Tapping into medical knowledge through an ambient user interface

by Jonas Knudsen

Medical knowledge has been expanding exponentially. The doubling time — or the pace at which knowledge expands — was estimated at 7 years in 1980, 3.5 years in 2010, and a projected 73 days by 2020, according to a 2011 study in Transactions of the American Clinical and Climatological Association.

The dilemma for clinicians is how to keep on top of every new piece of information. Not only does clinical research and evidence-based knowledge increase exponentially, but so too does the amount of personal information healthcare providers capture about the patient. Additionally, data from new sources — for example, Fitbits, Apple Watches and other health apps — are being integrated into the patient data ecosystem, including data from sources within the hospital and from medical devices.

With so much data from so many sources, there’s an urgent need for new ways to interact with clinical records. The traditional point-and-click interface developed by Microsoft 30 years ago as an administrative environment has been out of date for many years. In fact, it was never really intended to be adopted by industry as the primary way to interact with computers and applications. And in healthcare, where there are vast amounts of data, and healthcare professionals such as doctors and nurses are highly mobile, hygiene is a major factor (just think about the bacteria on keyboards and the mouse, which becomes an even bigger issue when computers are used by many staff members in a patient room) and speed and precision are crucial — the mouse, keyboard and manual data input just aren’t efficient.

Let’s take the clinical process during which doctors capture patient information based on a specific examination. From paper to manual computer input to digital dictation and speech recognition, the actual user interface the doctor uses to access the information changed very little. Most applications have been designed to support a methodical process of search, filter and access. But in a clinical environment, that three-step process is often repeated several times to get an overview of all the relevant information about a patient.

The challenge is twofold. First, clinical applications do not have adaptive user interfaces that can help search, filter and present relevant context-based information automatically based on the patient-doctor conversation. Second, the technology used to document data in most clinical applications does not support the use of automation to transform a conversation to a highly structured data format. That is one of the primary reasons that speech recognition hasn’t fulfilled its promise and has not been widely adopted in healthcare.

All of this means that clinicians are forced to spend time laboriously documenting information from their interaction with patients, and this the time they have to do what they are trained for, which is to deliver high-quality care.

Changing the dialogue through an ambient interface

The way to address these challenges is to deploy an ambient interface, which empowers clinicians to interact with each other through a digital environment that is aware of their presence. An ambient interface connects smart systems with users and is sensitive, adaptive and responsive to their needs, habits, conversations and gestures. Ambient interfaces offer greater interactivity and easier-to-find and more-relevant information about the individual clinical needs of patients, enabling clinicians to provide high-quality care and help patients achieve their objectives.

The ambient user interface offers relevant and context-based information to the doctor, based on machine learning algorithms. Once the interface has been configured, the doctor can determine the relevance of the data presented and the algorithms will adapt and learn. In this way, the clinician ensures he or she is presented with the data relevant to each patient and in the right context. Another benefit of an ambient interface is its speech recognition capabilities combined with machine learning that make it possible to deconstruct a doctor/patient conversation, identify relevant key words and arrange them into a predefined template. Rather than spending hours manually capturing the conversation, all the clinician needs to do is handle quality assurance and perhaps add a few terms to increase the granularity of the captured information.

The healthcare environment is becoming more challenging, putting clinicians under greater pressure and requiring organizations to find ways to improve the way they manage patient data and reduce costs. In this environment, hospital executives ­­— chief executive officers, chief information officers and chief marketing officers — must adopt a new approach to accessing and documenting data in digital applications. To that end, they must:

  1. Recognize that the user experience and the user interface are vital components in addressing strategic goals such as increasing operational efficiency, clinical quality and staff satisfaction.
  2. Include access to data and the adoption of the ambient user interface into their data management strategy.
  3. Require from vendors that digital applications of tomorrow are designed specifically for healthcare, not only in terms of functionality but also in design and interface.

With the right approach — one that recognizes the needs of clinicians and the expectations of patients — healthcare will be able to adapt and respond to the rapid rate at which medical knowledge is expanding.


Jonas Knudsen is part of the global health advisory group at DXC. He advises on planning and implementing effective digital health transformation initiatives to drive integrated and personalized care innovation. He was previously CIO of the New University Hospital Odense – Denmark (1200 beds) and head of IT strategy and planning in a EUR 1B hospital construction project. Previously associate professor of IT management at the University of Southern Denmark. Jonas holds a certification as an enterprise architect in TOGAF. Follow @JonasHKnudsen

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: