The World Health Organization projects a shortfall of 10 million health professionals by 2030, mainly in low- and lower-middle-income countries. According to the Association of American Medical Colleges (AAMC), the industry will face a shortage of 18K – 48K primary care physicians in 2019 – 2034. But why does this happen?
The supply of medical professionals cannot keep up with the demand due to several reasons: underinvestment in education and training in some countries, a large portion of doctors approaching retirement age, and the COVID-induced exodus, forcing clinicians to leave the field temporarily or permanently due to stress, risk, and frustrations.
Besides, clinicians quit jobs due to burnout and the issues connected with the EHR systems. According to KLAS research, around a third of doctors strongly disagree that initial EHR training prepared them well. Is it possible to improve custom EHR software to fight the scarcity? Various technologies can improve the EHR system’s performance and mitigate the painful moments of working with it.
The top technologies for this matter are AI, big data, and predictive modelling. So how can they help?
Artificial intelligence (AI)
An AI-enhanced EHR system can facilitate the work with the tool in two directions:
- Speeding up data extraction
With AI functionality, clinicians can reduce the time needed for the relevant data search and its extraction. The research published in JAMA in July 2021 featured the efforts of gastroenterologists from Stanford University Hospital (Stanford, California). The doctors and the AI-driven system had to analyse the electronic health records of random patients and deliver conclusions. Eventually, the AI-driven system saved 18 per cent of the time, while the quality of human vs AI-driven analysis was comparable. As for the participants’ feedback regarding the new feature, the overall majority (92 per cent) considered it positive and efficient.
- Facilitating document management
Clinical documentation comprises various data – clinical notes, medical images, and sensor data. Electronic health records hold it all. However, about 80 per cent of clinical data is unstructured and unsuitable for medical analysis. Enabling natural language processing can help with making use of this data.
Integrating NLP algorithms into the EHR system spares clinicians the need to fill in the EHR data manually. They can simply dictate the information, thus improving efficiency and reducing time and effort spent.
Besides, NLP can assist doctors in providing quality care. For instance, it can help to determine whether a follow-up is needed. NLP technologies help to recognise and classify various documents and the patient's medical track records. With a predefined checklist for assigning follow-ups, NLP technologies save doctors time, eliminating the need to manually go through multiple patient histories to ensure they do not miss their visits.
Big data analytics
Some social determinants, such as income, education level, employment status, and the need for social support can affect patients’ health immensely. Usually, this data comes from various government-supported and private sources (insurance companies, charities, and educational facilities). That data is raw and unstructured. This is where big data analytics can assist. Those solutions can extract unstructured data from many non-clinical sources, normalise it, and add it to patient profiles.
Uniting big data and electronic health records can help you improve various aspects of your clinic operation, namely:
- Fraud prevention. Enabling big data functionality within the EHR system allows users to identify patterns that may indicate fraudulent activity. This information can then be used to improve your clinic’s financial security.
- New therapy development. By analysing historical data with big data functionality, experts can identify patterns that highlight potential problems or areas for improvement. In addition, you can activate real-time alerts that help clinicians spot potential issues early.
- Developing efficient population health initiatives. By applying big data add-ons to EHR data sets, providers can perform efficient risk stratification. This effort allows clinicians to identify trends and patterns in patient populations.
For example, researchers from Carnegie Mellon University, Pittsburgh, Pennsylvania, employed big data analytics to detect seven subtypes in chronic kidney disease (CKD, patients. Moreover, the data gathered can help clinicians assess the relevant treatment decisions in terms of effectiveness and suitability.
Moreover, the amalgamation of big data analytics and health records data gave rise to the All of US project. This large-scale research program aims to promote precision medicine and accelerate research relying on EHR data.
The program welcomes representatives of all ethnic groups and sets up a connection between patient genomes and their phenotypes. Individuals share their health information and the relevant social determinants over time. After the submission, researchers study the data and deliver the insights to the participants. This research can help improve healthcare with better diagnostic tests or personalised treatments for different people.
Predictive analytics
Predictive analytics in healthcare is rapidly developing and helps providers increase cost-effectiveness and operational efficiency. The use of predictive analytics tools with EHR solutions can grant several benefits:
- More accurate diagnostics. With a predictive analytics tool on stage, doctors can upload a patient’s medical history into the tool, which then combs through the data and delivers the most likely diagnosis.
- Enabling preventive medicine. Clinicians can assess the risks of certain diseases in their patients and offer timely measures for prevention or severity reduction measures.
- Detecting at-risk patients on time. The tool can work on a larger scale – your patient populations. The add-on can dive into your EHR data to assess population health and identify at-risk patients. Such patient groups often need specific health plans and lifestyle recommendations.
Understanding and improving interoperability
EHR interoperability has long been a challenge in the industry. Today, several solutions can help solve this challenge:
- EHR data standardisation: The Fast Healthcare Interoperability Resources (FHIR) is to become a standard for EHR data worldwide. Observing FHIR allows different medical systems to share data seamlessly. However, the correct standard implementation is a challenge even for EHR developers. The thing is that implementations among partners should be compatible. This can be difficult to achieve if partners employ different vendors.
- Health information exchange (HEI). An HEI is a programmatic tool that lets providers, patients, and other partners exchange healthcare data securely. The HEI scale may differ. They can operate at the level of a city, a state, across neighboring states, or nationally.
Conclusion
Though most US healthcare providers employ EHR systems, their appeal to users still leaves much to be desired. About 30 per cent of clinicians even quit their jobs due to EHR-related issues. Is it possible to change the situation? We believe enhancing custom EHR software with some advanced functionalities can help.
Thus, enhancing the system with AI can accelerate the relevant data search, and the NLP module can spare clinicians the tedious manual work of filling in the slots altogether. With the big data analytics module in the system, providers can improve fraud prevention, streamline new therapy development, and enable risk stratification for efficient population health management. The predictive analytics module can help ensure swifter and more accurate diagnostics and timely prevention measures for individuals and patient populations.
It also makes sense to discuss the existing interoperability solutions with the vendor, select the one that suits you best, and have the vendor’s assistance in implementing it.
These measures for improving the EHR system performance can help refine doctors’ experience and reduce staff outflow in healthcare.
______
Inga Shugalo is a US-based Healthcare Industry Analyst at iTransition: Software Development Company.
