Is India a New Destination for Medical AI Development?
[Thursday, June 28, 2019] A prominent blog this week speculated why India seems to be becoming an attractive place for developers of Artificial Intelligence (AI) based medical technology. India offers some obvious advantages to developers over other countries in the region, however, one must consider some important lessons from the recent history of the clinical trial industry in the country before hyping the opportunities. Similar trends were seen about 10 years ago when the Indian clinical trial industry was first hyped for cost and logistical benefits and then crushed by accusations of data integrity and regulatory issues. For the last decade, Indian clinical trial industry has struggled to recover. Of the 84,863 active clinical trials listed on the clinicaltrials.gov registry, only 290 clinical trials for Phases 1-3, and about 400 Phase 4 studies are currently listed in India. That is an abysmally small number of clinical studies considering that India is home to one of the most populated, younger, highly educated, and rapidly urbanized populations in the world. There are some key differentiators for AI trials versus clinical trials for new drug, biologic or medical device products. AI technology generally would not be expected to cause adverse events of the kind seen with conventional investigational products reducing the liability for such trials. Also, such trials would likely not require regulatory approval from the Indian regulators under the current laws as the laws strictly apply to conventional prescription products and not to software-based products. Hence, trials can be conducted without much regulatory oversight. With a very large English-speaking younger tech-savvy population, much lower cost of data acquisition, along with easy transfer of data over internet, India could be the biggest destination for developers of AI technology. And it seems both big and small AI medical tech developers are looking at the opportunity. Several recently reported clinical trials in India with AI technologies seem to demonstrate the model could work. But before we hype it up, let us consider the following: India has few CROs that are experienced in developing clinical protocols for software as a medical device. The examples so far have been from Alphabet (the Google parent company) and Microsoft who already have huge network of developers and resources in India, and a few small players with partners. Such companies likely did most if the heavy lifting of planning and defining expectations before coming to India. For a company lacking internal resources or trustworthy partners, it would be challenging to conduct AI clinical trials in India. Second, while India has one of the largest numbers of smart phone users in any country, other smart technologies such a wearable devices are less common. India also lags behind in the use of electronic medical records and reliable data collection protocols. Another challenge is the use of endpoints based on patient reported outcome or other subjective measures. Most such endpoints have not been validated in this population. India is an enigma for clinical research, it offers several attractive features while presenting some challenges on the way, and like everything is life, there is work to be done on both side to make a relationship work. India should be the first on the list of places for testing AI technology for all developers who do their homework before committing. |
|