Today, many doctors find themselves drowning in a sea of data while spending more time on clicks and taps in electronic medical records than on face-to-face diagnosis and treatment of their patients.
This is a problem that I believe threatens to worsen outcomes and high costs of healthcare in the U.S. (the U.S. outranks most comparable countries in metrics such as disease burden, medical errors and waiting time to see a practitioner while having the highest per capita healthcare spending).
Artificial intelligence (AI) and machine learning (ML) technologies are helping transform this deluge of healthcare data into insights that can help hospitals and doctors become more efficient, accurate and accessible. I’ve seen this firsthand through my own company, which uses AI to analyze clinical notes. However, medicine is extremely complicated, and in order for AI and ML technologists to understand the problems clearly and ultimately develop effective solutions, it is critical that feedback from doctors be incorporated into the product development process from the very beginning.
For startups developing AI- and ML-based solutions in healthcare, having at least one doctor who works closely with your team can be extremely valuable for guiding your product development in the right direction. But as in all professions, there are a wide variety of skill sets, backgrounds and biases among physicians. So how can you determine which doctors will make the best advisors and are most likely to lead you on the right path? Here are a few points to consider:
1. Thinking From A Hierarchical Or Engineering Perspective
The ideal physician-advisor needs to be able to communicate clearly with a multidisciplinary team of engineers, designers, data scientists and other skilled individuals who might not have a medical background. It is critical to understand what information a doctor needs and when and where to deliver that information in the doctor’s workflow.
Your ideal doctor-advisor will be able to convey medical knowledge and complex clinical information in an understandable manner to nonmedical professionals. Does your candidate think hierarchically? Can they break complicated pieces of information down into multiple levels of interconnected components? This is also known as engineering thinking; the expectation is not to dumb things down, but to break complex clinical concepts and processes into logical components, that, when given to smart, nonmedical individuals, make sense.
2. Having The Ability To Convert Data Into Actionable Information
Taking care of patients requires doctors to take action: First, they take in a wide variety of data points, and then they make a call on the next steps. In today’s world of “data overload,” doctors are bombarded with data from many sources and are often not able to process or understand data quickly enough to make the best decisions for their patients.
AI solutions that can process this data and quickly provide useful recommendations to clinicians can help; however, doctors do not generally like or trust information from “black boxes.” They want to know why a recommendation is made. The “why” behind a piece of information is a key component that makes that information trusted and actionable. For a product engineer, this is not easy to address. In my experience, it takes a smart medical professional to understand what makes information actionable and how that information is optimally delivered to be useful to other clinicians.
3. Enabling A Practical Mindset
Your ideal doctor-advisor will focus on what is actionable and practical and be insightful and helpful when criticizing your product, rather than concentrating on aspects that cannot be altered or are too theoretical. There are always many constraints in healthcare environments, and these can provide significant barriers to the ideal solution. Many (if not most) of these constraints will be outside of your technology’s and the doctor’s control.
It is an advantage to work with a doctor who understands the constraints and focuses on what is practical for the current state of the particular product’s use environment. This will help your product deliver practical value sooner, even if it might take much longer to address some of the surrounding constraints that are prohibiting the ultimate desired solution.
4. Knowing The ‘Emotional Landmines’
Healthcare is a profession with many emotional landmines, or topics that are sensitive and can upset doctors if not approached carefully. For example, I’ve found that many doctors can become upset if a product tries to replace their role in treating the patient, rather than helping them do their job more effectively.
The ideal doctor-advisor will understand the frustrations of other healthcare professionals in their field and convey many possible ways your product can assist with doctors’ daily workflow and better their patients’ lives — without offending them. However, it is your job to ask the right questions to get the honest answers you need to better market your idea and improve upon your technology. Together with your doctor-advisor, you will discover new ways to eliminate emotional landmines and solve frustrations inherent in doctors’ daily routines.
It is indeed rare to find doctors who meet these criteria and who are also interested enough in your product’s potential to dedicate their extremely limited time to help advance the project. However, if you find the right clinician for the job and they agree to assist you in your research, the best approach would be to build your brand from the foundation with them. It will be valuable to your product if you both understand the usefulness and relevance of your product from the very beginning. Building your product along with the doctor from the start can make the whole process run more efficiently and effectively with greater chances for success.