The importance of efficiency, speed and productivity in healthcare delivery is indisputable. While the use of electronic health records (EHR) aims to help clinicians meet these demands, it has yet to optimize their productivity in a significant way. A US study by the American College of Physicians found that doctors spend an average of 16 minutes per patient on EHR functions across specializations. This is a concern for the US healthcare system, as extended care delivery times translate into higher costs for patients as well as physician burnout and job dissatisfaction.
A slew of healthcare and technology firms have stepped up to address this challenge and more through AI. In this article, we explore how exactly AI is making EHR systems more efficient.
1. Reducing administrative burden of clinical documentation
A study by the American Medical Association (AMA) and the University of Wisconsin, shows that nearly 50% of clinician time is spent on admin work, including documentation, order entry, billing and coding, and system security. Not only does this eat into valuable patient time, but also contributes to excessive work–life imbalance, dissatisfaction, high rates of attrition, and burnout. Automated capturing of clinical notes through natural language processing (NLP) reduces clinician admin work, freeing up more time to focus on patients.
AI-based speech-to-text technologies can help ease these pressures by minimizing much of these administrative tasks. EHR solutions embedded with an AI layer can document patient problems, diagnoses and procedures in compliant formats through voice-based commands. These smart EHR solutions make it easier to find specific patient information and even help clinicians convert their narratives into actionable information for real-time decision making.
Examples of firms that work in this field:
- 3M M*Modal
- Robin Healthcare
2. Automated extraction of patient information
Patient data needs to be easily accessible to providers for faster diagnosis and decision-making. Moreover, it should be clear and easy to read for clinicians to interpret the data accurately. However, sorting through large amounts of EHR data and picking the bits that apply to a patient’s condition is a huge challenge. In fact, 41% of hospitals reported that public health agencies’ inability to effectively receive patient data was one of the major barriers physicians faced during the COVID-19 pandemic.
AI-enabled EHR systems allow clinicians to rapidly access, extract and electronically export patient data with minimal error. For instance, at OneMedical, HCPs extract data from clinical documents using Athenahealth’s AI-enabled cloud-based EHR. Flatiron Health’s human “abstractors” can review provider notes and extract structured data, using AI to recognize key terms and reveal data insights.
Additionally, Amazon Web Services recently launched a cloud-based service known as Amazon Comprehend Medical, that can retrieve and index data from clinical notes. Other companies working in this area include Concord Technologies, Innodata Inc., and Intellidact AI.
3. Enhanced clinical decision-making
The humongous data created by EHR systems lends itself well to advanced AI and machine learning tools to uncover patient insights, predict high-risk conditions, and enable more personalized care. For instance, Epic Systems’ AI solutions help predict hospital readmissions, patient mortality and risk levels, sepsis, hospital-acquired diseases, patient deterioration among other potential dangers.
AI solutions that learn from new data and enable more personalized care are being developed by companies such as Google, Change Healthcare and
AllScripts. Google researchers are teaming up with healthcare networks to develop prediction models using big data to provide HCPs with alerts about serious conditions such as sepsis and heart failure. Google Cloud’s Vision AI also uses AI-derived image interpretation algorithms to extract health insights, and trigger real-time actions based on it.
Healthcare tech company Jvion has even developed a ‘cognitive clinical success machine’ application. It identifies patients at the highest risk for an adverse clinical condition or development and identifies patients who are most likely to respond to treatment.
4. Achieving EHR interoperability
The lack of a common standard for capturing, managing, and transmitting patient data makes it difficult to analyze information from different EHR systems. Without syntactic and semantic interoperability, there’s always a risk of information getting lost when shared digitally with multiple health providers. This creates delays and inaccuracies, ultimately affecting the quality of care and patient outcomes.
Some organizations have started to use AI and ML to solve interoperability issues in clinical documentation. Cerner’s AI platform uses AWS to allow for cloud-based EHR infrastructure interoperability with outside data sources. Beth Israel Deaconess Medical Center (BIDMC) is using AI and ML to ensure medical forms are completed ahead of surgery, and subsequently notifying nurses if the paperwork is missing.
The way forward
COVID-19 has put a greater focus on patient-centric care, which is only possible through smarter and more efficient EHR systems. The prolonged impact of the pandemic has also prompted the need for telehealth compatibility in EHR systems. The use of AI in EHR can make this integration seamless for both patients and HCPs.
While some healthcare networks are relying on EHR vendors to integrate AI capabilities to enhance specific application areas, others are also developing in-house AI solutions for their EHR systems. In order to expand their horizons, some EHR vendors are looking partner with technology firms to harness their AI tools.
Netscribes’ market research and innovation insights solutions help you keep track of the latest developments in the healthcare industry and provides data-driven insights to support strategic decisions. Contact us to know more.