Years after the shock of the COVID-19 pandemic, healthcare organisations still lack qualified personnel. The World Economic Forum estimates a projected shortfall of 10 million healthcare workers worldwide by 2030. Decision-makers at healthcare facilities hope to address the medical staff shortage with artificial intelligence.
Electronic health records systems are the most widely used healthcare-specific software in nearly all US hospitals. Healthcare specialists have access to a wide variety of EHR system features that cover the needs of multiple medical specialities. At the same time, clinicians often complain that electronic document management takes up a large portion of their time, as they usually need to fill in multiple forms. This leads to reduced time clinicians spend communicating with patients and increased overtime and burnout. Artificial intelligence can help solve this particular problem and assist clinicians in other aspects of their everyday tasks.
Patient data collection and structurisation
Writing comprehensive clinical notes and filling in patient forms are routine and time-consuming activities that medical professionals have to do every day. AI-powered tools can streamline both of these processes.
For example, EHR systems with voice recognition technology allow healthcare professionals to automatically input clinical notes into the system. Thanks to powerful voice-to-text capabilities, medical personnel can focus on patients without being distracted by typing notes on the computer, enabling natural conversation flow. After the consultation, clinicians only need to check the text captured by AI to avoid mistakes.
The system then can perform a routine form-filling task instead of clinicians: it detects and extracts important patient details from clinical notes, which doctors take during the consultation, and then fills out necessary forms with patient data. Relevant data can also be extracted from other sources of information. Such sources can include patient portals, health and wellness apps, and insurance contracts. However, it is crucial to remember that personal information processing cannot be executed without prior consent from the patient, regardless of its source.
Clinical data analysis for diagnostic and treatment support
Many EHR systems include a native analytics and reporting module for everyday clinical decision-making support. If such a module is enhanced with artificial intelligence, healthcare specialists can use it for a variety of purposes, including:
- detecting recurrent patterns in health metrics and behaviours;
- finding anomalies in patients’ vitals or symptoms;
- classifying the condition according to the International Classification of Diseases;
- detecting possible harmful interactions of medications;
- projecting the course of the particular disease or chronic condition;
- predicting possible health outcomes for different treatment plans.
AI can sift through patient data in minutes instead of the hours doctors spend finding and analysing the same amount of information. The findings are then presented in a user-friendly format to medical specialists who make diagnostic and treatment decisions.
The system can also review treatment plans and compare them with industry best practices and patient medical information. Any potential conflicts will be flagged to alert a treating clinician. An example of such a conflict would be a physician prescribing medication mentioned in a patient’s medical profile as an allergen or already prescribed by a different specialist. This approach helps doctors make more accurate decisions faster, personalising patient care and improving patient safety.
Virtual assistants for enhanced patient engagement
Currently, most EHRs have patient portals that allow patients to manage their health information, as well as view test results and treatment plans, get medication prescriptions, and find answers to frequently asked questions. Sometimes, such portals facilitate doctor-patient communication via chat.
Supplementing patient portals with intelligent chatbots can help patients receive a more personalised experience, easily find relevant information, and get help from a medical professional faster. AI-based bots can interact with a patient via chat, providing the requested information in a natural, easy-to-understand format. They can also extract valuable patient details from the conversation and input them into the EHR system. Intelligent bots can notify medical professionals if they detect key words or phrases in conversations, indicating that the patient's condition is dangerous. Such chatbots free up healthcare organisations’ contact center personnel and medical professionals, while ensuring that each patient receives attention and timely care.
So, what does the future hold for AI-supplemented EHR solutions?
Most healthcare organisations are at the beginning of their journey to implement AI-enhanced EHR software. While the benefits of such solutions are undeniable, there are still many challenges associated with their adoption.
Data privacy and security remain a major concern in the healthcare industry. While software providers must deliver EHR systems supplemented with robust security capabilities to withstand modern cyber threats, healthcare organisations must use such software according to the most recent security guidelines and regulations. Eliminating bias in AI algorithms is yet another challenge. It is crucial to avoid inaccurate diagnostic or treatment decisions due to the poor performance of AI models.
Following explainable AI principles is the best way for EHR software developers and healthcare organisations to minimise the possibility of AI-related security breaches and biased treatment. Finally, the cost of AI-enhanced EHR system implementation remains high. Healthcare organisations with limited budgets have to carefully evaluate which AI-enhanced features they want to enable within their EHR system. They often need experienced consultants’ assistance to optimise implementation costs and gain the most value from their software.
On a brighter note, AI capabilities will surely expand. More sophisticated algorithms will lead to minimised bias, better data analysis, accurate risk prediction, and personalised treatment approaches. As AI becomes a common part of EHR solutions, it will likely become more affordable. This, in turn, will improve healthcare professionals’ working conditions, positively influence their job satisfaction, and help provide better care to patients.
Mariia Kovalova is a Healthcare Technology Researcher at Itransition, a custom software development company headquartered in the US.
