Extracting data to deliver value-based care

Founded in 2016, Precision Driven Health is a research partnership between New Zealand’s health IT sector, health providers and universities, aimed at improving health outcomes through data science. Ahead of the HIMSS20 conference, Kevin Ross, CEO of the organization, talks about the significance of data in delivering value-based care and the real challenges of collaboration between different stakeholders in data.

What are your thoughts on value-based care (vs the traditional fee-for-service model) and what can be done with healthcare data to create more value and personalized care for patients?

Value-based care is far preferable to a fee-for-service model, as long as we agree on value, and can be transparent in our approach. At its extreme, a fee-for-service model rewards someone who takes many attempts to get it right, whereas value-based care rewards the delivery of what matters in the most efficient way. Data has a huge role to play in this, particularly in helping to define and analyze the most appropriate metrics for value. 

Data allows us to use a common language for value, and can make transparent the link between the activities undertaken in healthcare, and the value that is received by consumers. When data can be shared, people can see more clearly what they are receiving, and what is having the greatest impact. A fee-for-service tends to lead toward giving the same care for everyone, as we get good at efficiently delivering a particular procedure or program. 

Data science allows us to analyze what works for different people, and if the objective is to deliver value, then the world of possibilities opens up. One person may benefit from medication, while another benefits from improvements to their living circumstances; value-based care gives us a framework to consider these in a common way.

As CEO of Precision Driven Health, what are some of the lessons learnt through coordinating collaborations between health professionals and data scientists in applying data science to enable precision health to become a reality?

Everyone is keen to collaborate and learn from each other, but there is often a lot of work up front to establish common language and understanding. We come with our biases, and are often unaware of how much they influence our thinking and expectations. I find that most people oversimplify the role of another collaborator and assume that it will fit into their own way of viewing the world. 

For example, a data scientist thinks of a clinician as having a very transactional role of translating data into a diagnosis or next decision. While this may describe some of the core information exchange in clinical care, it is in reality only a small part of what happens – and much of what happens is not captured in tidy data science language. Similarly, our clinicians often think that data science is one of two extremes – either magic (just throw some data into a pot and out will come an answer), or predictable (the data will show that if we change X then Y will be improved). A good data science process involves a lot of workshopping, trial and error, human input and compromise.

What are your observations on the development of precision health in New Zealand? What opportunities do you see?

New Zealand has excellent quality data, enabled through a single identifier per person, known as the National Health Index. This allows twenty years of digital records to be linked and analyzed for historical patterns. Combining this data about a diverse population with excellent data scientists, clinicians and software developers, New Zealand is well positioned to be a testbed for innovations in precision health. Having a single-payer health system also helps to ensure that the best care for an individual usually corresponds with economic benefits to the whole system. 

However, despite all of these advantages, the task is still very difficult. Healthcare changes slowly, and we must build public trust for/in new uses of data, and new ways of delivering care. Our clinical carers are not trained in data science, and our data scientists are new to healthcare – so the changes that are readily seen in other industries have not yet taken hold on a large scale.

How do you think the use of healthcare data has led to better health outcomes in New Zealand? Could you give us some examples?

New Zealand has linked healthcare data to other government service data, including housing, education, justice. This has allowed us to understand the link between health and other social circumstances. As one example, a local health board has invested in improving the quality of homes in the neighborhood, to reduce the rates of respiratory issues caused by mold. 

At the other end of the system, we have developed a localized calculator for the outcomes from elective surgery, developed based on all surgeries in the past ten years. We can now use this information for anyone considering a procedure, and explain what the outcomes are like for people like them. For minority groups who are not represented well in international studies, this allows a much better-informed conversation between surgeon, patient and family that helps to make the best long-term decisions.

Kevin Ross is a panelist at the HIMSS20 AsiaPac Summit’s Panel Discussion titled Preparing Data and Redesigning Metrics for Value-Based Care, scheduled for Wednesday, March 11 from 8.30am to 9.30am in room W240C. Additional registration is required – click here for more details on the HIMSS20 AsiaPac Summit.

He is also the speaker for the session titled Ethical Machine Learning scheduled for Tuesday, March 10 from 1.30pm-2.30pm in room W207C. 

Source: Read Full Article