Physician, Heal Thyself with AI for Personalized Healthcare

Healthcare is at an interesting turn. While patients expect personalized care, physicians suffer severe burnout, mostly due to their administrative work. A survey indicated that almost 50% of physicians’ time is spent in updating Electronic Health Records (EHRs). Even when engaged directly with patients, they spend 52.9% of their time with EHRs and other activities.

Today, many across healthcare view AI as a potential time saver to ease physicians’ burden and provide more personalized patient care. This makes perfect sense, given that the healthcare industry overall possesses characteristics that could benefit from AI’s inherent ability to quickly process and produce human-like responses: highly regulated, massive volumes of structured and unstructured data, the need for accuracy and speed, and a wide array of interested stakeholders (not to mention that people’s health outcomes are at stake).

Let’s review the ways where AI’s value in healthcare is coming into focus.

Diagnosis and Decision-Making

Digital retailers provide personalized recommendations based on users’ past purchases and shopping preferences, based partly on AI algorithms that work in the background. There are similarities with AI’s potential in healthcare, as  AI can process high volumes of data points based on a patient’s medical reports, which is a daunting prospect for humans to take on alone.

Based on data, AI unveils patterns that help physicians diagnose and make the best decisions for patients. However, researchers warn of biases and errors in the AI-served data that can generate responses based on stereotypes and falsehoods. If AI algorithms used in healthcare are not representative of the communities they serve, the data produced may impact under-represented communities and the ones most in need of quality care. Health workers must ensure that AI algorithms are representative and that the proper data is sourced to provide accurate and fair outcomes.

Keyboard Time

Some physicians are already using AI for dictating notes, scheduling appointments, follow-ups, and digitization of medical records. Studies show that physicians who use AI for dictation and medical scribe services spend more with patients, allowing them to provide an elevated level of personalized care and consultation.

However, physicians cannot base their decision-making on transcriptions alone —  human beings don’t communicate solely through words. A lot is communicated through facial expressions and gestures during physician-patient interactions that physicians can also record. Physicians, therefore, would need to make changes in transcripts following those interactions, when required, to capture the full condition of their patients in the proper context.

Enhancing Physician Usability

Physicians are very busy people. AI-powered products can offer time-saving features and capabilities, but only if they are easy to use and intuitive. For example, some products help physicians prepare reports based on transcriptions, audio, and images quickly and efficiently, but physicians should know instantly where to click and generate reports.

It’s here that experiential designers come in. They design products not from their perspective but from that of users, ensuring that users will get the most out of their products and achieve their desired tasks. Experiential designers in healthcare follow a rigorous process from concept to design that itself can be aided by AI:  

Understanding the problem

Identifying the problem statement is the first step. Designers talk to business leaders and other stakeholders to gain insights into their priorities and pain points. They list several problem statements before finalizing one or two that an AI-powered product or product feature can address. AI can aid designers in several ways:

• Clarify the problem statement: Several AI tools help with brainstorming the problem area and give suggestions around the “How Might We Solve” statements.
• Provide opportunity scores: By offering scores for opportunities in problem statements, they help designers finalize the problem statement so that it aligns with the desired goals.  

Undertaking research

Experiential designers place a great deal of emphasis on understanding users’ pain points, aspirations, and requirements. In this case, users are physicians, nurses, and clinic staff. While each  may serve a different role in the healthcare ecosystem, their regular work is connected through a similar environment. Experiential designers may follow a combination of quantitative and qualitative research methods, such as surveys and one-on-one interviews, aided by AI:

• Research brief: AI tools can prepare research briefs to help with research.
• Interview scripts: They can prepare interview scripts for designers, which they can refer to while interviewing users.
• Transcribe user sessions: AI transcribes user sessions for designers, saving a massive amount of time. However, researchers warn that designers must not depend entirely on transcripts, as users may not verbalize everything. Designers must observe how users work, react, and consider these observations in their decision-making.
• Process data: AI can process massive data and identify usage patterns that help designers’ make decisions about final designs. Researchers caution that error and subjectivity can creep into AI-processed data and designers must validate the data through research before proceeding.
• Experiential designers also conduct market research to uncover competitor information. ChatGPT can research and share information on competitors and their products (but designers must validate this research before leveraging it as part of their processes).

Creating personas and journey maps

Insights from user research help designers create personas that will further tailor products to clinicians in specific roles and disciplines and with different responsibilities. Further, it helps them design journey maps to explore how to provide physicians with enjoyable experiences as they work with the AI-powered product. AI can assist in the creation of personas and also images based on text instructions. Designers can also use AI tools for wireframes and mockups.

Collectively, AI can be a critical tool for physicians to save time by automating most of their day-to-day and mundane administrative tasks. Additionally, AI can assist in diagnoses and decision-making when used in the proper context and validation. Finally, AI can also serve experiential designers in creating engaging AI-products specifically for physicians and others in the healthcare ecosystem.

At Persistent, we have a focused methodology called Digital Greenhouse, which is a human-centered, design-led approach to speed up digital initiatives and create a clear vision with ready healthcare prototypes in just four to six weeks. To jumpstart your digital journey with Persistent and learn more, click here.