Technology, such as automated, speech-based clinical documentation, may play a role in reducing physicians’ IT workload
Aug. 20, 2020
This week, as part of SURF-IoT’s continuing series of talks with mentors, Kai Zheng, professor of Informatics spoke with SURF-IoT fellows about research trends in biomedical and health informatics.
Medical/health informatics is an emerging scientific field, Zheng says. “It deals with the collection, storage, retrieval, communication and optimal use of health-related data, information and knowledge. And, it utilizes the methods and technologies of social and technology sciences for the purposes of problem solving and decision making in healthcare.”
Zheng, also an adjunct professor of emergency medicine at UC Irvine and SURF-IoT mentor, discussed a number of contemporary topics that have received attention lately in the biomedical and health informatics research community, such as clinician burnout attributable to health-IT use, speech-based clinical documentation, and natural language processing.
Clinical burnout among medical professionals is a “work-related syndrome involving emotional exhaustion, depersonalization, and a sense of low personal accomplishment,” Zheng says. Over the past decade, the U.S. has suffered an epidemic of clinician burnout with estimates among physicians often exceeding 50 percent. A study published in JAMIA (Journal of the American Medical Informatics Association) states that at a time when paper charts are being replaced by computerized medical records, about 70 percent of electronic health records users report IT-related stress. “There appears to be a strong association with the introduction of health electronic systems in healthcare that apparently is associated with a much higher frequency of burnout,” Zheng noted. This can result in higher incidents of mistakes with medical diagnosis and treatment, and more physicians leaving the profession.
A major source of burnout is attributed to the vast amount of time and energy physicians spend in documenting patient information. To meet federal regulations, physicians can spend four to five hours a day inputting this data into healthcare record-collecting systems, he says.
Zheng believes that technology, such as automated, speech-based clinical documentation, can play a role in reducing physicians’ IT workload. Recently developed, speech-based clinical documentation systems are designed to automatically generate text transcripts that are converted into clinical documentation. “In theory this should work pretty well. You put some microphones in the exam room to pick up the conversation,” he says. But there’s a lot of challenges of how to make this work technologically. There are also privacy and compliance concerns that need to be resolved, Zheng adds.
“Natural Language Understanding (NLU) is another domain that has received a lot of attention,” Zheng says. In the field of artificial intelligence, NLU is considered an AI-hard problem – the most difficult.
NLU utilizes algorithms in an attempt to parse text into syntactic and semantic categories, and to recognize concepts. The ability to extract richer data beyond diagnosis codes, laboratory and fiscal data is seen as having immense value for clinical care and medical research.
Medical text is particularly difficult to process, he says. “Two physicians reading the same piece of documentation may not agree on what is documented. There is frequent use of acronyms, abbreviations and ambiguous terms. For instance, CA. Does that mean California or cancer? That is very challenging for computers to understand the meaning of the text based on the surrounding context.”
Zheng’s SURF-IoT research project, ” Understanding Patient Questions About Their Medical Data and How Such Questions are Answered through the Collective Wisdom in Online Patient Communities” will use a qualitative content-analysis approach to understand patient questions about their medical records, and specifically about laboratory test results and medications, and the answers that they receive and find helpful.
As patients have greater access to their medical data, research suggests they often have difficulties understanding the information contained in the records. In order to successfully use the available data for decision making, patients often tap into supplemental sources of knowledge such as online health forums.
This content is usually posted publicly, which provides an opportunity for researchers to understand the nature of patients’ questions about their clinical data and about their own health. It offers direct evidence of comprehension issues, as well as additional forms of support. Similarly, the nature of the answers to these questions, especially those that patients find useful, suggests the support those patients may gain through the forum discussions, he says.
Although the insights from these analyses may have broader implications for patient education and patient-provider communication, we are particularly interested in opportunities to improve the design of the technologies that give patients access to this clinical data,” Zheng says.
SURF-IoT 2020 talk series is presented via ZOOM and open to the public.
To view a schedule of future speakers and register for attendance, visit events.
Watch Prof. Zheng’s presentation here.
– Sharon Henry
