in the data, and that the informed consent is thorough, accurate and complete.

Version management—By managing the versions of documents on the blockchain, compliance to guidelines and regulations is ensured, and a standard is established. Each user/organization will view exactly the same data at any given point of time without any conflict unlike the traditional process of eTMF.(

 Users/organizations can even view the version history for auditing purposes.

Figure 4 eTMF using Permissioned Blockchain

The eTMF-blockchain technology offers a myriad of application benefits. These benefits predominately cater to the enhancement of efficiencies in terms of data quality and integrity while ensuring compliance to ethical standards. This will allow continuous sharing of clinical data between trial entities making it more reliable, credible, and transparent.

 This will also pave the way toward exploring a myriad of new application-based, future-proof solutions, such as:

Bifurcating channels for public and private data sharing.

Adding options for the provision of consensual access to regulatory bodies.

Eradicating paper as well as traditional eTMF systems by replacing them with e-forms.

Creating artificial intelligence- enabled tools to investigate data correctness and the execution of clinical trials.

is the Head of Life Sciences ADD Research & Innovation, and ADD A&I;

is a Developer; all of TCS ADD Platforms. 

is the Former Head of Research and Innovation for TCS ADD Platforms. 

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3121265/

https://www.pharmavoice.com/contributed-article/modern-technologies-partnerships-enabling-next-generation-patient-centric-research/

https://informaconnect.com/internet-of-things-clinical-trials-challenges-opportunities/

https://www.ema.europa.eu/en/documents/scientific-guideline/guideline-content-management-archiving-clinicaltrial-master-file-paper/electronic_en.pdf

https://www.ema.europa.eu/en/documents/scientific-guideline/reflection-paper-good-clinical-practicecompliance-relation-trial-master-files-paper/electronicmanagement-audit-inspection-clinical-trials_en.pdf

https://research.vu.nl/ws/portalfiles/portal/106914288/Triall_Whitepaper_v1.1_Dec_2019.pdf

https://www.gminsights.com/industry-analysis/contract-research-organization

cro-market?utm_source=globenewswire.com&utm_medium=referral&utm_campaign=Paid_globenewswire

https://www.grandviewresearch.com/industry-analysis/eclinical-solutions-market

https://www.triall.io/wp-content/uploads/2021/04/Triall-Whitepaper-v2.1.pdf

https://unthsc-ir.tdl.org/handle/20.500.12503/27036

https://hackernoon.com/6-medical-blockchain-solutions-for-clinical-trials-684s34nz

https://scirp.org/reference/referencespapers.aspx?referenceid=3044024

https://arxiv.org/ftp/arxiv/papers/1906/1906.11078.pdf

https://www.tandfonline.com/doi/full/10.1080/10580530.2020.1720046

https://www.blockchain-council.org/blockchain/data-immutability-works-blockchain/

https://jme.bmj.com/content/medethics/47/4/244.full.pdf

https://openblockchain.readthedocs.io/en/latest/tech/application-ACL/

https://medium.com/codechain/byzantine-failure-why-blockchain-development-is-difficult-1d2da8de9f03

https://medium.com/gigtricks/how-hyperledger-reinforce-trust-transparency-b216efa14f8e

https://www.clinicalprofessionals.co.uk/news/blockchain-and-its-implications-for-clinical-research/

eTMF-blockchain technology offers a myriad of application benefits to data quality and integrity while ensuring compliance to ethical standards. 

Dr. Ashish Indani

Articles

AI/ML Approaches to Assisted Medical Writing—Part 2

AI/ML Approaches to Assisted Medical Writing—Part 1

The Future of Regulatory Intelligence With Conversational AI

Future of Clinical Trial Documentation Management: eTMF Integrated with Blockchain